Next Article in Journal
Composite Quantum Coriolis Forces
Previous Article in Journal
Performance of a New Sixth-Order Class of Iterative Schemes for Solving Non-Linear Systems of Equations
Previous Article in Special Issue
Optimality of a Network Monitoring Agent and Validation in a Real Probe
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Model Predictive Control of Parabolic PDE Systems under Chance Constraints

1
Group of Process Optimization, Institute for Automation and Systems Engineering, Technische Universität Ilmenau, P.O. Box 100565, 98684 Ilmenau, Germany
2
German Research Chair, African Institute of Mathematical Sciences (AIMS), KN 3 Rd, Kigali, Rwanda
*
Author to whom correspondence should be addressed.
Mathematics 2023, 11(6), 1372; https://doi.org/10.3390/math11061372
Submission received: 13 February 2023 / Revised: 5 March 2023 / Accepted: 7 March 2023 / Published: 12 March 2023
(This article belongs to the Special Issue Stochastic Control Systems: Theory and Applications)

Abstract

:
Model predictive control (MPC) heavily relies on the accuracy of the system model. Nevertheless, process models naturally contain random parameters. To derive a reliable solution, it is necessary to design a stochastic MPC. This work studies the chance constrained MPC of systems described by parabolic partial differential equations (PDEs) with random parameters. Inequality constraints on time- and space-dependent state variables are defined in terms of chance constraints. Using a discretization scheme, the resulting high-dimensional chance constrained optimization problem is solved by our recently developed inner–outer approximation which renders the problem computationally amenable. The proposed MPC scheme automatically generates probability tubes significantly simplifying the derivation of feasible solutions. We demonstrate the viability and versatility of the approach through a case study of tumor hyperthermia cancer treatment control, where the randomness arises from the thermal conductivity coefficient characterizing heat flux in human tissue.

1. Introduction

Partial differential equations (PDEs) are models commonly used for describing spatial and/or temporal variations in natural, physical, and industrial processes [1,2]. Conventionally, the parameters involved in PDEs, e.g., thermal conductivity, diffusivity, viscosity, charge density, and material density [1,3], are taken to be deterministic coefficients that are sometimes prescribed from experiences or fixed through experimental measurements. Nevertheless, such nominal values do not precisely reflect the nature of the parameters, since they fluctuate in the real process and, hence, are uncertain. In addition, the operational environment of the physical system and uncontrolled external sources, e.g., ambient temperature and pressure, can also impose a significant impact. These uncertainties constitute what is known as input uncertainties (or random data), which propagate through the system and incur discrepancies between the real (actual) responses and the outputs predicted by the nominal PDE model.
Various simulation-based uncertainty quantification (UQ) methods have been developed for capturing propagated uncertainties from the inputs to the outputs (state variables) of a random PDE system [4,5,6,7,8,9,10,11,12,13,14,15,16,17]. Such studies have recently triggered various investigations on the mathematical analysis, numerical solution methods, and applications of PDE optimization problems with input uncertainties [11,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33]. In particular, studies on the state-constrained optimization of PDE systems with random data constitute the state-of-the-art. For such problems, an appropriate definition of inequality constraints of random state variables is basically a perquisite to render the optimization problem as well-defined. Hence, recent studies [19,20,30,33] considered chance constrained PDE optimization problems, i.e., to impose chance constraints on the state variables. However, these studies were focused more on elliptic PDE systems. Owing to the wide-spread use of parabolic PDEs in various engineering applications with distributed parameter systems and time-dependent behaviors, studies on optimization problems with state-constrained parabolic PDE systems involving random data are necessary.
Model predictive control (MPC) provides advanced control strategies for systems with complex state-space models involving control and state constraints. Recent studies show that the MPC scheme has further potential uses in engineering applications with distributed parameter models, such as parabolic PDEs, e.g., thermal processes [34,35,36] and process engineering applications [37,38,39,40,41,42], etc. Nevertheless, these studies investigated deterministic MPC without considering uncertainties. MPC for state-constrained distributed parameter systems with random data has not yet been well-studied.
Since reliable predictions are heavily dependent on the accuracy of the underlying model, a deterministic MPC scheme is bound to lead to infeasible state trajectories with inappropriately designed control strategies. Therefore, this work studies the model predictive control of processes described by parabolic PDE systems with random parameters. We propose a chance constrained MPC scheme for the control of such systems with random coefficients by imposing probabilistic measures for the reliable satisfaction of the state constraints.
The MPC of parabolic PDE systems with chance constrained state variables calls for the solution of a chance constrained PDE optimization problem on each prediction horizon. For a numerical solution, we transform it through an appropriate discretization into a high dimensional chance constrained optimization problem. Although many studies have been made to develop solution approaches [43,44,45,46], open challenges remain in particular for addressing large-scale complex problems. In general, there are three major difficulties in solving chance constrained problems on finite or infinite dimensional spaces: (a) the probability function of a chance constraint can be non-convex and non-differentiable, (b) this function is usually hard to evaluate, and (c) the feasibility to the chance constraint can be hard to verify. Even the knowledge of convexity and differentiability may not ensure a simpler evaluation of chance constraints. Therefore, for complex models, chance constraints can only be evaluated through approximation methods.
In this study, we focus on the development of a numerical solution approach to chance constrained MPC of semi-linear parabolic PDE systems. Hence, to circumvent potential non-smoothness and inherent computational difficulties associated with chance constraints, we extend the applicability of our smoothing inner–outer approximation method [47,48] to chance constrained MPC of such PDE systems. The proposed approach entails two important advantages. First, on each prediction horizon, the chance constrained problem is approximated by two computationally amenable smooth (inner and outer) optimization problems. The respective feasible sets and optimal solutions of the smoothing problems are convergent to the feasible set and solutions of the chance constrained problem (see [20,47,48] for detailed theoretical analysis). Second, the approximation scheme automatically generates probability tubes, whereby the inner approximation guarantees feasibility of the predicted state trajectories to the chance constrained MPC problem. In this way, the effort for constructing a probability tube, which is the case in traditional literature concerning tube-based stochastic MPC [49,50,51,52,53,54], is avoided.
One of the important areas to deal with challenging PDE models is medicine, e.g., physiologically based pharmacokinetic modeling for drug development [55], hyperthermia treatment in cancer therapy [56,57,58] and medical image analysis [59,60]. To demonstrate our approach, we extend the PDE-constrained optimal control problem of hyperthermia treatment by considering uncertain parameters, imposing chance constraints, and applying the receding horizon strategy for computing optimal as well as reliable control strategies for cancer therapy.

2. Problem Statement and Assumptions

2.1. Problem Statement

We consider a chance constrained model predictive controller (CCMPC) of a semi-linear parabolic PDE system:
( CCMPC )
min u J ( u ) = E t ^ z t ^ z + H y ( u , ξ ; t , · ) y d ( t , · ) L 2 ( D ) 2 + λ 2 u ( t , · ) L 2 ( D ) 2 d t subject to :
y t x κ ( ξ ) x y = f ( u , t , x ) in Q × Ω ,
κ 0 x y · n = g ( y , t , x ) on ( t ^ z , t ^ z + H ] × D × Ω ,
y ( t ^ z , x ) = y t ^ z ( x ) in D ,
P r y ( u , ξ ; t , x ) y m a x α , in Q ,
u m i n u ( t , x ) u m a x , in Q ,
t ^ z + 1 : = t ^ z + Δ t , z = 1 , , N t ,
where D R n ( n = 1 , 2 , 3 ) is a bounded spatial domain with boundary D ; x D represents the spatial position; ∇ denotes derivative with respect to x; n stands for a unit outward normal vector to D ; the interval ( t ^ z , t ^ z + H ] is a prediction time-horizon with a finite length H; Q : = ( t ^ z , t ^ z + H ] × D ; ξ Ω R p is a vector of random input variables from a complete probability space Ω , Σ , P r with σ -algebra Σ and probability measure P r ( · ) associated with a given continuous probability density function ϕ ( ξ ) , which can be Gaussian or non-Gaussian distributed. Moreover, E [ · ] represents the expected value operator with respect to ϕ ( ξ ) .
κ ( ξ ) is considered as a random coefficient in the PDE system with nominal value κ 0 = E [ κ ( ξ ) ] , where ξ represents model parameter uncertainty. The system is driven by a distributed forcing term f ( u , t , x ) with a control variable u. Equation (3) describes the interaction of the system with the surrounding environment through the boundary D , which is specified by the boundary function g ( y , t , x ) .
The control variable u is deterministic and belongs to the set U = { u L 2 ( Q ) | u m i n u u m a x } . The state variable y depends on the control u and the random parameter ξ , which is expressed as y ( u , ξ ; t , x ) , ( t , x ) Q . Since y is a random variable, the chance constraint (5) specifies the probability of holding restrictions on the state variable, where α 0 , 1 is a user-specified level of reliability. Since the chance constraint (5) is required to hold point-wise over the spatial domain D at each time instant t ( t ^ z , t ^ z + H ] , this formulation has the advantage that the user can selectively specify higher or lower reliability levels at given locations x and instant of time t.
Associated to the CCMPC defined above, we further define
  • the probability function
    p ( u ; t , x ) : = P r y ( u , ξ ; t , x ) y m a x ,
  • feasible set
    P = { u U | p ( u ; t , x ) α , ( t , x ) Q } .

