Bi-Level Sustainability Planning for Integrated Energy Systems Considering Hydrogen Utilization and the Bilateral Response of Supply and Demand

Abstract

Under the background of “double carbon” and sustainable development, aimed at the problem of resource capacity planning in the integrated energy system (IES), at improving the economy of system planning operation and renewable energy (RE) consumption, and at reducing carbon emissions, this paper proposes a multi-objective bi-level sustainability planning method for IES considering the bilateral response of supply and demand and hydrogen utilization. Firstly, the multi-energy flow in the IES is analyzed, constructing the system energy flow framework, studying the support ability of hydrogen utilization and the bilateral response of supply and demand to system energy conservation, emission reduction and sustainable development. Secondly, a multi-objective bi-level planning model for IES is constructed with the purpose of optimizing economy, RE consumption, and carbon emission. The non-dominated sorting genetic algorithm II (NSGA-II) and commercial solver Gurobi are used to solve the model and, through the simulation, verify the model’s effectiveness. Finally, the planning results show that after introducing the hydrogen fuel cells, hydrogen storage tank, and bilateral response, the total costs and carbon emissions decreased by 29.17% and 77.12%, while the RE consumption rate increased by 16.75%. After introducing the multi-objective planning method considering the system economy, RE consumption, and carbon emissions, the system total cost increased by 0.34%, the consumption rate of RE increased by 0.6%, and the carbon emissions decreased by 43.61t, which effectively provides reference for resource planning and sustainable development of IES.

1. Introduction

The depletion of fossil energy reserves and the greenhouse effect have intensified the pressure on energy conservation and emission reduction in various countries to promote energy reform. The proportion of clean and carbon-free RE generation in the energy field is increasing. However, the uncertainty and randomness of RE generation have led to increasingly prominent RE consumption problems [1,2]. Faced with the dilemma of ecological reform and difficulties in RE consumption, China formulated the dual carbon target at the 75th session of the United Nations General Assembly to establish a cleaner RE supply system and realize green and low-carbon transformation in the energy field [3].

The IES can break the single form of energy flow in the original power system, utilizing the complementary characteristics of electricity, thermal, cooling, hydrogen, and other energy sources. The flexibility of energy flow in each link of source–grid–load–storage can be effectively utilized to improve the proportion of RE consumption and reduce carbon emissions [4,5,6]. As an essential direction for future clean and low-carbon energy development, IES planning research with multi-energy flow is necessary. This research aims to optimize the system’s internal resource planning and contribute to the realization of green and low-carbon transformation in the energy field.

Energy storage is an essential schedulable and flexible resource in the IES, and can realize the time shift of RE generation and improve the consumption rate [7,8]. With the advantages of long storage time, high energy density, and being pollution-free, hydrogen has an extensive range of application prospects in RE consumption, transportation, and energy supply [9,10]. Reference [11] developed an IES optimization model with water and hydrogen integration, improving the overall resource, social, and economic value. Reference [12] constructed a comprehensive hydrogen utilization unit including the electrolytic bath (EB), hydrogen storage tank (HST), hydrogen fuel cell (HFC), methane reactor, and other equipment. The hydrogen utilization unit is integrated in the optimal scheduling of the multi-energy systems with a high proportion of wind turbines (WT) and photovoltaic (PV). Reference [13] established an electricity–thermal–hydrogen integrated energy planning model to improve the RE consumption ability through hydrogen multi-energy conversion. References [14,15] introduced electrolytic water hydrogen production and storage links into the park system in the system planning model, effectively broadening the RE consumption scope and improving energy utilization efficiency. However, the above applications mainly focus on hydrogen production to promote RE consumption in the IES operation stage and do not study the role and influence of the full utilization link of hydrogen in IES resource planning.

On the energy load side, there is a substantial mismatch between RE generation and multi-loads. On the one hand, multi-loading system users can adjust their load consumption and time to achieve comprehensive energy demand response (DR), which is essential in promoting the real-time matching between system energy supply and load demand [16,17]. Reference [18] established the IES planning model considering the electricity price incentive DR and optimized the load curve to improve the system economy. Reference [19] proposed an optimization model that includes a DR program based on incentives for the energy management system and aims to maximize the rewards for users participating in DR. Based on the time-sharing electricity price, References [20,21] constructed the scheduling model of IES to coordinate various forms of energy to match the RE generation. On the other hand, in the face of the different peak characteristics of the electric load and thermal load in the IES, the organic Ranken cycle (ORC) waste thermal generation is introduced into the source side to effectively improve the flexibility of the generation side [22]. Reference [23] introduces waste thermal generation equipment, flexible cogeneration units, and multi-energy DR to form a generalized DR model. Reference [24] introduces power-to-gas, HFC, and ORC to construct variable efficiency operation models of cogeneration units, which can effectively improve energy utilization efficiency and the operation economy. However, the above studies focused on the DR mode of the load-side multi-energy loads and did not deeply explore the combined utilization of load-side DR and source-side ORC waste thermal generation in IES planning.

With the gradual transformation of the energy structure to clean and low-carbon energy, in addition to the economy of IES, it is necessary to study carbon emissions and RE consumption in IES planning [25,26]. The research is mainly divided into two parts. In the first part, carbon emissions or RE consumption rates are used as constraints in the IES planning model [27]. The other part introduces the carbon emission and RE curtailment costs in the planning model, which can use the economic objective to constrain carbon emission and RE curtailment [6]. Reference [28] introduces carbon emission cost into the economic optimization objective of the IES planning model to ensure the economy and environmental protection of IES planning results. Reference [29] constructs a reward–penalty step carbon trading model and proposes the IES planning method with the objective of minimizing operation costs and carbon trading costs. To improve the RE consumption level, Reference [30] optimizes the planning capacity of renewable resources by considering the RE curtailment cost. In addition, Reference [31] analyzes the importance of economic, low-carbon, clean, and self-balancing features in future system planning for the new power system with a high proportion of RE and multi-energy loads.

The following can be noted from the existing literature: (1) Hydrogen utilization in IES operation is a research hotspot, but the research about the hydrogen utilization in the IES resource planning is finite. (2) The DR of load-side resources has been widely considered, but the impact of the DR and ORC waste thermal generation on the IES resource planning needs to be studied. (3) The studies focus on the resource planning of IES with the single objective of economy, ignoring the factors such as RE utilization and low carbon emission. Inspired by this, this paper proposes an IES multi-objective bi-level planning method considering the bilateral response of supply and demand and hydrogen utilization. The main contributions are as follows:

(1) The supporting ability of hydrogen generation and storage utilization for RE consumption and storage, and the supply ability of HFC to meet multi-energy loads are fully utilized. The whole link utilization of hydrogen is considered in the IES planning model.