2.2. Assumptions

Although formal mathematical analytic investigations are not provided here, the discussion needs to be framed in an appropriate function space. Moreover, to render the problem as well-posed as well as to validate the numerical solution approach, the problem setting needs to satisfy standard assumptions and the relevant function spaces need to be defined.
  • Let L 2 ( Q ) be the space of square-integrable functions over Q, then define
    F : = v : Ω L 2 ( Q ) , v is measurable , v F : = Ω v ( ξ , · ) L 2 ( Q ) ϕ ( ξ ) d ξ < + .
  • Let L 2 ( S ) be the space of square-integrable functions over S : = ( t ^ z , t ^ z + H ] × D , then define
    G : = v : Ω L 2 ( S ) , v is measurable , v G : = Ω v ( ξ , · ) L 2 ( S ) ϕ ( ξ ) d ξ < + .
  • Using the Sobolev space H 1 ( D ) and the Hilbert space W : = L 2 Ω , H 1 ( D )
    V : = v L 2 Ω , H 1 ( D ) v W 2 = Ω D v 2 + κ | v | 2 d x ϕ ( ξ ) d ξ < + .
The following assumptions are standard in the literature for PDE systems with a random parameter.
A1: 
Uniform ellipticity condition. There are positive constants κ m a x , κ m i n with κ m a x > κ m i n , such that the random coefficient κ satisfies
κ m i n κ ( ξ ) κ m a x , almost everywhere .
A2: 
For each fixed u U , the forcing term f has the property that f ( u , · , · ) L 2 ( Q ) .
A3: 
For each fixed u U , the boundary function g ( u , ξ ; t , x ) = g ( y ( u , ξ ; t , x ) ; t , x ) has the property that g ( u , · , · , · ) G .
Here, A1 is a realistic assumption from a practical point of view. It is a crucial assumption also to guarantee that the underlying differential operator of the PDE system is uniformly elliptic. Thereby, under the function spaces defined and for the problem settings specified in assumptions A1–A3, the parabolic PDE system Equation (2) with boundary (3) and initial Equation (4) can be guaranteed to have a unique weak solution, which is continuously dependent on f and g [61].
Hence, we assume that, for each pair ( u , ξ ) Q × Ω , the parabolic PDE system Equations (2)–(4) hve a unique solution y ( u , ξ ; t , x ) , ( t , x ) Q , so that y ( · , · ; t , x ) , ( t , x ) Q is continuous with respect to ( u , ξ ) . Moreover, it is assumed that y ( u , · ; t , · ) L p ( Ω ; V ) , where V is a Banach space, e.g., V = L 2 ( D ) or V = H 0 1 ( D ) , and
y ( u , · , t , · ) L p ( Ω ; V ) p = Ω D y ( u , ξ ; t , x ) p d x ϕ ( ξ ) d ξ = E y ( t , · , ξ ) V p <
for each t > 0 .
Over each prediction horizon of the MPC scheme, the chance constrained optimal control problem (CCPDE) with the random parabolic PDE system Equations (2)–(4) is to be solved based on the deterministic initial condition Equation (4). The solution yields an optimal control sequence u * over ( t ^ z , t ^ z + H ] , and an optimal feedback is obtained by applying the first element of u * to the PDE system. To employ the receding horizon strategy, it is necessary to update the initial time instance t ^ z of the prediction horizon by Equation (7), where Δ t is a sampling time, and N t is the number of time discretization points in the operational period ( t ^ 1 , t ^ 1 + H ] of the finite length H .

3. Solution Approach to CCMPC

3.1. Finite Dimensional Representation

The CCMPC of parabolic PDE systems calls for the solution of a CCPDE optimization problem on each prediction horizon. The theoretical investigation of this problem can be worked out in a more general context of Banach spaces (see [20]).
Moreover, based on the discussion in Section 2, we assume that for each given admissible control u and realization of the random variable ξ , the PDE system of Equations (2)–(4) can be solved to obtain y ( u , ξ ; t , x ) . As a result, on each prediction horizon, the problem reduces to
( CCPDE r )
min u J ( u ) = E t ^ z t ^ z + H y ( u , ξ ; t , · ) y d ( t , · ) L 2 ( D ) 2 + λ 2 u ( t , · ) L 2 ( D ) 2 d t subject to :
p ( u ; t , x ) : = P r y ( u , ξ ; t , x ) y m a x α , ( t , x ) Q ,
u U .
For numerical computations, the reduced formulation CCPDE r Equations (8)–(10) need to be transformed into a finite dimensional representation on the prediction horizon through an appropriate space-time discretization technique. Suppose { t ^ z , t ^ z + Δ t , , t ^ z + ( N t 1 ) Δ t } is a set of appropriately chosen time instances, where the number of time discretization points N t = [ H Δ t ] + 1 , and { x 1 , x 2 , , x N x } are spatial discretization points generated according to a given finite element method from the domain D, respectively. Subsequently, these yield a vector of discretized controls u Equation (11), a corresponding admissible set U d Equation (12), a discretized cost functional J ( u ) Equation (13), and a set of chance constraints p j , k Equation (14) corresponding to each time instance and space point, as follows
u = [ u ( t 1 , x 1 ) , u ( t 2 , x 2 ) , , u ( t N t , x N x ) ] T R N t × N x ,
U d = { u R N t × N x | u ̲ u u ¯ } ,
J ( u ) = E j = 1 N t k = 1 N x y ( u ( t j , x k ) , ξ , t j , x k ) y d ( t j , x k ) 2 + λ 2 u ( t j , x k ) 2 ,
p j , k ( u ) = P r { y ( u , ξ , t j , x k ) y m a x ( t j , x k ) 0 } α , j = 1 , , N t ; k = 1 , , N x ,
where u ̲ = [ u m i n , , u m i n ] R N t × N x , u ¯ = [ u m a x , , u m a x ] R N t × N x . Consequently, a finite dimensional representation of CCPDE r takes the form
( C C O P T ) min u U d J ( u ) subject to :
p j , k ( u ) = P r { s j , k ( u , ξ ) 0 } α , j = 1 , , N t ; k = 1 , , N x ,
where s j , k ( u , ξ ) : = y ( u , ξ , t j , x k ) y m a x ( t j , x k ) and with the feasible set
P = u U d | p j , k ( u ) α , j = 1 , , N t ; k = 1 , , N x .
As mentioned above, the major computational hurdle associated with chance constrained optimization problems is the difficulty in evaluating the probability
p j , k ( u ) = P r s j , k ( u , ξ ) 0 = ξ Ω | s j , k ( u , ξ ) 0 ϕ ( ξ ) d ξ .
The probability function p j , k ( u ) usually does not have a closed form analytic representation. Except for special types of sets ξ Ω | s j , k ( u , ξ ) 0 , it is generally difficult to directly evaluate the integral on the right hand-side of Equation (18) to obtain the value of p j , k ( u ) for a given u. Furthermore, p j , k ( u ) can be non-differentiable, non-convex, and commonly guaranteeing the feasibility of an optimal solution is not trivial. Moreover, the theoretical assumptions for the differentiability of p j , k ( u ) are quite restrictive and also the knowledge of its differentiability furnishes little computational advantages. Therefore, to overcome these difficulties, we use our recently proposed smoothing inner–outer approximation strategy [47]. The viability of this approximation strategy has been demonstrated for both finite [47,48] and infinite dimensional [20] CCOPT problems. Here, we extend the inner–outer approach for the solution of chance constrained MPC problems for parabolic PDE systems. Using this method, we guarantee the feasibility of the state trajectories in MPC, despite the presence of uncertainties and, in particular, without requiring an extra effort for constructing a probability tube.

3.2. Inner-Outer Approximation Strategy

Here, we briefly summarize and adapt the inner–outer approximation approach of Geletu et al. [47] to a CCMPC problem, highlighting its relevant properties.

3.2.1. The Smoothing Approximation Problems

By dropping the indices from Equation (16) for the sake of brevity, consider a single chance constraint
p ( u ) = P r { s ( u , ξ ) 0 } α .
Define two parametric functions ψ ( τ , u ) and φ ( τ , u ) , with τ ( 0 , 1 ) , based on the Geletu–Hoffmann function Θ ( τ , s ) (Geletu et al. [47], also [20,48]) such that
ψ ( τ , u ) = E Θ ( τ , s ( u , ξ ) ) and φ ( τ , u ) = E Θ ( τ , s ( u , ξ ) ) ,
where
Θ ( τ , s ) = 1 + m 1 τ 1 + m 2 τ exp s τ ,
and m 1 , m 2 are constants with 0 < m 2 m 1 / ( 1 + m 1 ) .
Now, we define two parametric optimization problems in Table 1
with
M ( τ ) = { u U d | ψ ( u , τ ) 1 α } and S ( τ ) = { u U d | φ ( u , τ ) α }
are feasible sets of the problems (IA) τ and (OA) τ , respectively.

3.2.2. Properties of the Approximations

  • Properties of the Approximation Functions
    (a)
    Uniform limit. Property 2 of Theorem 3.6 in [47] implies that both functions 1 ψ ( τ , u ) and φ ( τ , u ) closely resemble the function p ( u ) , as the parameter τ moves closer to zero from above. This uniform approximation property has two important consequences.
    ( a 1 )
    p ( · ) is a continuous function (Corollary 3.8 in [20]).
    ( a 2 )
    The function ψ ( τ , · ) and φ ( τ , · ) are differentiable with regard to u (Theorem 3.6 in [47]).
    Note that a 1 and a 2 hold true irrespective of the differentiability of the probability function p ( · ) . Therefore, a gradient-based optimization algorithm can be used to solve the smooth approximation problems I A τ and O A τ , respectively, to obtain approximate optimal solutions to the (generally non-smooth) CCOPT problem.
    (b)
    If the same assumptions as in [62] hold true, then the probability function p ( · ) is differentiable and the gradients ( 1 ψ ( τ , u ) ) and φ ( τ , u ) can be shown to converge, with regard to τ , to the gradient p ( u ) (Theorem 16 in [47] and Theorem 4.10 in [48]).
  • Properties of the inner–outer approximation problems
    Recall that the feasible set of CCOPT is given by P = P r { s ( u , ξ ) 0 } α .
    (a)
    Inner–outer approximation. The feasible set M ( τ ) is always a subset (inner approximation) of the feasible set P of CCOPT, while the feasible set S ( τ ) is a superset (outer approximation) of the feasible set P of CCOPT; i.e.,
    M ( τ ) P S ( τ ) , for any τ ( 0 , 1 ) .
    (b)
    Monotonicity of the inner–outer approximations to ensure the enclosure of the feasible set:
    M ( τ 2 ) M ( τ 1 ) P S ( τ 1 ) S ( τ 2 ) , for 0 < τ 1 < τ 2 < 1 .
    (c)
    Convergence of the feasible sets of the approximations:
    lim τ 0 + M ( τ ) = P , lim τ 0 + S ( τ ) = P ,
    where the first convergence needs the regularity condition c l u U | p ( u ) > α = P (see Geletu et al. [47]).
    (d)
    Obviously, J I A ( · ) of ( I A ) τ is an upper bound to the objective function of ( O A ) τ and converge to each other as τ moves closer to 0 from above.
  • The algorithm for computation
    The following algorithm presents the computation scheme of the inner–outer approximation approach for solving a CCOPT problem.
In Algorithm 1, J I A ( u τ k * ) and J O A ( u ˜ τ k * ) represent the optimal objective function values of ( I A ) τ k and ( O A ) τ k at the respective optimal points u τ k * and u ˜ τ k * . The algorithm stops when the difference of optimal values of the inner and outer approximations is less than a given threshold.
Algorithm 1: Computation scheme for the inner–outer approximation method.
1: Choose an initial parameter τ 0 ( 0 ,   1 ) ;
2: Solve the optimization problems ( IA ) τ 0 and ( OA ) τ 0 ;
3: Select the termination tolerance tol;
4: Set k←0;
5: while ( | J I A ( u τ k * ) J O A ( u ˜ τ k * ) | > tol) do
6: Reduce the parameter τk (e.g., τk+1 = ρτk, for ρ ∈ (0, 1));
7:  Set kk + 1;
8: Solve the optimization problems ( IA ) τ k and ( OA ) τ k ;
9: end while
Remark 1.
Referring the chance constraint P r s ( u , ξ ; t , x ) 0 α , i n Q = ( t ^ , t ^ + H ] × D  of CCMPC with  s ( u , ξ ; t , x ) = y ( u , ξ ; t , x ) y m a x , define
ψ ( τ , u ; t , x ) = E Θ ( τ , s ( u , ξ ; t , x ) ) a n d φ ( τ , u ; t , x ) = E Θ ( τ , s ( u , ξ ; t , x ) ) .
Using the same parametric function  Θ ( τ , · )  as in Equation (19) and let  P ( t ) = { u U d | p ( u ; t , x ) α , x D } , M ( τ ; t ) = { u U d | ψ ( u , τ ; t , x ) 1 α , x D }  and  S ( τ ; t ) = { u U d | φ ( u , τ ; t , x ) α , x D } , t ( t ^ , t ^ + H ] . It follows that,
M ( τ ; t ) P ( t ) S ( τ ; t ) , t ( t ^ , t ^ + H ] .
This implies that the feasible sets of the inner and outer approximations automatically define deterministic inner and outer enveloping tubes to the feasible set of the chance constraint over each prediction horizon   ( t ^ , t ^ + H ]  (see Figure 1). These tubes guarantee the feasibility of the solution on each prediction horizon and will be made tighter by decreasing parameter τ by Δ τ , from one prediction horizon to the next, yielding robust constraint satisfaction. The tubes are constructed automatically, which neither incurs additional computation expenses nor does it require additional assumptions to simplify the problem.

4. Numerical Implementation

The CCMPC of the parabolic PDE system is reformulated as CCPDE with time discretization, where on each prediction horizon a CCPDE problem is solved by updating the initial condition Equation (4) at the time moment t ^ z following the receding horizon strategy Equation (7). For this purpose, we use the backward difference for the time discretization with N t time instances and finite element space discretization of the PDE system with N x space points, respectively. N quasi-Monte Carlo samples are generated for ξ . As a result, we obtain two finite dimensional NLP problems for the inner and outer approximations as follows:
( NLP I A ) τ min u 1 N i = 1 N j = 1 N t k = 1 N x y ( u ( t j , x k ) , ξ i , t j , x k ) y d 2 + λ 2 u ( t j , x k ) 2 subject to :
Ψ ( τ , u ( t j , x k ) ) 1 α , j = 1 , , N t ; k = 1 , , N x ,
u m i n u ( t j , x k ) u m a x , j = 1 , , N t ; k = 1 , , N x ,
where Ψ ( τ , u ( t j , x k ) ) = 1 N i = 1 N 1 + m 1 τ 1 + m 2 τ exp ( y m a x y ( u ( t j , x k ) , ξ i , t j , x k ) ) ,
( NLP O A ) τ min u 1 N i = 1 N j = 1 N t k = 1 N x y ( u ( t j , x k ) , ξ i , t j , x k ) y d 2 + λ 2 u ( t j , x k ) 2 subject to :
Φ ( τ , u ( t j , x k ) ) α , j = 1 , , N t ; k = 1 , , N x ,
u m i n u ( t j , x k ) u m a x , j = 1 , , N t ; k = 1 , , N x ,
where Φ ( τ , u ( t j , x k ) ) = 1 N i = 1 N 1 + m 1 τ 1 + m 2 τ exp ( y ( u ( t j , x k ) , ξ i , t j , x k ) y m a x ) .
Numerically, the spatial discretization of the PDEs is realized by the finite element method in the FEniCS library (FEniCS project [63,64]). Alternatively, libraries supporting spectral discretization could be used. Spectral discretization can improve the numerical accuracy of discretization, allowing larger intervals and leading to higher computational efficiency. However, in comparison to such libraries, FEniCS also allows the solution of the PDE-constrained optimization problems through the Dolfin-Adjoint package [65]. Despite the automated finite element spatial discretization provided by FEniCS, Dolfin-Adjoint does not directly support the solution of state-constrained optimization problems with PDEs. Therefore, to solve the problems NLP I A and NLP O A under FEniCS, we embed constraints (22) and (26) into the respective objective functions. In particular, when state variables are involved in inequality constraints explicitly or implicitly, the Moreau–Yosida regularization [66,67] can render the problem amenable for solving using FEniCS. Thus, we use the smoothed version of the Moreau–Yosida regularization and express the regularized problems as
( RNLP I A ) τ
min u 1 N i = 1 N j = 1 N t k = 1 N x y ( u ( t j , x k ) , ξ i , t j , x k ) y d 2 + λ 2 u ( t j , x k ) 2 + γ 2 R I A ( τ , u ( t j , x k ) ) 2 subject to :
u m i n u ( t j , x k ) u m a x , j = 1 , , N t ; k = 1 , , N x ,
( RNLP O A ) τ
min u 1 N i = 1 N j = 1 N t k = 1 N x y ( u ( t j , x k ) , ξ i , t j , x k ) y d 2 + λ 2 u ( t j , x k ) 2 + γ 2 R O A ( τ , u ( t j , x k ) ) 2 subject to :
u m i n u ( t j , x k ) u m a x , j = 1 , , N t ; k = 1 , , N x ,
where γ is a penalization parameter, and R I A and R O A are (discretized) Moreau–Yosida regularization terms expressed as
R I A ( τ , u ( t j , x k ) ) = Ψ ( τ , u ( t j , x k ) ) 1 + α + ( Ψ ( τ , u ( t j , x k ) ) 1 + α ) 2 + χ 2 ,
R O A ( τ , u ( t j , x k ) ) = α Φ ( τ , u ( t j , x k ) ) + ( α Φ ( τ , u ( t j , x k ) ) ) 2 + χ 2 .
In addition, the Moreau–Yosida regularization strategy inserts the L 1 penalty function max { 0 , f ( a ) } to the objective function for the constraints of the form f ( a ) 0 . However, to avoid working with this non-smooth expression, we use the equality max { 0 , f ( a ) } = ( f ( a ) + | f ( a ) | ) 2 and replace the absolute value by the smoothing approximation | f ( a ) | f ( a ) 2 + χ as in Equations (33) and (34), where χ R + is a sufficiently small smoothing parameter.
Figure 2 shows our implementation details. On each prediction horizon, the time and space discretization as well as random sampling are applied to obtain the finite dimensional representation of smoothing inner and outer approximations of the PDE-constrained optimization problem, (RNLP I A ) τ and (RNLP O A ) τ , respectively. Note that the samples are generated only once and then used to reformulate the initial problem on each prediction horizon. Subsequently, the Moreau–Yosida regularization is employed to regularize (RNLP I A ) τ and (RNLP O A ) τ . The regularized problems are then solved using Algorithm 1 for decreasing values of the approximation parameter τ coupled with a sampling strategy for the random variable ξ to evaluate the expected values in Equation (20) and their gradients ( 1 ψ ( τ , u ) ) and φ ( τ , u ) .
We implement RNLP I A and RNLP O A directly in the finite element code and solve them using Dolfin-Adjoint, which provides automatic differentiation and derives adjoint and tangent linear models from the FEniCS-based forward model for the gradient calculation. At each iteration of the optimization algorithm, y ( u , ξ i , x , t j ) of the PDE system Equations (2)–(4) will be obtained for each time instant of the random variable ξ i . N-times of solutions of the PDE are necessary, since y ( u , ξ i , x , t j ) is required to calculate the values and gradients of the cost functional in RNLP I A and RNLP O A .
Furthermore, problems RNLP I A and RNLP O A are solved in a double loop: in the first loop over the initial time instance t ^ z , updated by Equation (7), and in the second loop over decreasing values of τ . To speed up the computation, we use parallel computing for the finite element code by spreading the solution calculations of the PDE system for each realization of κ ( ξ i ) and by allowing function evaluation and gradient computation over several processor threads through the MPI standard. It is also worth noting that MPI is the only method that allows running such Dolfin-Adjoint finite element code in parallel. Finally, on each prediction horizon, the resulting NLP problems are solved in parallel by the primal-dual interior-point method (the IPOPT library [68]) with the parallel linear solver HSL_MA97 to obtain optimal control sequences. The solution is used as an initial guess for the optimization on the next prediction horizon, which means a so-called warm MPC strategy.