(2) The integrated DR mechanism of electricity-thermal-cooling-hydrogen and the ORC waste thermal generation is incorporated into the IES operation. The bilateral response of supply and demand is utilized to alleviate the pressure on the system’s energy supply and reduce the reliance on the external electricity grid.

(3) A multi-objective bi-level planning model of IES is established to minimize the total planning cost and carbon emissions and maximize the RE consumption rate. The planning results of the cases verify the effectiveness and superiority of the proposed planning method.

2. IES Framework with Multi-Energy Loads and Mathematical Model

The operation framework of the IES with hydrogen utilization and the bilateral response of supply and demand is shown in Figure 1. WT, PV, external grid, and HFC in the system are used to meet the electricity demand of conventional electric load and flexible load such as EB, electric chillers (EC) and electric vehicle (EV). HFC combined with the electric heating unit (EH) supplies the thermal load. EC combined with the absorption chiller (AC) meets the system cooling load demand. The hydrogen energy supply of the system mainly comes from EB and HST.

As shown in Figure 1, IES integrates various forms of energy, including electricity, thermal, cooling, hydrogen, and transportation through multiple energy conversion equipment. The system is connected to the external power grid through tie lines. On the one hand, the hydrogen utilization unit comprises EB, HST, and HFC. During periods of high RE generation and low multi-load demand, there is a system energy curtailment, and the EB can actively respond to the RE consumption demand. The HST provides sufficient consumption space for the EB to realize the energy conversion of green electricity to green hydrogen. During periods of low RE generation and high multi-load demand, there is a system energy shortage, and HFC can convert hydrogen into electricity and thermal energy to bridge the system energy gap, thereby realizing the conversion of green hydrogen to green energy. On the other hand, the DR is introduced to adjust the fluctuation of the multi-load curve, broadening the online space of RE.

At the same time, the ORC waste thermal generation equipment is utilized to form the flexible response of the source side and can flexibly adapt to the HFC energy output and the electricity and thermal demand. The ORC has similar effects to the DR of the load side, forming the bilateral response of supply and load to improve the system energy utilization efficiency.

2.1. Mathematical Model of Equipment

2.1.1. EB Model

The relationship between hydrogen production and electricity consumption in the EB can be expressed as follows [32]:

$$M_t^{{\mathrm{H}}_2}={\eta }_{{\mathrm{H}}_2}{\nu }_{{\mathrm{H}}_2}{P}_t^{{\mathrm{H}}_2}$$

(1)

where $M_t^{{\mathrm{H}}_2}$ is the hydrogen production of the EB; ${\eta }_{{\mathrm{H}}_2}$ is the hydrogen production efficiency; ${\nu }_{{\mathrm{H}}_2}$ is the hydrogen production per unit of electricity; and $P_t^{{\mathrm{H}}_2}$ is the electricity consumed by the EB.

2.1.2. HST Model

$$M_t^{\mathrm{HST}}=\left(1-u_{\mathrm{loss}}^{\mathrm{HST}}\right)M_{t-1}^{\mathrm{HST}}+{\eta }^{\mathrm{cha}}{M}_t^{\mathrm{cha}}-{\eta }^{\mathrm{dis}}{M}_t^{\mathrm{dis}}$$

(2)

where $M_t^{\mathrm{HST}}$ and $M_{t-1}^{\mathrm{HST}}$ are the amount of hydrogen stored in the HST during t and t − 1 period; $u_{\mathrm{loss}}^{\mathrm{HST}}$, ${\eta }^{\mathrm{cha}}$ and ${\eta }^{\mathrm{dis}}$ are the self-loss coefficient, hydrogen charging, and hydrogen discharging efficiency of the HST, respectively; $M_t^{\mathrm{cha}}$ and $M_t^{\mathrm{dis}}$ are the hydrogen charging and discharging energy of the HST.

2.1.3. HFC Model

As the essential coupling equipment for hydrogen conversion, HFC can convert the energy generated by hydrogen combustion into electricity and thermal generation [33]. The model is as follows:

$$\left\{\begin{array}{l}{P}_t^{\mathrm{hfc}}={\eta }_{\mathrm{e}}^{\mathrm{hfc}}{M}_t^{\mathrm{hfc}}/{\nu }_{\mathrm{hfc}}\\ H_t^{\mathrm{hfc}}={\eta }_{\mathrm{h}}^{\mathrm{hfc}}{M}_t^{\mathrm{hfc}}/{\nu }_{\mathrm{hfc}}\end{array}\right.$$

(3)

where $P_t^{\mathrm{hfc}}$ and $H_t^{\mathrm{hfc}}$ are the electricity and thermal generation of HFC; ${\eta }_{\mathrm{e}}^{\mathrm{hfc}}$ and ${\eta }_{\mathrm{h}}^{\mathrm{hfc}}$ are the electricity and thermal conversion efficiency of HFC; $M_t^{\mathrm{hfc}}$ is the hydrogen energy consumption of HFC; ${\nu }_{\mathrm{hfc}}$ is the needed hydrogen to produce a unit of energy for HFC.

2.1.4. EH Model

The conversion rate of EH can be close to 100%, making it a crucial thermoelectric coupling device in the IES. The operation model of EH is as follows:

$$H_t^{\mathrm{eh}}={\eta }_{\mathrm{eh}}{P}_t^{\mathrm{eh}}$$

(4)

where $H_t^{\mathrm{eh}}$ and $P_t^{\mathrm{eh}}$ are the thermal generation and electricity consumption of EH, respectively; ${\eta }_{\mathrm{eh}}$ is the operating efficiency of EH.

2.1.5. EC and AC Model

The EC and AC jointly supply the cooling demand, using electricity and thermal energy as raw materials, respectively. The operation models are shown in Equations (5) and (6):

$$C_t^{\mathrm{ec}}={\eta }_{\mathrm{ec}}{P}_t^{\mathrm{ec}}$$

(5)

$$C_t^{\mathrm{ac}}={\eta }_{\mathrm{ac}}{H}_t^{\mathrm{ac}}$$

(6)

where $C_t^{\mathrm{ec}}$ and $C_t^{\mathrm{ac}}$, $P_t^{\mathrm{ec}}$ and $H_t^{\mathrm{ac}}$, ${\eta }_{\mathrm{ec}}$ and ${\eta }_{\mathrm{ac}}$ are the cooling generation, operating efficiency, and energy consumption of the EC and the AC, respectively.