5. Case Study

To demonstrate the proposed approach, we consider here a hyperthermia cancer treatment problem as a case study. Hyperthermia treatment (HT) consists of the heating of malignant tumors to subdue or eradicate the growth of tumor cells from a given organ [56,57]. Depending on the tissue area to be covered, the therapy is divided into three categories: local, regional, and whole-body HT. While the latter two methods act on the whole body or limbs, a local hyperthermia affects only cancerous tissues [69], for instance, a small skin area in case of skin cancer. This allows us to accurately model the tumor domain, formulate an optimization problem, and find the optimal strategy for thermal treatment, which is considered in this case study.

5.1. Mathematical Model

The Pennes bioheat equation [70]
ρ · c t · T t x κ x T + ω · c b · T T a = Q ( u ) , in ( t , t + H ] × D
is a parabolic PDE frequently used to compute the temperature distribution T in the therapy tissue domain D in response to the heating power distribution Q, which, in terms of local HT, can be waves of different frequency ranges and magnetic fields. Equation (35) has the following coefficients and notations: ρ —tissue density, c t —specific heat capacity of tissue, T a —arterial blood temperature, c b —specific heat capacity of blood, κ —tissue thermal conductivity, and ω —perfusion of blood.

5.2. Source of Uncertainty

Equation (35) is widely used in modeling of hyperthermia treatment and often considered to be deterministic. However, its parameters (35) are found to differ from patient to patient as human tissue characteristics are not identical among all individuals [71,72,73]. In addition, laboratory experiments and measurements for a single patient introduce observational error, for instance, owing to the complexity of the measured tissue or organ. Therefore, the model parameters ρ ; c t ; κ ; ω and c b are known to be uncertain [73,74]. As a result, these uncertainties propagate through the model to the predicted values of temperature distribution T in the tumor domain and to the amount of heat transferred to healthy neighboring tissues, making them also uncertain [75,76,77] and sensitive to the parameter variations [78].
In this study, we assume the thermal conductivity κ to be a time- and space-independent random parameter and its distribution can be obtained from clinical data. This consideration is justified by the fact that in local hyperthermia a relatively small area of tissue is subjected to therapy. As a consequence, the tumor may have properties near to be homogeneous across the domain; i.e., a variation of κ in space and time may be insignificant.

5.3. Optimization Problem

Our aim is to regulate the thermal dose through a chance constrained MPC scheme and to keep the tumor tissue D t uniformly heated to a desired temperature T d during the treatment (see Figure 3).
At the same time, the heat injection should satisfy the required temperature restriction with a high reliability by avoiding over-heating of the tumor tissue as well as the healthy neighboring tissues D \ D t under all boundary conditions. In this regard, we formulate the following chance constrained HT MPC problem:
( CCHTMPC )
min u E t ^ z t ^ z + H T ( u , ξ ; t , · ) T d ( t , · ) L 2 ( D ) 2 + λ 2 u ( t , · ) L 2 ( D ) 2 d t subject to :
ρ · c t · T t x κ ( ξ ) x T + ω · c b · T T a = W ( u ) , in Q × Ω ,
κ 0 x T · n = h T T a on ( t ^ z , t ^ z + H ] × D × Ω ,
T ( t ^ z , x ) = T t ^ z ( x ) in D ,
P r { T ( u , ξ ; t , x ) T m a x } α in Q ,
u m i n u ( t , x ) u m a x in Q ,
t ^ z + 1 : = t ^ z + Δ t , z = 1 , , N t ,
where the control vector u is bounded by u m i n and u m a x and denotes the heat supply to the tumor region D t , the intensity of u is specified by a regularization parameter λ > 0. The heat leaving the controlled domain D through the boundary D due to interaction with the environment is determined by the Robin boundary condition Equation (38). h is the heat transfer coefficient, and T t ^ z is the initial temperature on D that sets the initial condition Equation (39). At the initial moment t ^ 1 = 0 initial temperature T ( 0 , x ) = T s with T s as the skin temperature. T m a x is the maximum allowed temperature distributed on D to avoid overheating of the body, which will be ensured by the upper chance constraint (40).
In previous deterministic studies on HT planning problems, several authors used state constraint T T m i n on D (e.g., [76,79,80]), where T m i n is the minimum allowed therapy temperature. In a stochastic case, it would take the form of lower chance constraint P r { T T m i n } α in Q . As it will be shown in the next section, this constraint is redundant, since even without it, the temperature holds sufficiently near to T d .

5.4. Parameters and Coefficients

In Equation (37), we set ρ = c t = ω = c b = 1 for simplicity. The heat transfer coefficient is set to h = 10 5 . κ 0 = 10 2 is used as the nominal value of the thermal conductivity to generate its random values by κ ( ξ i ) = κ 0 ξ i . The random parameter ξ is described by the four-parameter beta distribution, the probability density function of which β s ( x ; s , l , a , b ) has the following relation with the standard two-parameter beta distribution β ( x ; a , b )
β s ( x ; s , l , a , b ) = 1 s β ( x l s , a , b ) ,
where s and l stand for the scale and shift parameters, respectively.
Using the quasi-Monte-Carlo method and β s ( x ; 0.2 , 0.9 , 2.0 , 2.0 ) , a sample array of N = 300 samples is generated. Furthermore, to obtain early prediction of constraint violations, to produce corresponding corrective actions, and to ensure numerical stability of the controller, a finite prediction horizon of the length H = 2 (scaled time) with sampling time Δ t = 0.025 is set. This corresponds to N t = 81 time instances on each prediction horizon. Usually, in MPC implementation, the computation time for the update should be smaller than the sampling time. However, our numerical study is focused on the investigation of the applicability of the proposed method. In the domain D = [ 0 , 1 ] × [ 0 , 1 ] , we specify N x = 21 × 21 = 441 points and 800 Lagrange elements of the first degree, 40 % of which cover the tumor sub-domain D t = [ 0.3 , 0.7 ] × [ 0.3 , 0.7 ] . As the control is applied only to the sub-domain of the CCHTMPC problem, the final representation has 9 × 9 × 80 = 6480 decision variables on each prediction horizon that need to be obtained while the prediction horizon is moving along the operational period (see Figure 4).
The tumor temperature during the HT should be maintained in the range of [ 42 , 44 ] C. For the numerical calculation, the temperature values in degrees Celsius are scaled to dimensionless quantities. For instance, the maximum allowed temperature of 44 C is scaled to T m a x = 2.05 , the initial temperature, assuming it to be the skin temperature of 32 C, is set to T s = 0 . The desired temperature of 43.7 C is linearly interpolated between T m a x and T s and corresponds to T d = 2.0 . Instead of the arterial blood, the capillary blood is considered and assumed to have the temperature of 34.1 C, which is scaled to T a = 0.36 . Other parameters are listed in Table 2.

6. Results and Discussions

This section presents the results of the numerical experiments. All computations are carried out on a 64-bit Ubuntu Linux machine of version 20.04.2 LTS, equipped with 16 Threads 3.60 GHz Intel Xeon Gold 6244 CPU and 60 GB RAM.

6.1. Results of Optimal Control

For an illustration of the properties of the proposed approach, we first demonstrate the solutions of the inner and outer NLP problems of CCHTMPC only on the first prediction horizon ( 0 , H ] , which corresponds to the solution of an optimal control problem—CCPDE (i.e., open-loop control without feedback). The optimal control sequences u ( 0 , H ] * ( t , x ) obtained by a decreasing set of the approximation parameter τ are then applied to the PDE system Equations (37)–(39), leading to optimal temperature distributions T * ( u ( 0 , H ] * ( t , x ) , ξ i , t , x ) , i = 1 , , N ; i.e., for the N values of κ ( ξ i ) , from which the mean value E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] is calculated and presented over time in Figure 5, Figure 6 and Figure 7.
The temperature distributions presented in Figure 5 show that the outer problem with a high value of τ = 0.1 provides an infeasible solution that exceeds the maximum allowed temperature T m a x = 2.05 , while the solution of the inner problem remains feasible, but far from the desired temperature T d = 2.0 . With the decrease of τ , as shown in Figure 5, Figure 6 and Figure 7 and Figure A1 and Figure A2, the solutions of the inner and the outer problem converge to the desired temperature, and with τ = 5 × 10 5 the solution will not violate the chance constraint any more (Figure 7).
The convergence of the inner and outer solutions can be seen more clearly in Figure 8, where the maximum values of mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] are presented along the time and for each value of τ . This Figure 8, similar to Figure 1, shows how the solutions of the outer problem (red curves) narrow to the border of the feasible set of the original problem at T m a x = 2.05 due to the constraining effect of the outer deterministic tubes, whereas the solutions of the inner problem (blue curves) expand to the feasible set of the original problem due to the relaxing effect of the inner deterministic tube, as the parameter τ decreases. It can also be seen that a strong initial heat impulse makes the state constraint prone to violation in the early moments of control t = [ 0.0 , 0.25 ] , which could be worsened if a lower reliability level is imposed on the chance constraints. This can be explained by the fact that the minimum allowed temperature T m i n = 42 C is much closer to the maximum allowed temperature T m a x = 44 C than to the initial temperature T s = 32 C, which would force the controller to give a greater thermal momentum to the tumor body at the beginning in order to approach the feasible region T T m i n as fast as possible.
At the solution, the objective function values obtained under the optimal control for decreasing τ values are presented in Figure 9. It can be seen, for τ = 0.01 , the objective function values move close to each other, and for smaller τ their difference is quite small. This indicates the convergence of the inner and outer solutions. However, 5 × 10 5 is the smallest possible value of τ in the current settings, since there will be a problem of vanishing gradients attributed to the form of the approximation function in Equation (19) when τ is smaller than this value. This is due to the fact that it moves closer to the Heaviside function for small values of τ [48]. It should also be noted that the Moreau–Yosida regularization terms Equations (33) and (34) are excluded from the calculation of the optimal objective function values presented in Figure 9, since they contain the parameter τ and affect the objective function values under the same control but with different τ values.
Now, we evaluate the realized reliability level α r e a l of the solutions. This can be calculated as a percentage of the resulting temperature realizations that satisfy the temperature constraint under the optimal control sequence u ( 0 , H ] * and given samples of ξ :
α r e a l = 1 N i = 1 N I ( u * ( t , x ) , ξ i ) ,
I ( u * ( t , x ) , ξ i ) = 1 , T * ( u * ( t , x ) , ξ i , t , x ) T m a x 0 , T * ( u * ( t , x ) , ξ i , t , x ) > T m a x .
The result shows that, with τ = 5 × 10 5 , we have α r e a l O A = α r e a l I A = 0.973 . It means that, despite the fact that the maximum allowed temperature is very close to the desired temperature, which even in the deterministic case significantly makes it difficult to find admissible values of control, the chance constraint resulted from the optimal control holds.
The computation time for solving NLP problems on the first prediction horizon, for τ = 5 × 10 5 , is on average 12.75 h. The decrease of τ and the corresponding tightening of the feasible set is accompanied by a certain computational penalty, which is presented in Table 3. Here, we show only the time taken by the IPOPT optimization and discard the time needed to generate and compile the finite element code. The increase in computation time is possibly also caused by the vanishing gradient problem described above.
From Table 3 and Figure 8, we can see the effect of τ values in this case. It is seen from Figure 8 that only for τ 10 3 the solutions of inner and outer problems lie in the desired range, i.e., they reach T d but do not violate T m a x . Table 3 shows that by τ 10 3 , the computation needed for solving one problem takes approximately the same time. Thus, the computation time does not heavily depend on the τ value.

6.2. Results of MPC

Now, we present the solutions of the MPC scheme; i.e., u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) of the inner and outer problems of CCHTMPC, obtained by collecting the first elements of optimal control sequences for all prediction horizons u ( t ^ z , t ^ z + H ] * ( t , x ) , t ^ z ( t ^ 1 , t ^ 1 + H ] . The number of samples used for the random parameter is N = 300 . The approximation parameter is set to the smallest possible value τ = 5 × 10 5 . We decrease it from one prediction horizon to the next one by Δ τ = 10 9 .
The resulting mean profile of the optimal temperature realizations E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] over time is shown in Figure A3. It is similar to the temperature profile of the optimal control shown in Figure 7. The difference between the inner and outer solutions is hardly visible; i.e., the mean temperature realizations in the tumor domain converge by the MPC scheme.
To investigate the feasibility of the resulting solutions, the maximum values of the mean temperature distributions on all prediction horizons E [ T * ( u ( t ^ z , t ^ z + H ] * ( t , x ) , ξ , t , x ) ] , t ^ z ( t ^ 1 , t ^ 1 + H ] are presented by the gray curves in Figure 10, and the maximum values of the mean temperature distributions of CCHTMPC E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] (red and blue curves), respectively. It can be clearly seen that the CCHTMPC solution (of the MPC) differs from the CCPDE (of the optimal control). By using MPC, the initial impulse tending to violate the state constraint is not as strong as before, i.e., the temperature does not reach T m a x . At the same time, the solutions after that impulse become much smoother and do not fluctuate anymore.
To evaluate the satisfaction of the chance constraints by the MPC, we plot the temperature realizations of the inner and outer approximate solutions of the CCHTMPC problem for all N = 300 samples over time; i.e., T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ i , t , x ) , i = 1 , , N with gray curves in Figure 11a for the inner problems and in Figure 11b for the outer problem as well as their mean temperature distributions E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] .
As can be seen, only a small percentage of the trajectories violates the state constraint; i.e., the computed control sequence is feasible for most samples of the random variable. It corresponds to the realized reliability levels α r e a l O A , M P C = α r e a l I A , M P C = 0.987 . Both realized reliability levels are higher than the specified value α = 0.95 and are also higher than those of the open-loop optimal control α r e a l O A = α r e a l I A = 0.973 . In general, the reliability level should be specified according to the constraints and severity of randomness. The stronger they are, the harder it is to find a feasible solution. In accordance with this, the chosen reliability level must relax the problem enough to have a solution, while providing a sufficient level of performance.
The computation time of CCHTMPC NLP problems using a solution of open-loop optimal control as an initial guess, i.e., on the single prediction horizon for τ = 5 × 10 5 is on average 2.7 h, which is much faster than the computation time of open-loop control. All this demonstrates the significant advantages of MPC due to its feedback feature. However, the computation time still has to be decreased for a real application, as a long computation time results in a long MPC sampling time, which makes the MPC unrealistic.

7. Conclusions and Future Work

The CCMPC of PDE systems poses an enormous computational challenge. In this study, a smoothing inner–outer approximation approach is proposed for the solution of CCMPC of parabolic PDE systems. Through a series of transformations, the original problem is converted to a couple of high-dimensional deterministic state-constrained NLP problems, the solution of which is an approximation to that of the initial problem. Each of these NLPs is solved in a double loop for updates of the initial condition to employ the receding horizon strategy and for decreasing the approximation parameter τ to guarantee convergence of the approximation to the original problem. The Moreau–Yosida regularization is applied to address the state constraints of the resulting NLPs. The proposed MPC scheme gains the following advantages: (1) feasibility is guaranteed, irrespective of the distribution of input uncertainties (bounded or unbounded uncertainties); (2) through the construction of deterministic tubes, the feasibility is ensured on each prediction horizon, yielding robust constraint satisfaction; and (3) the deterministic tubes are constructed automatically, simplifying derivation of a feasible solution. To verify the properties of the proposed approach, a case study of hyperthermia cancer treatment is performed. Numerical results demonstrate the viability of the proposed method, validate the stated properties, and also indicate some directions for further studies.
The approach turned out to be quite computationally demanding, which points to the need for an accurate temporal discretization and a prediction horizon reduction. The latter can be carried out by analyzing the stability properties and deriving a minimal stabilizing prediction horizon for the MPC. These and other questions, such as reducing the computation time, application-based specification of parameters associated with chance constraints, and discretization improvement, e.g., estimation of errors, and use of spectral techniques will be the focuses of our future research.