3. Bilateral Response of Supply and Demand Based on Multi-Loads and ORC

3.1. DR Modeling of Conventional Electric Load, HV, and EV Charging Load

To actively play the flexible regulation ability of the system load side, in addition to the conventional electric load to participate in the DR, this paper directs the EV and hydrogen vehicle (HV) charging loads, a vital controllable resource, to participate in the DR. The conventional electric, EV, and HV charging loads can be divided into fixed, reducible, and transferable types. The fixed load does not participate in the DR. The reducible load can be reduced to a certain extent during the scheduling cycle. The transferable load can transfer part of the load to another period and lessen the peak–valley difference [34].

$$P_{m,t}^{\mathrm{load}}=P_{m,t}^{\mathrm{fix}}+P_{m,t}^{\mathrm{cut}}+P_{m,t}^{\mathrm{ela}}$$

(7)

where m represents the load type; $P_{m,t}^{\mathrm{load}}$ is the m-th type load value; $P_{m,t}^{\mathrm{fix}}$ is the fixed load value; $P_{m,t}^{\mathrm{cut}}$ is the reduced load value; $P_{m,t}^{\mathrm{ela}}$ is the transferable load value.

3.2. The DR Modeling of Thermal and Cooling Load Considering the Fuzzy Temperature Perception of the Human Body

Considering the energy storage characteristics of the environment and the fuzziness of temperature perception of the human body, the thermal and cooling loads emerge as high-quality, flexible resources. The user’s ambient temperature is controlled within the acceptable temperature range of the human body, thereby achieving flexible thermal and cooling demand regulation effects. The models of flexible thermal and cooling loads can be expressed in Equations (8) and (9), respectively.

$$H_t^{\mathrm{load}}=Gw\left(T_{\mathrm{in},t}^{\mathrm{h}}-T_{\mathrm{out},t}^{\mathrm{h}}\right)+CG\left(T_{\mathrm{in},t}^{\mathrm{h}}-T_{\mathrm{in},t-1}^{\mathrm{h}}\right)$$

(8)

$$C_t^{\mathrm{load}}=Gw\left(T_{\mathrm{out},t}^{\mathrm{c}}-T_{\mathrm{in},t}^{\mathrm{c}}\right)+CG\left(T_{\mathrm{in},t-1}^{\mathrm{c}}-T_{\mathrm{in},t}^{\mathrm{c}}\right)$$

(9)

where $H_t^{\mathrm{load}}$ and $C_t^{\mathrm{load}}$ are the thermal load in winter and cooling load in summer; G is the thermal area; C is the thermal capacity per unit thermal area; w is the thermal dissipation coefficient of the temperature difference between inside and outside buildings; $T_{\mathrm{in},t}^{\mathrm{h}}$ and $T_{\mathrm{out},t}^{\mathrm{h}}$ are the indoor and outdoor temperatures in winter, respectively; $T_{\mathrm{out},t}^{\mathrm{c}}$ and $T_{\mathrm{in},t}^{\mathrm{c}}$ are the indoor and outdoor temperatures in summer, respectively.

The thermal and cooling loads participating in DR are as follows:

$$H_t^{\mathrm{dr}}=Sw\left(T_{\mathrm{in},\mathrm{best}}^{\mathrm{h}}-T_{\mathrm{out},t}^{\mathrm{h}}\right)-H_t^{\mathrm{load}}$$

(10)

$$C_t^{\mathrm{dr}}=Sw\left(T_{\mathrm{out},t}^{\mathrm{c}}-T_{\mathrm{in},\mathrm{best}}^{\mathrm{c}}\right)-C_t^{\mathrm{load}}$$

(11)

where $H_t^{\mathrm{dr}}$ and $C_t^{\mathrm{dr}}$ are the power of thermal and cooling loads participating in DR; $T_{\mathrm{in},\mathrm{best}}^{\mathrm{h}}$ and $T_{\mathrm{in},\mathrm{best}}^{\mathrm{c}}$ are the optimal temperature of the human body in winter and summer.

3.3. Flexible Response Model of ORC Waste Thermal Generation

It is challenging for the traditional load-side DR to guarantee a flexible energy supply and refined energy utilization. Considering the low thermal load in spring, summer, and autumn, and the difference in electricity and thermal demand, the ORC waste thermal power generation technology is introduced to improve the energy utilization efficiency of the system. It converts the rich thermal energy into electricity and improves the utilization rate of HFC. With the introduction of the ORC waste thermal generation, the HFC thermal energy generation is refined into two parts: $H_t^{\mathrm{hfcl}}$ directly supply the thermal load, and $H_t^{\mathrm{orc}}$ is sent to the ORC to produce electricity $P_t^{\mathrm{orc}}$. The ORC waste thermal generation model is as follows:

$$\left\{\begin{array}{l}{H}_t^{\mathrm{hfc}}=H_t^{\mathrm{orc}}+H_t^{\mathrm{hfcl}}\\ P_t^{\mathrm{orc}}={\eta }^{\mathrm{orc}}{H}_t^{\mathrm{orc}}\\ H_{\min }^{\mathrm{orc}}\le H_t^{\mathrm{orc}}\le H_{\max }^{\mathrm{orc}}\\ \Delta H_{\min }^{\mathrm{orc}}\le H_t^{\mathrm{orc}}-H_t^{\mathrm{orc}}\le \Delta H_{\max }^{\mathrm{orc}}\end{array}\right.$$

(12)

where ${\eta }^{\mathrm{orc}}$ is the power generation efficiency of ORC; $H_{\max }^{\mathrm{orc}}$ and $H_{\min }^{\mathrm{orc}}$ are the upper and lower limits of the ORC input power; $\Delta H_{\max }^{\mathrm{orc}}$ and $\Delta H_{\min }^{\mathrm{orc}}$ are the upper and lower limits of ORC ramping.

4. IES Resource Bi-Level Planning Model

A multi-objective bi-level planning model is established to comprehensively optimize the investment and operation costs, RE consumption rate, and carbon emissions. The upper layer optimizes the equipment capacity of IES, and the lower layer regularizes the optimal operation strategy of IES under typical days. The IES planning framework is shown in Figure 2.