Author Contributions

Conceptualization and methodology, R.V. and A.G.; software and validation, R.V.; formal analysis, A.G.; supervision, A.G. and P.L.; writing—original draft preparation, R.V.; writing—review and editing, A.G. and P.L.; project administration, P.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Deutsche Forschungsgemeinschaft (DFG) grant number LI 806/13-4. We acknowledge support for the publication costs by the Open Access Publication Fund of the Technische Universität Ilmenau.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

Abbreviations

The following abbreviations are used in this manuscript:
MPCmodel predictive contro
PDEpartial differential equation
UQuncertainty quantification
CCMPCchance constrained model predictive control
CCPDEchance constrained optimal control
CCOPTchance constrained optimization
IAinner-approximation
OAouter-approximation
NLPnonlinear programming
RNLPregularized nonlinear programming
IPOPTInterior Point OPTimizer
HThyperthermia treatment
CCHTMPCchance constrained hyperthermia treatment model predictive control

Appendix A

Figure A1. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] of optimal control on the first prediction horizon by τ = 0.001 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Figure A1. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] of optimal control on the first prediction horizon by τ = 0.001 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Mathematics 11 01372 g0a1
Figure A2. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 0.0001 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Figure A2. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 0.0001 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Mathematics 11 01372 g0a2
Figure A3. Mean temperature realizations E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] of CCHTMPC by τ = 5 × 10 5 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 0.5 ; (d) O A , t = 1.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 0.5 ; (h) I A , t = 1.0 .
Figure A3. Mean temperature realizations E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] of CCHTMPC by τ = 5 × 10 5 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 0.5 ; (d) O A , t = 1.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 0.5 ; (h) I A , t = 1.0 .
Mathematics 11 01372 g0a3

References

  1. Farlow, S. Partial Differential Equations for Scientists and Engineers; Dover Publications: New York, NY, USA, 1993; pp. 1–32. [Google Scholar]
  2. Ockendon, J.; Howison, S.; Lacey, A.; Movchan, A. Applied Partial Differential Equations; Oxford University Press: Oxford, UK, 2003; pp. 1–5. [Google Scholar]
  3. Strauss, W. Partial Differential Equations: An Introduction; Wiley: Danvers, MA, USA, 2007; pp. 6–54. [Google Scholar]
  4. Babuska, I.; Nobile, F.; Tempone, R. A Stochastic Collocation Method for Elliptic Partial Differential Equations with Random Input Data. SIAM J. Numer. Anal. 2007, 45, 1005–1034. [Google Scholar] [CrossRef]
  5. Barth, A.; Lang, A.; Schwab, C. Multilevel Monte Carlo method for parabolic stochastic partial differential equations. BIT Numer. Math. 2013, 53, 3–27. [Google Scholar] [CrossRef] [Green Version]
  6. Deb, M.; Babuška, I.; Oden, J. Solution of stochastic partial differential equations using Galerkin finite element techniques. Comput. Methods Appl. Mech. Eng. 2001, 190, 6359–6372. [Google Scholar] [CrossRef]
  7. Frauenfelder, P.; Schwab, C.; Todor, R.A. Finite elements for elliptic problems with stochastic coefficients. Comput. Methods Appl. Mech. Eng. 2005, 194, 205–228. [Google Scholar] [CrossRef]
  8. Ghanem, R.; Spanos, P. Stochastic Finite Elements: A Spectral Approach; Civil, Mechanical and Other Engineering Series; Dover Publications: New York, NY, USA, 2003; pp. 15–65. [Google Scholar]
  9. Gittelson, C.J. Stochastic Galerkin discretization of the log-normal isotropic diffusion problem. Math. Model. Methods Appl. Sci. 2010, 20, 237–263. [Google Scholar] [CrossRef]
  10. Kornhuber, R.; Schwab, C.; Wolf, M.W. Multilevel Monte Carlo Finite Element Methods for Stochastic Elliptic Variational Inequalities. SIAM J. Numer. Anal. 2014, 52, 1243–1268. [Google Scholar] [CrossRef] [Green Version]
  11. Kunoth, A.; Schwab, C. Sparse Adaptive Tensor Galerkin Approximations of Stochastic PDE-Constrained Control Problems. SIAM/ASA J. Uncertain. Quantif. 2016, 4, 1034–1059. [Google Scholar] [CrossRef]
  12. Kuo, F.; Scheichl, R.; Schwab, C.; Sloan, I.; Ullmann, E. Multilevel Quasi-Monte Carlo Methods for Lognormal Diffusion Problems. Math. Comput. 2015, 86. [Google Scholar] [CrossRef]
  13. Kuo, F.Y.; Schwab, C.; Sloan, I.H. Quasi-Monte Carlo Finite Element Methods for a Class of Elliptic Partial Differential Equations with Random Coefficients. SIAM J. Numer. Anal. 2012, 50, 3351–3374. [Google Scholar] [CrossRef]
  14. Matthies, H.; Keese, A. Galerkin methods for linear and nonlinear elliptic stochastic partial differential equations. Comput. Methods Appl. Mech. Eng. 2005, 194, 1295–1331. [Google Scholar] [CrossRef] [Green Version]
  15. Mugler, A.; Starkloff, H.J. On Elliptic Partial Differential Equations with Random Coefficients. Stud. Univ.-Babeș-Bolyain Math. 2011, 56, 473–487. [Google Scholar]
  16. Nobile, F.; Tempone, R. Analysis and implementation issues for the numerical approximation of parabolic equations with random coefficients. Int. J. Numer. Methods Eng. 2009, 80, 979–1006. [Google Scholar] [CrossRef]
  17. Wan, X.; Karniadakis, G.E. Solving elliptic problems with non-Gaussian spatially-dependent random coefficients. Comput. Methods Appl. Mech. Eng. 2009, 198, 1985–1995. [Google Scholar] [CrossRef]
  18. Borzi, A.; Schulz, V.; Schillings, C.; Winckel, G. On the treatment of distributed uncertainties in PDE-constrained optimization. GAMM-Mitteilungen 2010, 33, 230–246. [Google Scholar] [CrossRef]
  19. Farshbaf-Shaker, M.; Henrion, R.; Hoemberg, D. Chance Constraints in PDE Constrained Optimization; Preprint; Weierstraß-Institut für Angewandte Analysis und Stochastik: Berlin, Germany, 2017. [Google Scholar] [CrossRef]
  20. Geletu, A.; Hoffmann, A.; Schmidt, P.; Li, P. Chance constrained optimization of elliptic PDE systems with a smoothing convex approximation. ESAIM Control Optim. Calc. Var. 2020, 26, 1–28. [Google Scholar] [CrossRef] [Green Version]
  21. Göttlich, S.; Kolb, O.; Lux, K. Chance-constrained optimal inflow control in hyperbolic supply systems with uncertain demand. Optim. Control Appl. Methods 2021, 42, 566–589. [Google Scholar] [CrossRef]
  22. Hou, L.; Lee, J.; Manouzi, H. Finite element approximations of stochastic optimal control problems constrained by stochastic elliptic PDEs. J. Math. Anal. Appl. 2011, 384, 87–103. [Google Scholar] [CrossRef] [Green Version]
  23. Kouri, D.; Surowiec, T. Risk-Averse PDE-Constrained Optimization Using the Conditional Value-At-Risk. SIAM J. Optim. 2016, 26, 365–396. [Google Scholar] [CrossRef]
  24. Kouri, D.P. A Multilevel Stochastic Collocation Algorithm for Optimization of PDEs with Uncertain Coefficients. SIAM/ASA J. Uncertain. Quantif. 2014, 2, 55–81. [Google Scholar] [CrossRef]
  25. Kouri, D.P.; Heinkenschloss, M.; Ridzal, D.; van Bloemen Waanders, B.G. Inexact Objective Function Evaluations in a Trust-Region Algorithm for PDE-Constrained Optimization under Uncertainty. SIAM J. Sci. Comput. 2014, 36, A3011–A3029. [Google Scholar] [CrossRef]
  26. Lee, H.C.; Lee, J. A Stochastic Galerkin Method for Stochastic Control Problems. Commun. Comput. Phys. 2013, 14. [Google Scholar] [CrossRef]
  27. Martinez-Frutos, J.; Kessler, M.; Munch, A.; Periago, F. Robust optimal Robin boundary control for the transient heat equation with random input data. Int. J. Numer. Methods Eng. 2016, 108, 116–135. [Google Scholar] [CrossRef]
  28. Martinez-Frutos, J.; Kessler, M.; Periago, F. Robust Optimal shape design for an elliptic PDE with uncertainty in its input data. ESAIM Control Optim. Calc. Var. 2015, 21, 901–923. [Google Scholar] [CrossRef]
  29. Mordukhovich, B. Optimal control and feedback design of state-constrained parabolic systems in uncertainty conditions. Appl. Anal. 2011, 90, 1075–1109. [Google Scholar] [CrossRef]
  30. Schmidt, P.; Geletu, A.; Li, P. Stochastische Optimierung parabolischer PDE-Systeme unter Wahrscheinlichkeitsrestriktionen am Beispiel der Temperaturregelung eines Stabes. Automatisierungstechnik 2018, 66, 975–985. [Google Scholar] [CrossRef]
  31. Ridzal, D.; Waanders, B.; Kouri, D.; Heinkenschloss, M. A Trust-Region Algorithm with Adaptive Stochastic Collocation for PDE Optimization under Uncertainty. SIAM J. Sci. Comput. 2012, 35. [Google Scholar] [CrossRef] [Green Version]
  32. Tiesler, H.; Kirby, R.M.; Xiu, D.; Preußer, T. Stochastic Collocation for Optimal Control Problems with Stochastic PDE Constraints. SIAM J. Control Optim. 2012, 50, 2659–2682. [Google Scholar] [CrossRef] [Green Version]
  33. van Ackooij, W.; Pérez-Aros, P. Generalized Differentiation of Probability Functions Acting on an Infinite System of Constraints. SIAM J. Optim. 2019, 29, 2179–2210. [Google Scholar] [CrossRef]
  34. Alla, A.; Volkwein, S. Asymptotic stability of POD based model preditive control for a semilinear parabolic PDE. Adv. Comput. Math. 2015, 41, 1073–1102. [Google Scholar] [CrossRef] [Green Version]
  35. Altmüller, N.; Grüne, L. Distributed and boundary model predictive control for the heat equation. GAMM-Mitteilungen 2012, 32, 131–145. [Google Scholar] [CrossRef]
  36. Mechelli, L.; Volkwein, S. POD-based economic model predictive control for heat-convection phenomena. In Numerical Mathematics and Advanced Applications ENUMATH 2017; Adrian, F., KumarInga, R., Martin, J., Pop, I., Eds.; Springer International Publishing: Cham, Switzerland, 2019; pp. 663–671. [Google Scholar] [CrossRef] [Green Version]
  37. Dubljevic, S.; Mhaskar, P.; El-Farra, N.H.; Christofides, P. Predictive Control of Transport-Reaction Processes. Comput. Chem. Eng. 2005, 29, 2335–2345. [Google Scholar] [CrossRef]
  38. Dubljevic, S.; Christofides, P. Predictive output feedback control of parabolic partial differential equations (PDEs). Ind. Eng. Chem. Res. 2006, 45, 8421–8429. [Google Scholar] [CrossRef]
  39. Dubljevic, S.; El-Fara, N.; Mhaskar, P.; Christofides, P. Predictive control of parabolic PDEs with state and control constraints. Int. J. Robust Nonlinear Control 2006, 16, 749–772. [Google Scholar] [CrossRef]
  40. Lao, L.; Ellis, M.; Christofides, P. Economic model predictive control of parabolic PDE systems: Addressing state estimation and computational efficiency. J. Process. Control 2014, 24, 448–462. [Google Scholar] [CrossRef]
  41. Lao, L.; Ellis, M.; Christofides, P. Economic Model Predictive Control of Transport-Reaction Processes. Ind. Eng. Chem. Res. 2014, 53, 7382–7396. [Google Scholar] [CrossRef]
  42. Shang, H.; Forbes, J.; Guay, M. Computationally efficient model predictive control for convection dominated parabolic systems. J. Process. Control 2007, 17, 379–386. [Google Scholar] [CrossRef]
  43. Li, P.; Arellano-Garcia, H.; Wozny, G. Chance Constrained Programming Approach to Process Optimization Under Uncertainty. Comput. Chem. Eng. 2008, 32, 25–45. [Google Scholar] [CrossRef]
  44. Wendt, M.; Li, P.; Wozny, G. Nonlinear Chance-Constrained Process Optimization under Uncertainty. Ind. Eng. Chem. Res. 2002, 41, 3621–3629. [Google Scholar] [CrossRef]
  45. Flemming, T.; Bartl, M.; Li, P. Set-Point Optimization for Closed-Loop Control Systems under Uncertainty. Ind. Eng. Chem. Res. 2007, 46, 4930–4942. [Google Scholar] [CrossRef]
  46. Klöppel, M.; Geletu, A.; Hoffmann, A.; Li, P. Using Sparse-Grid Methods to Improve Computation Efficiency in Solving Dynamic Nonlinear Chance-Constrained Optimization Problems. Ind. Eng. Chem. Res. 2011, 50, 5693–5704. [Google Scholar] [CrossRef]
  47. Geletu, A.; Hoffmann, A.; Klöppel, M.; Li, P. An Inner-Outer Approximation Approach to Chance Constrained Optimization. SIAM J. Optim. 2017, 27, 1834–1857. [Google Scholar] [CrossRef]
  48. Geletu, A.; Klöppel-Gersdorf, M.; Hoffmann, A.; Li, P. A tractable approximation of non-convex chance constrained optimization with non-Gaussian uncertainties. Eng. Optim. 2015, 47, 495–520. [Google Scholar] [CrossRef]
  49. Cannon, M.; Buerger, J.; Kouvaritakis, B.; Raković, S. Robust tubes in nonlinear model predctive control. IEEE Trans. Autom. Control 2011, 56, 1942–1947. [Google Scholar] [CrossRef]
  50. Cannon, M.; Kouvaritakis, B.; Raković, S.; Cheng, Q. Stochastic tubes in model predictive control with probabilistic constraint. IEEE Trans. Autom. Control 2011, 56, 194–200. [Google Scholar] [CrossRef]
  51. Langson, W.; Chryssochoos, I.; Rakovic, S.; Mayne, D. Robust model predictive control using tubes. Automatica 2004, 40, 125–133. [Google Scholar] [CrossRef]
  52. Mayne, D.; Kerrigan, E.; van Wyk, E.; Falugi, P. Tube-based robust nonlinear model predictive control. Robust Nonlinear Control 2011, 21, 1341–1353. [Google Scholar] [CrossRef]