To ensure the safe and stable operation of the IES at each period, this paper considers hydrogen utilization and the bilateral response on the supply and demand side in the operation stage, improving the system’s ability to deal with uncertain risks. Firstly, according to the initial resource planning scheme, the optimal operation results of IES are obtained by considering the operation cost of each typical day, and the operation results and objective parameters are fed to the upper model. Secondly, based on the feedback results, the upper model improved the resource planning scheme and passed it to the lower layer for operation. Finally, the optimal resource planning scheme and the typical days’ operation plan were obtained by iterative interaction between the upper and lower layers.

4.1. The Upper-Level Planning Model Objective Functions

Under resource investment and planning constraints, the IES resource planning aims to achieve optimal planning and operational economy, maximize RE consumption, and minimize carbon emissions. The multi-objective functions are as follows:

4.1.1. Objective Function 1

The economic objective function of the IES resource optimal planning is to minimize the average annual total cost $C_{\mathrm{total}}$ during the system planning period. The economic objective function consists of two parts: the annualized investment cost $C_{\mathrm{inv}}$ and the operating cost $C_{\mathrm{ope}}$.

$$\min C_{\mathrm{total}}=C_{\mathrm{inv}}+C_{\mathrm{ope}}$$

(13)

$C_{\mathrm{inv}}$ includes the annualized investment cost of WT, PV, EB, and other equipment, as shown below.

$$\left\{\begin{array}{l}{C}_{\mathrm{inv}}={\displaystyle \sum _{n=1}^N}{c}_n{w}_{n}d\\ N\in \left\{\mathrm{WG},\mathrm{PV},\mathrm{EB},\mathrm{HST},\mathrm{HFC},\mathrm{CEV},\mathrm{CHV},\mathrm{ORC},\mathrm{EC},\mathrm{AC},\mathrm{EH}\right\}\end{array}\right.$$

(14)

where $d=r{\left(1+r\right)}^m/{\left(1+r\right)}^m-1$ is the cost coefficient of return on resource investment; r and m are the interest rate and service life of the equipment; N is the type of equipment. $c_n$ and $w_n$ are the unit investment cost and investment capacity of resource n, respectively.

4.1.2. Objective Function 2

The RE consumption rate in IES can be expressed as follows:

$$\max P_{\mathrm{RE}}={\displaystyle \sum _{s=1}^S}{\displaystyle \sum _{t=1}^T}{w}_s\left(\frac{P_{s,t}^{\mathrm{wt}}+P_{s,t}^{\mathrm{pv}}}{p_{s,t}^{\mathrm{wt}}{w}_{\mathrm{wt}}+p_{s,t}^{\mathrm{pv}}{w}_{\mathrm{pv}}}\right)$$

(15)

where $w_s$ is the number of typical days in each season; $S=4$, representing the four seasons; $p_{s,t}^{\mathrm{wt}}$ and $p_{s,t}^{\mathrm{pv}}$ are the forecast information of WT and PV generation in period t of a typical day in season s, respectively; $P_{s,t}^{\mathrm{wt}}$ and $P_{s,t}^{\mathrm{pv}}$ are the actual generation of WT and PV, respectively.

4.1.3. Objective Function 3

Carbon emission is an important indicator of environmental benefits. IES carbon emissions $M_{\mathrm{ce}}$ are calculated by electricity purchased from the external grid, as shown below:

$$\min M_{\mathrm{ce}}={\displaystyle \sum _{s=1}^S}{\displaystyle \sum _{t=1}^T}{w}_s{\alpha }_{\mathrm{grid}}{P}_{s,t}^{\mathrm{grid}}$$

(16)

where ${\alpha }_{\mathrm{grid}}$ is the carbon emission factor of electricity purchased from the external grid; $P_{s,t}^{\mathrm{grid}}$ is the electricity purchased from the external grid in period t of a typical day in season s.

4.2. Upper-Level Model Constraints

Considering the actual operation scenario and the limitation of site size, resource planning needs to set an upper limit on the planning capacity of equipment, as shown below.

$$0\le R_n^{\mathrm{inv}}\le R_{n,\max }^{\mathrm{inv}}$$

(17)

where $R_n^{\mathrm{inv}}$ and $R_{n,\max }^{\mathrm{inv}}$ are the planning capacity and the upper limit of equipment n, respectively.

To ensure the reliability of the energy supply, the resource planning capacity of IES should not be less than the maximum demand of the multi-energy load in each typical day, as referred to in reference [35].

4.3. The Lower-Level Operating Model Objective Function

The objective of the IES lower operating model is to minimize the sum of typical daily operation costs $C_{\mathrm{ope}}$, which includes the electricity purchased cost $C_{\mathrm{grid}}$, the RE curtailment cost $C_{\mathrm{ab}}$, the DR compensation cost $C_{\mathrm{dr}}$, the HV and EV charging cost $C_{\mathrm{vehicle}}$, and the environmental cost $C_{\mathrm{env}}$. The objective function can be expressed as follows.

$$\min C_{\mathrm{ope}}=C_{\mathrm{grid}}+C_{\mathrm{ab}}+C_{\mathrm{dr}}+C_{\mathrm{vehicle}}+C_{\mathrm{env}}$$

(18)

$$C_{\mathrm{grid}}={\displaystyle \sum _{s=1}^S}{\displaystyle \sum _{t=1}^T}{w}_s{w}_t^{\mathrm{e}}{P}_{s,t}^{\mathrm{grid}}$$

(19)

$$C_{\mathrm{ab}}={\displaystyle \sum _{s=1}^S}{\displaystyle \sum _{t=1}^T}{w}_s{w}_{\mathrm{ab}}\left(p_{s,t}^{\mathrm{wt}}{w}_{\mathrm{wt}}+p_{s,t}^{\mathrm{pv}}{w}_{\mathrm{pv}}-P_{s,t}^{\mathrm{wt}}-P_{s,t}^{\mathrm{pv}}\right)$$

(20)

$$C_{\mathrm{dr}}={\displaystyle \sum _{s=1}^S}{\displaystyle \sum _{t=1}^T}{\displaystyle \sum _{m=1}^M}{w}_s{w}_m\left(\left|P_{m,t}^{\mathrm{ela}}\right|+P_{m,t}^{\mathrm{cut}}+H_t^{\mathrm{dr}}+C_t^{\mathrm{dr}}\right)$$

(21)

$$C_{\mathrm{vehicle}}={\displaystyle \sum _{s=1}^S}{\displaystyle \sum _{t=1}^T}{w}_s\left(w_t^{\mathrm{e}}{P}_{s,t}^{\mathrm{ev}}+w^{\mathrm{h}}{M}_{s,t}^{\mathrm{hv}}\right)$$

(22)

$$C_{\mathrm{env}}={\displaystyle \sum _{s=1}^S}{\displaystyle \sum _{t=1}^T}{\displaystyle \sum _{u=1}^U}{w}_s{R}_u\left(k_u^{\mathrm{grid}}{P}_{s,t}^{\mathrm{grid}}+k_u^{\mathrm{ov}}{V}_{s,t}^{\mathrm{ov}}\right)$$

(23)

where $w_t^{\mathrm{e}}$ is the time-of-use electricity price; $w_m$ is the compensation price of type m load participating in DR; $P_{s,t}^{\mathrm{ev}}$ and $M_{s,t}^{\mathrm{hv}}$ are the charging demands of EV and HV; $w^{\mathrm{h}}$ is the hydrogen price; U is the number of pollutant types, including CO₂, SO₂, NO_x, etc.; $R_u$ is the environmental value of pollutant u; and $k_u^{\mathrm{grid}}$ and $k_u^{\mathrm{ov}}$ are the emission intensity of pollutant u about the electricity purchased and the driving of fuel vehicles (FV).

4.4. Lower-Level Constraints

The lower-level operating constraints consider the supply and demand balance of electricity, thermal, cooling, and hydrogen, conventional equipment operation constraints, and the load-side DR constraints.

4.4.1. Multi-Energy Power Balance Constraints

$$P_{s,t}^{\mathrm{load}}+P_{s,t}^{{\mathrm{H}}_2}+P_{s,t}^{\mathrm{ev}}+P_{s,t}^{\mathrm{ec}}+P_{s,t}^{\mathrm{eh}}=P_{s,t}^{\mathrm{wt}}+P_{s,t}^{\mathrm{pv}}+P_{s,t}^{\mathrm{grid}}+P_{s,t}^{\mathrm{hfc}}+P_{s,t}^{\mathrm{orc}}$$

(24)

$$M_{s,t}^{\mathrm{dis}}+M_{s,t}^{{\mathrm{H}}_2}=M_{s,t}^{\mathrm{cha}}+M_{s,t}^{\mathrm{hfc}}+M_{s,t}^{\mathrm{hv}}$$

(25)

$$H_{s,t}^{\mathrm{hfcl}}+H_{s,t}^{\mathrm{eh}}=H_{s,t}^{\mathrm{ac}}+H_{s,t}^{\mathrm{load}}$$

(26)

$$C_{s,t}^{\mathrm{ec}}+C_{s,t}^{\mathrm{ac}}=C_{s,t}^{\mathrm{load}}$$

(27)

Equations (24)–(27) represent the supply and demand balance of electricity, hydrogen, thermal, and cooling, respectively.

4.4.2. Equipment Operation Constraints

The multi-energy conversion equipment operation of IES should meet the planned resource capacity constraints and upper and lower limit constraints of ramping capacity.

$$\left\{\begin{array}{l}{P}_{\min }^n\le P_t^n\le P_{\max }^n\\ P_{\mathrm{down}}^n\le P_{t+1}^n-P_t^n\le P_{\mathrm{up}}^n\end{array}\right.$$

(28)

where $P_{\max }^n$ and $P_{\min }^n$ are the upper and lower limits of the generation of equipment n, respectively; $P_{\mathrm{up}}^n$ and $P_{\mathrm{down}}^n$ are the upper and lower limits of the ramping rate of equipment n, respectively.

4.4.3. Energy Storage Equipment Constraints

The HST is the bridge between the hydrogen production and utilization of IES. To ensure the operation continuity in the next scheduling period, the rated storage capacity constraint, the exchange power constraint, and the scheduling continuity constraint should be satisfied.

$$\left\{\begin{array}{l}0\le M_t^{\mathrm{HST}}\le M_{\max }^{\mathrm{HST}}\\ {0\le M}_t^{\mathrm{cha}}\le M_{\mathrm{cha},\max }^{\mathrm{HST}}\\ {0\le M}_t^{\mathrm{dis}}\le M_{\mathrm{cha},\max }^{\mathrm{dis}}\\ M_0^{\mathrm{HST}}=M_{\mathrm{T}}^{\mathrm{HST}}=0.5M_{\max }^{\mathrm{HST}}\end{array}\right.$$

(29)

where $M_{\max }^{\mathrm{HST}}$ is the upper limit of HST capacity; $M_{\mathrm{cha},\max }^{\mathrm{HST}}$ and $M_{\mathrm{cha},\max }^{\mathrm{dis}}$ are the upper limit of HST capacity, respectively; $M_0^{\mathrm{HST}}$ and $M_{\mathrm{T}}^{\mathrm{HST}}$ are the residual hydrogen energy at the initial period and the final period of the HST, respectively.

4.4.4. External Grid Interaction Constraints

IES is connected to the external grid through tie lines. When the internal energy supply of IES is insufficient, the IES purchases electricity from the external grid to meet the load demand. However, considering the effects of the randomness of WT and PV generation for the external grid operation, the electricity transmission from the integrated energy system to the main grid is not considered in this paper.

$$P_{\min }^{\mathrm{grid}}\le P_{s,t}^{\mathrm{grid}}\le P_{\max }^{\mathrm{grid}}$$

(30)

where $P_{\min }^{\mathrm{grid}}$ and $P_{\max }^{\mathrm{grid}}$ are the limit values of the electricity purchased from the external grid, respectively.

4.4.5. Load-Side DR Constraints

For conventional electric load, EV, and HV charging load to participate in DR, the following constraints should be satisfied:

$$\left\{\begin{array}{l}{P}_{m,\min }^{\mathrm{cut}}\le P_{m,t}^{\mathrm{cut}}\le P_{m,\max }^{\mathrm{cut}}\\ P_{m,\min }^{\mathrm{ela}}\le P_{m,t}^{\mathrm{ela}}\le P_{m,\max }^{\mathrm{ela}}\\ {\sum }_0^T{P}_{m,t}^{\mathrm{ela}}=0\end{array}\right.$$

(31)

where $P_{m,\max }^{\mathrm{cut}}$ and $P_{m,\min }^{\mathrm{cut}}$ are the upper and lower limits of the type load m that can be reduced, respectively; $P_{m,\max }^{\mathrm{ela}}$ and $P_{m,\min }^{\mathrm{ela}}$ are the upper and lower limits of the type load m that can be transferred, respectively.

For flexible thermal and cooling loads to participate in DR, the user’s ambient temperature should meet the human body’s temperature comfort range, as shown in Equations (32) and (33).

$$T_{\mathrm{best}}^{\mathrm{h}}+I_{\min }^{\mathrm{h}}\le T_{\mathrm{in},t}^{\mathrm{h}}\le T_{\mathrm{best}}^{\mathrm{h}}+I_{\max }^{\mathrm{h}}$$

(32)

$$T_{\mathrm{best}}^{\mathrm{c}}+I_{\min }^{\mathrm{c}}\le T_{\mathrm{in},t}^{\mathrm{c}}\le T_{\mathrm{best}}^{\mathrm{c}}+I_{\max }^{\mathrm{c}}$$

(33)

where $I_{\min }^{\mathrm{h}}$ and $I_{\max }^{\mathrm{h}}$, $I_{\min }^{\mathrm{c}}$ and $I_{\max }^{\mathrm{c}}$ are the upper and lower limits of the acceptable ambient temperature changes in winter and summer, respectively.

5. Solution Method

The above bi-level planning model is a nonlinear and multi-objective planning problem. In the upper-level optimization problem, the relationship among multiple objective functions is usually non-dominated, and the optimal solution of a single objective function rarely dominates all feasible solutions of the other objective functions. Therefore, the NSGA-II is introduced to solve the above problem, which adopts a fast non-dominated sorting technique, crowding comparison operator, and elitism strategy to deal with nonlinear multi-objective optimization problems. A multi-objective non-dominated optimal solution set with uniform distribution and sound diversity is obtained [36]. For the multi-constrained nonlinear model in the lower layer, GUROBI, a commercial solver with efficient searchability and suitable for multiple output constraints, is used to optimize the equipment operation plan of each typical day. Thereby, a solution algorithm suitable for the two-layer resource planning model is formed. The specific process is shown in Figure 3.

6. Case Studies

The IES with hydrogen utilization and the bilateral response of supply and demand shown in Figure 1 are used as the test system to verify the effectiveness of the proposed planning model and solution algorithm. Each typical day is treated as a 24 h scheduling cycle, with the scheduling interval set to 1 h. The parameters of NSGA-II are configured as follows: the population size is 50; the maximum number of iterations is 50; the genetic algorithm’s crossover probability is 0.9; the crossover distribution index is 20; the mutation probability is 0.1; and the mutation distribution index is 20. The time-of-use electricity prices are listed in

Table 1. The time-of-use electricity price.

Periods	Specific Period	Electricity Price (CNY/kWh)
Peak period	8:00–11:00, 18:00–21:00	0.804
Flat period	6:00–7:00, 12:00–17:00	0.55
Valley period	1:00–5:00, 22:00–24:00	0.295

. The unit investment cost and service life of IES resources are shown in

Table 2. The unit investment cost and service life of IES resources [35].

Equipment	Parameter	Value	Equipment	Parameter	Value
WT	Unit investment cost (CNY/kW)	8000	PV	Unit investment cost (CNY/kW)	5000
	Lifetime (years)	25		Lifetime(years)	25
EB	Unit investment cost (CNY/kW)	2800	HFC	Unit investment cost (CNY/kW)	3010
	Lifetime (years)	20		Lifetime(years)	20
HST	Unit investment cost (CNY/kW)	2079	CEV	Unit investment cost (CNY/kW)	1000
	Lifetime (years)	20		Lifetime(years)	15
CHV	Unit investment cost (CNY/kW)	3325	ORC	Unit investment cost (CNY/kW)	3600
	Lifetime (years)	15		Lifetime (years)	20

.

Regarding scenario setting, due to prediction information errors, the uncertainty of multi-energy loads, and RE should be considered. In this paper, typical seasonal scenarios are generated using K-means clustering to address this uncertainty. First, it is assumed that the multi-energy loads and RE for each season follow a normal distribution with the mean equal to the prediction information and a variance of 0.4 times the prediction information. A total of 10,000 historical scenarios are generated for each season. The detailed clustering process is described in the literature [37]. Finally, the clustering yields the multi-energy load demand and RE scenarios for each season, shown in Figure 4. Case studies are run on a computer with a CPU Core i5-13400 and 16 GB RAM, and the GUROBI solver (Gurobi Optimization, LLC, Beaverton, OR, USA) was used through the YAMLIP toolbox in MATLAB R2024b to perform the solutions.

6.1. Analysis of Planning Results

6.1.1. Analysis of the Non-Dominated Optimal Solutions Set

When the number of objective functions is more than two, the non-dominated genetic algorithm cannot obtain a unique optimal solution and generates a non-dominated optimal solution set composed of multiple planning alternatives. The relationship of planning cost, RE consumption rate, and carbon emissions is shown in Figure 5, Figure 6 and Figure 7.

As shown in Figure 5, Figure 6 and Figure 7, the three objective functions are somewhat related. (1) System economy, RE consumption rate, and carbon emission are mutually exclusive, and multiple objective functions cannot be satisfied simultaneously. Among them, the increase in economic cost will lead to an increase in the RE consumption rate and a reduction in carbon emissions. This is because with the increase in economic costs, the planning capacity of RE and flexible resources such as EB, HST, and HFC increases. Meanwhile, the self-balancing ability of IES increases, and the dependence on the external grid is reduced. Therefore, carbon emissions are reduced, and the RE consumption rate is increased. (2) The RE consumption rate is inversely correlated with carbon emissions. This is because with the increase in electricity purchased from the external grid, the needed RE generation will be directly reduced, and the RE consumption rate will be reduced. At the same time, the external grid is the carbon emission source of IES, increasing the carbon emissions.

6.1.2. Selection of Optimal Planning Results

As mentioned above, in the multi-objective optimization problems, the optimal solution is obtained from the set of non-dominated solutions according to different indices. Each solution in the solution set is a feasible system planning scheme. Therefore, the scheme selection process is necessary to determine the optimal planning scheme. However, since the planning objective of IES aims to minimize costs, minimize carbon emissions, and maximize the consumption rate of RE, multiple optimization goals cannot be unified in terms of units. This paper employs the TOPSIS method to assess the proximity of each index to both optimal and inferior solutions, determining the multi-objective optimal compromise solution. The basic steps are as follows:

(1) Standardization of the objective function value.

Due to the different units of objective function values in multi-objective planning, the processing of forward and backward transformation needs to be carried out for the objective function values, as shown in Equation (34).

$$\left\{\begin{array}{l}{x}_{ij}^{-}=\frac{x_j^i-x_{\min }^i}{x_{\max }^i-x_{\min }^i}\\ x_{ij}^{-}=\frac{x_{\max }^i-x_j^i}{x_{\max }^i-x_{\min }^i}\end{array}\right.$$

(34)

where $x_j^i$ is the actual value of the i-th objective in the j-th planning scheme; $x_{\max }^i$ and $x_{\min }^i$ are the maximum and minimum values of the i-th objective, respectively; $x_{ij}^{+}$ and $x_{ij}^{-}$ are the forward indicators and reverse indicators, respectively.

(2) Assign weights to the objective function value.

The weight of each objective function is set to the same, and the three objective function results all follow the assumption of 1/3 for optimal solution selection [38]. Therefore, the weighting of the objective function value can be expressed as follows:

$$z_j^i=W^i{x}_j^i$$

(35)

where $W^i$ is the weight for the i-th objective, $W^i=1/3$.

(3) Determine the positive ideal solution $z_{+}$ and the negative ideal solution $z^{-}$

$$\left\{\begin{array}{l}{z}^{+}=\left\{\max z_j^i\left|j=1, 2, 3,\dots m\right.\right\}\\ z^{-}=\left\{\min z_j^i\left|j=1, 2, 3,\dots m\right.\right\}\end{array}\right.$$

(36)

(4) Calculate the relative distance between the planning scheme and the optimal solution $D_i^{+}$, as well as the worst solution $D_i^{-}$

$$\left\{\begin{array}{l}{D}_i^{+}=\sqrt{{\displaystyle \sum _{i=1}^3}{\left(z_j^i-z_j^{+}\right)}^2}\\ D_i^{-}=\sqrt{{\displaystyle \sum _{i=1}^3}{\left(z_j^i-z_j^{-}\right)}^2}\end{array}\right.$$

(37)

(5) Evaluate the relative closeness $R_j$ of the j-th planning scheme to the optimal solution

$$R_j=\frac{D_i^{-}}{D_i^{+}+D_i^{-}}$$

(38)

The higher the relative closeness is, the closer the j-th planning scheme is to the optimal solution, and the farther it is from the worst solution. Therefore, the planning scheme is better. The ideal solution based on the Topsis method is obtained by equal-weighted calculation, and the results are shown in

Table 3. IES resource planning results under the optimal scheme.

Optimal Scheme	Total Cost/105 CNY	RE Consumption Rate/%	Carbon Emission/104 kg
Ideal solution	105.08	99.81	2.450

and

Table 4. IES resource planning results under the optimal solution.

Resource Type	Resource Name	Planning Capacity (kW/kg)
Energy supply unit	WT	3014.4
	PV	1065.4
	CEV	1079.9
	CHV	26.58
Energy conversion unit	EB	1684.6
	HFC	767.45
	ORC	301.49
	EH	518.5
	EC	517.38
	AC	0
Energy storage unit	HST	198.29

.

Table 4 shows the equipment capacity planning results of the ideal solution. According to the planning results in Table 4, it can be seen that the planning capacity of WT, PV, EB, HST, and HFC is relatively high. On the one hand, this is because WT and PV are the primary electricity sources in the IES to ensure the balance between supply and demand, except for the external electricity grid. On the other hand, EB, HST, and HFC are critical energy conversion equipment to optimize the energy flow and utilization of the system, which can smooth the fluctuation of WT and PV generation well and promote real-time matching of energy supply and load demand.

6.2. Resource Planning Results Under Different Planning-Operating Cases

To precisely compare and analyze the effectiveness of the proposed model in optimizing resource planning and the typical daily multi-energy supply capacity of IES, four planning cases are set as follows:

Case 1: Only consider the planning of conventional generation resources (WT, PV, EH, EB, EC), taking the system planning economy as the objective.

Case 2: Consider the hydrogen utilization in the system operation and the planning of HFC and HST, only taking the system planning economy as an objective.

Case 3: Introduce the source-side ORC waste thermal generation equipment and load-side multi-energy load DR, constituting the bilateral response of supply and demand, to participate in the system operation, only taking the system planning economy as the objective.

Case 4: The proposed multi-objective bi-level planning method.

Table 5. IES resource planning results under different cases.

Resources	Case1	Case2	Case3	Case4
WT/kW	3302.5	3255.8	3037.9	3014.4
PV/kW	1796.1	1515.1	1072.0	1065.4
EB/kW	1393.6	1644.0	1704.8	1784.6
HST/kg	--	200.79	200.05	198.29
HFC/kW	--	607.9	769.98	773.45
ORC/kW	--	--	295.8	301.49
CEV/kW	1270.5	1270.5	1089.9	1079.9
CHV/kg	31.27	31.27	26.58	26.58
EH/kW	1308.8	690.8	517.87	518.5
EC/kW	475.7	477.2	517.38	517.38
AC/kW	535.6	255.48	0	0

shows the results of different planning cases.

Table 6. IES operation results under different cases.

Objective Value/Unit	Case1	Case2	Case3	Case4
Investment cost/105 CNY	60.35	60.95	55.16	55.1
Operating cost/105 CNY	87.5	52.32	49.56	49.8
Total cost/105 CNY	147.85	113.27	104.72	105.08
RE consumption rate/%	82.46	99.75	99.21	99.81
Carbon emission/104 kg	29.78	14.61	6.811	2.450
Energy purchased cost/105 CNY	13.92	0.43	0.126	0.123
EV, HV charging cost/105 CNY	47.81	47.81	45	45
DR compensation cost/105 CNY	--	--	0.576	0.827
Environmental cost/105 CNY	4.81	3.78	3.76	3.75
RE curtailment cost/105 CNY	20.95	0.289	0.096	0.095

shows the system investment and operation cost results in different cases.

6.3. The Effect of Hydrogen Energy Utilization and the Bilateral Response of Supply and Demand on the System Operating Results

To further clarify the effectiveness of hydrogen utilization and the bilateral response of supply and demand in the IES, the typical daily operation results of Case 1 and Case 4 in winter are taken as examples for analysis. Figure 8 shows the supply and demand balance of electricity, thermal, and hydrogen in the operation periods.

As shown in Figure 8, Case 1 has no hydrogen storage equipment, and the IES has to curtail RE generation during WT and PV high generation. Meanwhile, during the peak period of EV and HV charging load, from 16:00 to 20:00, there is a shortage of electricity supply. Consequently, the system must purchase electricity from the external grid to meet the balance of power supply and demand during this period of high electricity prices. In Case 4, HST and HFC are introduced to realize the inter-time transfer of electricity and hydrogen. When the generation of the source-side unit is much higher than the load demand, such as during the period from 22:00 to 3:00, the EB can operate at high power to absorb the surplus RE generation. Thereby, the hydrogen demand of HV and HFC is met, and the excess hydrogen is input into the HST. Therefore, in periods of high hydrogen load, the system can quickly meet the HV demand without consuming massive electricity to produce hydrogen in real time.

Given the thermal load of IES in each period, compared with Case 1, the thermal load determined by the indoor temperature in Case 4 is equivalent to an adjustable interval, which HFC and EH jointly supply. During the peak periods of energy consumption, such as 16:00–18:00, the indoor temperature is reduced within the allowable range, alleviating the energy consumption demand and energy supply pressure of HFC and EH.

6.4. The Effect of Different Numbers of Vehicles on the System Planning Results

With the continuous advancement of clean and low-carbon energy structures, electrified transportation has become an essential tool to promote the realization of the “double carbon goal.” To verify the influence of EV and HV on system resource planning, different numbers of vehicles are set to participate in IES planning and operation. The planning results are shown in

Table 7. IES planning results under different numbers of vehicles.

Vehicles/Number			Total Cost/ 105 CNY	RE Consumption Rate/%	Carbon Emission/104 kg
EV	HV	FV
600	400	1800	69.40	99.82	4.67
800	600	1400	89.73	99.84	4.50
1000	800	1000	105.04	99.81	2.45
1200	1000	600	105.06	99.81	2.22

and Figure 9.

As shown in Figure 9 and Table 7, with the increase in the number of EV and HV, the electricity and hydrogen demand on the load side increases continuously. The system needs to plan more resource capacity to ensure the balance between multi-energy supply and demand. Thus, the total cost and the planning capacity of each resource increase continuously. On the one hand, in the face of increasing load demand, WT planning capacity continued to increase from 2345.3 kW to 3326.3 kW, but PV planning capacity changed little. This is because there is a significant difference between the EV and HV charging load peak period and the high PV generation period. Therefore, the system prioritizes investing in WT. On the other hand, in response to the growing demand for EV and HV charging, the planning capacity of CEV and CHV has increased from 625.5 kW and 11.25 kg to 1159.5 kW and 29.69 kg. The energy supply services for vehicles are enhanced. In addition, the capacity of EB, HST, and HFC shows an increasing trend, which offers sufficient flexible adjustment space for RE consumption about the mismatch between RE generation and multi-loads. The flexibility and reliability of the energy consumption are improved.

6.5. The Effect of Vehicle Numbers on the IES Planning

With the continuous improvement of various types of technologies, the resource investment cost in different years will change. It is necessary to analyze the effect of varying resource investment costs on system planning. Figure 10 shows the sensitivity analysis of the resource investment cost.

As shown in Figure 10, as the unit investment cost is reduced from 100% to 80%, the system investment and operation costs are reduced by 19.85% and 0.18%, respectively. The carbon emission is reduced by 28.29%. The reduction in unit investment cost is more conducive to the system energy saving and emission reduction. The main reason is that as the resource investment cost decreases, the system becomes more inclined to invest and plan resources to ensure a self-balanced supply and demand of multi-energy, reducing its dependence on the external power grid. Consequently, the carbon emission of IES also decreases.

6.6. The Effect of Different Weight Coefficients of the Objective Function on System Planning Results

To analyze the effect of the weight coefficients of the objective function, different weights of the objective function are set. The objective function values and resource planning results of the IES under different weights were compared and analyzed, as shown in

Table 8. Planning results under different weight coefficients of the objective functions.

Weight Coefficient			Total Cost/ 105 CNY	RE Consumption Rate/%	Carbon Emission/104 kg
${\mathit{W}}^1$	${\mathit{W}}^2$	${\mathit{W}}^3$
1	0	0	104.72	99.21	6.811
1/3	1/3	1/3	105.08	99.81	2.450
0	1	0	105.64	99.95	1.06
0	0	1	105.58	99.90	0.88

.

As shown in Table 8, since both the annualized total cost and carbon emissions aim to be minimized, as the weight of the objective function continuously decreases, both the total cost and carbon emissions show a decreasing trend, decreasing from 105.58 × 10⁵ CNY to 104.72 × 10⁵ CNY and from 6.811 × 10⁴ kg to 0.88 × 10⁴ kg, respectively. On the contrary, the RE consumption rate aims to be maximized. As the weight continuously increases, the consumption rate keeps rising, from 99.21% to 99.95%. Therefore, during the actual construction stage of the IES, it is necessary to reasonably set the proportion of different objective function weights and reasonably guide the planning and construction of system resources.

7. Conclusions

This paper proposes an IES multi-objective bi-level planning method considering hydrogen utilization and the bilateral response of supply and demand. This method introduced the ORC at the source side and the multi-energy load DR at the load side to optimize the energy flow and load curve of IES. Additionally, hydrogen utilization was considered to realize the time shift of RE. Based on this, a two-layer resource planning model combining IES planning and operation with multi-energy flow was established to optimize the comprehensive cost, RE consumption rate, and carbon emissions. Through comparative analysis of numerical examples, the following conclusions are drawn:

(1) According to the multi-objective solution set, the system economy, RE consumption rate, and carbon emissions dominate each other, and the multiple objective functions cannot be achieved simultaneously. The economy is inversely correlated with the RE consumption rate and carbon emissions. IES has higher RE consumption and lower carbon emissions, and the system needs to sacrifice economic costs.

(2) The proposed planning method combines hydrogen utilization and the bilateral response of source and demand, and can effectively reduce the dependence on the external grid and promote RE consumption of IES. After introducing hydrogen utilization and the bilateral response into the system, the investment and operating costs were reduced, the carbon emission was reduced by 91.77%, and the RE consumption increased by 21.04%.

It should be noted that the resource planning model proposed in this paper can effectively account for the IES planning-operation economy, RE consumption, and carbon emissions. However, it is simplistic to consider the uncertainty of RE and multi-energy loads during the operation stage of this paper. Therefore, in future studies, the author will focus on the uncertainty of RE and multi-energy loads in detail, using stochastic programming and robust programming to address it, and analyze its impact on system resource planning.