  53. Raković, S.; Kouvaritakis, B.; Cannon, M.; Panos, C.; Findeisen, R. Parameterized tube model predictive control. IEEE Trans. Autom. Control 2010, 57, 2746–2758. [Google Scholar] [CrossRef]
  54. Raković, S.; Kouvaritakis, B.; Findeisen, R.; Cannon, M. Homothetic Tube Model Predictive Control. Automatica 2012, 48, 1631–1638. [Google Scholar] [CrossRef]
  55. Boger, E.; Wigström, O. A Partial Differential Equation Approach to Inhalation Physiologically Based Pharmacokinetic Modeling. CPT Pharmacometrics Syst. Pharmacol. 2018, 7. [Google Scholar] [CrossRef] [Green Version]
  56. Deuflhard, P.; Seebass, M.; Stalling, D.; Beck, R.; Hege, H.C. Hyperthermia Treatment Planning in Clinical Cancer Therapy: Modelling, Simulation, and Visualization. Sci. Comput. Model. Appl. Math. 1997, 3, 9–17. [Google Scholar]
  57. Deuflhard, P.; Schiela, A.; Weiser, M. Mathematical Cancer Therapy Planning in Deep Regional Hyperthermia. Acta Numer. 2012, 21, 307–378. [Google Scholar] [CrossRef]
  58. Quintero, M.; Cordovez, J. Looking for an Efficient and Safe Hyperthermia Therapy: Insights from a Partial Differential Equations Based Model. Differ. Equ. Dyn. Syst. 2015, 25, 137–150. [Google Scholar] [CrossRef]
  59. Weickert, J.; Schnörr, C. PDE–Based Preprocessing of Medical Images Optimization. University of Mannheim. Working Paper. 2000. Available online: http://ub-madoc.bib.uni-mannheim.de/1840 (accessed on 12 February 2023).
  60. Mang, A.; Gholami, A.; Davatzikos, C.; Biros, G. PDE-constrained optimization in medical image analysis. Optim. Eng. 2018, 19, 765–812. [Google Scholar] [CrossRef] [Green Version]
  61. Evans, L. Partial Differential Equations, 2nd ed.; American Mathematical Society: Providence, RI, USA, 2010; pp. 311–320. [Google Scholar]
  62. Uryasev, S. Derivatives of probability functions and some applications. Ann. Oper. Res. 1995, 56, 287–311. [Google Scholar] [CrossRef]
  63. Logg, A.; Wells, G.; Mardal, K.A. Automated Solution of Differential Equations by the Finite Element Method. The FEniCS Book; Springer: Berlin/Heidelberg, Germany, 2012; Volume 84, pp. 175–224. [Google Scholar] [CrossRef]
  64. Alnaes, M.; Blechta, J.; Hake, J.; Johansson, A.; Kehlet, B.; Logg, A.; Richardson, C.; Ring, J.; Rognes, M.; Wells, G. The FEniCS project version 1.5. Arch. Numer. Softw. 2015, 3, 9–23. [Google Scholar] [CrossRef]
  65. Mitusch, S.; Funke, S.; Dokken, J.S. dolfin-adjoint 2018.1: Automated adjoints for FEniCS and Firedrake. J. Open Source Softw. 2019, 4, 1292. [Google Scholar] [CrossRef]
  66. Hintermüller, M.; Hinze, M. Moreau-Yosida Regularization in State Constrained Elliptic Control Problems: Error Estimates and Parameter Adjustment. SIAM J. Numer. Anal. 2009, 47, 1666–1683. [Google Scholar] [CrossRef] [Green Version]
  67. Krumbiegel, K.; Ira, N.; Arnd, R. Sufficient Optimality Conditions for the Moreau–Yosida-Type Regularization Concept Applied to the Semilinear Elliptic Optmimal Control Problems with Pointwise State Constraints; Weierstraß-Institut für Angewandte Analysis und Stochastik: Berlin, Germany, 2011; Volume 2. [Google Scholar]
  68. Wächter, A.; Biegler, L. On the Implementation of an Interior-Point Filter Line-Search Algorithm for Large-Scale Nonlinear Programming. Math. Program. 2006, 106, 25–57. [Google Scholar] [CrossRef]
  69. Cheng, Y.; Weng, S.; Yu, L.; Zhu, N.; Yang, M.; Yuan, Y. The Role of Hyperthermia in the Multidisciplinary Treatment of Malignant Tumors. Integr. Cancer Ther. 2019, 18, 153473541987634. [Google Scholar] [CrossRef] [Green Version]
  70. Pennes, H. Analysis of tissue and arterial blood temperatures in the resting human forearm. J. Appl. Physiol. 1948, 1, 93–122. [Google Scholar] [CrossRef]
  71. Cheng, K.; Stakhursky, V.; Stauffer, P.; Dewhirst, M.; Das, S. Online feedback focusing algorithm for hyperthermia cancer treatment. Int. J. Hyperth. 2007, 23, 539–554. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  72. Deng, Z.S.; Liu, J. Blood perfusion-based model for characterizing the temperature fluctuation in living tissues. Phys. A Stat. Mech. Its Appl. 2001, 300, 521–530. [Google Scholar] [CrossRef]
  73. Liu, J. Uncertainty analysis for temperature prediction of biological bodies subject to randomly spatial heating. J. Biomech. 2002, 34, 1637–1642. [Google Scholar] [CrossRef] [PubMed]
  74. Neufeld, E. High Resolution Hyperthermia Treatment Planning. Ph.D. Thesis, ETH Zürich, Zürich, Switzerland, 2008. [Google Scholar]
  75. Greef, M.; Kok, H.; Correia, D.; Borsboom, P.P.; Bel, A.; Crezee, H. Uncertainty in hyperthermia treatment planning: The need for robust system design. Phys. Med. Biol. 2011, 56, 3233–3250. [Google Scholar] [CrossRef]
  76. Canters, R.; Franckena, M.; Zee, J.; Rhoon, G. Optimizing deep hyperthermia treatments: Are locations of patient pain complaints correlated with modelled SAR peak locations? Phys. Med. Biol. 2011, 56, 439–451. [Google Scholar] [CrossRef]
  77. Voyer, D.; Musy, F.; Nicolas, L.; Perrussel, R. Comparison of Methods for Modeling Uncertainties in a 2D Hyperthermia Problem. Prog. Electromagn. Res. B 2009, 11, 189–204. [Google Scholar] [CrossRef] [Green Version]
  78. Clegg, S.; Samulski, T.; Murphy, K.; Rosner, G.; Dewhirst, M. Inverse techniques in hyperthermia: A sensitivity study. IEEE Trans. Biomed. Eng. 1994, 41, 373–382. [Google Scholar] [CrossRef]
  79. Köhler, T.; Maass, P.; Wust, P.; Virchow, R.; Berlin, K. Efficient Methods in Hyperthermia Treatment Planning. In Surveys on Solution Methods for Inverse Problems; Springer: Vienna, Austria, 2001; pp. 155–167. [Google Scholar] [CrossRef]
  80. Schiela, A.; Weiser, M. Barrier methods for a control problem from hyperthermia treatment planning. In Proceedings of the Recent Advances in Optimization and its Applications in Engineering; Springer: Heidelberg, Germany, 2010; pp. 419–428. [Google Scholar] [CrossRef]
Figure 1. Deterministic tubes to the feasible sets of inner and outer problems over the prediction horizon for decreasing values of the approximation parameter τ 1 > τ 2 > τ 3 > τ 4 .
Figure 1. Deterministic tubes to the feasible sets of inner and outer problems over the prediction horizon for decreasing values of the approximation parameter τ 1 > τ 2 > τ 3 > τ 4 .
Mathematics 11 01372 g001
Figure 2. Implementation flow chart.
Figure 2. Implementation flow chart.
Mathematics 11 01372 g002
Figure 3. Controlled tissue domain.
Figure 3. Controlled tissue domain.
Mathematics 11 01372 g003
Figure 4. Moving prediction horizon of finite dimensional representation.
Figure 4. Moving prediction horizon of finite dimensional representation.
Mathematics 11 01372 g004
Figure 5. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 0.1 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Figure 5. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 0.1 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Mathematics 11 01372 g005
Figure 6. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 0.01 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Figure 6. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 0.01 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Mathematics 11 01372 g006
Figure 7. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 5 × 10 5 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Figure 7. Mean temperature realizations E [ T * ( u ( 0 , H ] * ( t , x ) , ξ , t , x ) ] on the first prediction horizon by τ = 5 × 10 5 ; (a) O A , t = 0.025 ; (b) O A , t = 0.2 ; (c) O A , t = 1.0 ; (d) O A , t = 2.0 ; (e) I A , t = 0.025 ; (f) I A , t = 0.2 ; (g) I A , t = 1.0 ; (h) I A , t = 2.0 .
Mathematics 11 01372 g007
Figure 8. Maximum values of E [ T * ( u ( 0 , H ] * , ξ , t , x ) ] over time for different values of τ .
Figure 8. Maximum values of E [ T * ( u ( 0 , H ] * , ξ , t , x ) ] over time for different values of τ .
Mathematics 11 01372 g008
Figure 9. Optimal functional values of the inner and outer problems for decreasing values of τ .
Figure 9. Optimal functional values of the inner and outer problems for decreasing values of τ .
Mathematics 11 01372 g009
Figure 10. Maximum values of E [ T * ( u ( t ^ z , t ^ z + H ] * ( t , x ) , ξ , t , x ) ] , t ^ z ( t ^ 1 , t ^ 1 + H ] (CCPDEs) and E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] (IA and OA of CCHTMPC) by τ = 5 × 10 5 .
Figure 10. Maximum values of E [ T * ( u ( t ^ z , t ^ z + H ] * ( t , x ) , ξ , t , x ) ] , t ^ z ( t ^ 1 , t ^ 1 + H ] (CCPDEs) and E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] (IA and OA of CCHTMPC) by τ = 5 × 10 5 .
Mathematics 11 01372 g010
Figure 11. Temperature distributions T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ i , t , x ) , i = 1 , , N (all solutions) and E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] (mean trajectory) over time for τ = 5 × 10 5 .
Figure 11. Temperature distributions T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ i , t , x ) , i = 1 , , N (all solutions) and E [ T * ( u ( t ^ 1 , t ^ 1 + H ] * ( t , x ) , ξ , t , x ) ] (mean trajectory) over time for τ = 5 × 10 5 .
Mathematics 11 01372 g011
Table 1. Inner and outer approximation problems.
Table 1. Inner and outer approximation problems.
Inner-approximation problemOuter-approximation problem
( I A ) τ min u J ( u ) subject to : ψ ( u , τ ) 1 α , u U d , ( O A ) τ min u J ( u ) subject to : φ ( u , τ ) α , u U d ,
Table 2. Optimization parameters.
Table 2. Optimization parameters.
Parameter Value
u m a x Maximum admissible control30
u m i n Minimum admissible control0
λ Control regularization parameter 10 5
χ Moreau–Yosida regularization smoothing parameter 1.6 × 10 3
γ Moreau–Yosida regularization penalization parameter 5.0 × 10 5
α Reliability level0.95
m 1 , m 2 Geletu–Hoffmann function constants1.0, 0.4
Table 3. Time consumption in hours for the solution of inner and outer problems with decreasing τ .
Table 3. Time consumption in hours for the solution of inner and outer problems with decreasing τ .
Problem τ = 10 1 τ = 10 2 τ = 10 3 τ = 10 4 τ = 5 × 10 5
OA1.253.6710.3110.6510.18
IA1.149.2915.3215.0415.20
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Voropai, R.; Geletu, A.; Li, P. Model Predictive Control of Parabolic PDE Systems under Chance Constraints. Mathematics 2023, 11, 1372. https://doi.org/10.3390/math11061372

AMA Style

Voropai R, Geletu A, Li P. Model Predictive Control of Parabolic PDE Systems under Chance Constraints. Mathematics. 2023; 11(6):1372. https://doi.org/10.3390/math11061372

Chicago/Turabian Style

Voropai, Ruslan, Abebe Geletu, and Pu Li. 2023. "Model Predictive Control of Parabolic PDE Systems under Chance Constraints" Mathematics 11, no. 6: 1372. https://doi.org/10.3390/math11061372

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop