Search Results (536)

In recent years, inverse reinforcement learning algorithms have garnered substantial attention and demonstrated remarkable success across various control domains, including autonomous driving, intelligent gaming, robotic manipulation, and automated industrial systems. Nevertheless, existing methodologies face two persistent challenges: (1) finite or non-optimal expert demonstration and (2) ambiguity in which different reward functions lead to same expert strategies. To improve and enhance the expert demonstration data and to eliminate the ambiguity caused by the symmetry of rewards, there has been a growing interest in research on developing inverse reinforcement learning based on the maximum entropy method. The unique advantage of these algorithms lies in learning rewards from expert presentations by maximizing policy entropy, matching expert expectations, and then optimizing the policy. This paper first provides a comprehensive review of the historical development of maximum entropy-based inverse reinforcement learning (ME-IRL) methodologies. Subsequently, it systematically presents the benchmark experiments and recent application breakthroughs achieved through ME-IRL. The concluding section analyzes the persistent technical challenges, proposes promising solutions, and outlines the emerging research frontiers in this rapidly evolving field. Full article

Ducted fan flying robots with robotic arms can perform physical interaction tasks in complex environments such as indoors. However, the coupling effects between the aerial platform, the robotic arm, and physical environment pose significant challenges for the robot to accurately approach and stably contact the target. To address this problem, we propose a unified control framework for a ducted fan flying robot that encompasses both flight planning and physical interaction. This contribution mainly includes the following: (1) A nonlinear model predictive control (NMPC)-based trajectory optimization controller is proposed, which achieves accurate and smooth tracking of the robot’s end effector by considering the coupling of redundant states and various motion and performance constraints, while avoiding potential singularities and dangers. (2) On this basis, an easy-to-practice hierarchical control framework is proposed, achieving stable and compliant contact of the end effector without controller switching between the flight and interaction processes. The results of experimental tests show that the proposed method exhibits accurate position tracking of the end effector without overshoot, while the maximum fluctuation is reduced by up to 75.5% without wind and 71.0% with wind compared to the closed-loop inverse kinematics (CLIK) method, and it can also ensure continuous stable contact of the end effector with the vertical wall target. Full article

Scopula subpunctaria (Herrich-Schaeffer), is a significant insect pest affecting tea plantations in China; however, its thermal developmental characteristics remain inadequately understood. This study examined the immature developmental stages of S. subpunctaria under eight constant temperature regimes (13, 16, 19, 22, 25, 28, 31, and 33 °C) in controlled laboratory conditions. Results indicated an inverse relationship between temperature and the total duration of the immature stages (egg to pupa), with developmental time decreasing from 105.8 days at 13 °C to 29.3 days at 31 °C. Specifically, the developmental durations for eggs, larvae, and pupae ranged from 5.4 to 20.3 days, 15.4 to 52.3 days, and 8.1 to 33.3 days, respectively, in 13 °C to 31 °C temperature range. Using an ordinary linear model, the estimated lower developmental threshold temperatures were 8.61 °C for eggs, 8.40 °C for larvae, 9.39 °C for pupae, and 8.85 °C for the total immature stage, with corresponding thermal constants of 114.94, 302.11, 149.93, and 558.99 degree-days (DD), respectively. Comparative analysis of eleven nonlinear models revealed substantial variation in estimates of lower and upper temperature thresholds, while estimates of optimal temperatures showed minor differences. Based on statistical criteria and biological relevance, the Briere-2 model was selected to characterize egg development, the Lactin-1 model for larval development, and the Briere-1 model for pupal and total immature stages. Stage transition models for eggs, larvae, pupae, and the total immature period were constructed using a two-parameter Weibull function integrated with the respective nonlinear models. This study provides foundational insights into the thermal developmental characteristics of S. subpunctaria and offers predictive tools for forecasting stage-specific emergence in tea plantations. Full article

Improving the isotropic magnetic properties of FeSi electrical steels has traditionally focused on enhancing their crystallographic texture and microstructural morphology. Strengthening the cube texture within a ferritic matrix of optimal grain size is known to reduce core losses and increase magnetic induction. However, conventional cold rolling followed by annealing remains insufficient to optimise the magnetic performance of thin FeSi strips fully. This study explores an alternative approach based on grain boundary migration driven by temperature gradients combined with deformation gradients, either across the sheet thickness or between neighbouring grains, in thin, weakly deformed non-oriented (NO) electrical steel sheets. The concept relies on deformation-induced grain growth supported by rapid heat transport to promote the preferential formation of coarse grains with favourable orientations. Experimental material consisted of vacuum-degassed FeSi steel with low silicon content. Controlled deformation was introduced by temper rolling at room temperature with 2–40% thickness reductions, followed by rapid recrystallisation annealing at 950 °C. Microstructure, texture, and residual strain distributions were analysed using inverse pole figure (IPF) maps, kernel average misorientation (KAM) maps, and orientation distribution function (ODF) sections derived from electron backscattered diffraction (EBSD) data. This combined thermomechanical treatment produced coarse-grained microstructures with an enhanced cube texture component, reducing coercivity from 162 A/m to 65 A/m. These results demonstrate that temper rolling combined with dynamic annealing can surpass the limitations of conventional processing routes for NO FeSi steels. Full article

Oocytes from women of advanced reproductive age exhibit diminished developmental potential, but the underlying mechanisms remain incompletely defined. Oocyte maturation depends on translational control of maternal mRNA synthesized during growth. We performed a computational analysis on human oocytes from women <30 versus ≥40 years and observed that mRNA GC content correlates negatively with half-life in oocytes from young (<30 yr) but positively with oocytes from aged (>40 yr) women. In young oocytes, longer mRNA half-life is associated with lower protein abundance, whereas in aged oocytes GC content correlates positively with protein abundance. During the GV-to-MII transition, codon composition stratifies stability: codons that support rapid translation (optimal) stabilize mRNA, while slow-translating codons (non-optimal) promote decay. With reproductive aging, GC-containing codons become more optimal and align with increased protein abundance. These findings indicate that reproductive aging remodels codon-optimality-linked, translation-coupled mRNA decay, stabilizing a subset of GC-rich maternal mRNA that may be prone to excess translation during maturation. Our analysis is explicitly within human reproductive aging; it does not revisit cross-species stability rules. Instead, it shows that sequence–stability relations are reprogrammed with age within human oocytes, including an inversion of the GC–stability association during GV-to-MII transition. Disruption of the normal mRNA clearance program in aged oocytes may compromise oocyte competence and alter maternal mRNA dosage, with downstream consequences for early embryonic development. Full article

Background/Objectives: Human papillomavirus (HPV)-attributable cancers are a major public health problem worldwide. However, HPV vaccination rates vary significantly and are often not optimal. This study aimed to assess the effects of multilevel interventions on improving HPV vaccination. Methods: A systematic literature review and a meta-analysis were carried out, taking into account randomized controlled trials. Outcomes of interest were HPV vaccination initiation and completion. A random-effect meta-analysis using the generic inverse variance method was carried out, with a risk ratio (RR) with a 95% confidence interval (CI) as the pooled effect estimate. Results: A literature search identified 15 relevant studies, all conducted in high-income countries. Multilevel interventions significantly improved HPV vaccination coverage and initiation (RR = 1.26, 95% CI 1.16–1.38, p < 0.00001 and RR = 1.14, 95% CI 1.04–1.24, p = 0.004, respectively) compared to usual care. Sensitivity analyses showed that the results remained relatively robust. Subgroup analysis by targeted levels of intervention indicated that multilevel interventions had an effect across all comparisons and outcomes except for HPV vaccination completion for interventions that targeted four levels of influence. Conclusions: Based on evidence from high-income settings, multilevel interventions are effective in improving HPV vaccination rates. Future studies should expand the focus to areas with limited resources too and aim to provide more detailed data, avoid registering outcomes via self-report, and create sustainable strategies that can persist beyond a study’s duration and possibly become part of policies for improving HPV vaccination coverage. Full article

To address the technical challenges in dynamic monitoring of grout diffusion patterns under complex geological conditions, in this study, a distributed parallel grouting monitoring system based on electrical resistivity tomography was developed. The system achieves three-dimensional visualization of grout propagation through hardware architecture innovation and the integration of inversion algorithms. At the hardware level, a cascadable distributed data acquisition terminal was designed, employing a dynamic optimization strategy for electrode combinations. This breakthrough overcomes traditional serial acquisition limitations. Algorithmically, a Bayesian estimation-based geological unit merging inversion model was proposed; it dynamically calculates merging thresholds through the noise posterior probability, achieving an improvement of more than 30% in the inversion boundary resolution compared with traditional least squares methods. Numerical simulations and physical experiments demonstrated that dipole arrays with 0.5 m electrode spacing exhibit optimal sensitivity to variations in grout resistivity, accurately capturing electrical response characteristics during diffusion. In practical roadbed grouting applications, the system yielded a grout diffusion radius showing only a 0.3 m deviation from the core sampling verification results, with three-dimensional imaging clearly depicting the diffusion morphology. This system provides reliable technical support for the precise control and quality assessment of underground engineering grouting processes. Full article

To mitigate thermal cracking in concrete box girders during construction, this study introduces an inversion method for thermal parameters by integrating machine learning with finite element simulation. The research aims to accurately identify key thermal parameters—thermal conductivity k, total hydration heat Q₀, convection coefficient h, and reaction coefficient m—through an efficient and reliable data-driven approach. An orthogonal experimental design was used to construct a representative sample database, and a Bayesian-optimized XGBoost (BO-XGBoost) model was developed to establish a nonlinear mapping between temperature peaks and thermal parameters. Validated against field monitoring data from a prestressed concrete continuous rigid-frame bridge, the method demonstrated high accuracy: the inversiontemperature curves closely matched measured data, with a maximum peak temperature error of only 1.40 °C (relative error 2.5%). Compared to conventional machine learning models (DT, SVR, BP and LSTM), BO-XGBoost showed superior predictive performance and convergence efficiency. The proposed approach provides a scientific basis for real-time temperature control and crack prevention in concrete box girders and is applicable to temperature field analysis in mass concrete structures. Full article

Dark fermentation (DF) has gained increasing interest over the past two decades as a sustainable route for biohydrogen production; however, understanding how reproducible the process can be, both from macro- and microbiological perspectives, remains limited. This study assessed the reproducibility of a parallel continuous DF system using fruit-vegetable waste as a substrate under strictly controlled operational conditions. Three stirred-tank reactors were operated in parallel for 90 days, monitoring key process performance indicators. In addition to baseline operation, different process enhancement strategies were tested, including bioaugmentation, supplementation with nutrients and/or additional fermentable carbohydrates, and modification of key operational parameters such as pH and hydraulic retention time, all widely used in the field to improve DF performance. Microbial community structure was also analyzed to evaluate its reproducibility and potential relationship with process performance and metabolic patterns. Under these conditions, key performance indicators and core microbial features were reproducible to a large extent, yet full consistency across reactors was not achieved. During operation, unforeseen operational issues such as feed line clogging, pH control failures, and mixing interruptions were encountered. Despite these disturbances, the system maintained an average hydrogen productivity of 3.2 NL H₂/L-d, with peak values exceeding 6 NL H₂/L-d under optimal conditions. The dominant microbial core included Bacteroides, Lactobacillus, Veillonella, Enterococcus, Eubacterium, and Clostridium, though their relative abundances varied notably over time and between reactors. An inverse correlation was observed between lactate concentration in the fermentation broth and the amount of hydrogen produced, suggesting it can serve as a precursor for hydrogen. Overall, the findings presented here demonstrate that DF processes can be resilient and broadly reproducible. However, they also emphasize the sensitivity of these processes to operational disturbances and microbial shifts. This underscores the necessity for refined control strategies and further systematic research to translate these insights into stable, high-performance real-world systems. Full article

Within the field of permanent magnet synchronous motor sensorless speed control systems, we present a novel scheme with a Multi-dimensional Taylor Network (MTN)-based nonlinear observer as the core, supplemented by two auxiliary MTN modules to realize closed-loop control: (1) MTN Model Identifier: Provides real-time PMSM nonlinear dynamic feedback for the observer; (2) MTN Adaptive Inverse Controller: Compensates for load disturbances using the observer’s estimated states. The study focuses on optimizing the MTN observer to address key limitations of existing methods (high computational complexity, lack of stability guarantees, and low estimation accuracy). Compared with the neural network observer, this MTN-based scheme stands out due to its straightforward structure and significantly reduced approximately 40% computational complexity. Specifically, the intricate calculations and high resource consumption typically associated with neural network observers are circumvented. Subsequently, by leveraging Lyapunov theory, an adaptive learning rule for the MTN weights is meticulously devised, which seamlessly bridges the theoretical proof of the nonlinear observer’s stability. Simulation results demonstrate that the proposed MTN observer achieves rapid convergence of speed and position estimation errors (with steady-state errors within ±0.5% of the rated speed and ±0.02 rad for rotor position) after a transient period of less than 0.2 s. Even when stator resistance is increased by tenfold to simulate parameter variations, the observer maintains high estimation accuracy, with speed and position errors increasing by no more than 1.2% and 0.05 rad, respectively, showcasing strong robustness. These results collectively confirm the efficacy and practical value of the proposed scheme in PMSM sensorless speed control. Full article

The adaptability of a hexapod robot to complex terrain is highly dependent on its own posture, which directly affects its stability and flexibility. In order to adapt to a change in terrain, it is necessary to adjust posture in real time when walking. At the same time, external factors such as ground state and landing impact will also interfere with posture. Therefore, it is necessary to maintain balance after adjustment. This paper proposes a pose adjustment method utilizing joint angle control. It enhances robot stability, flexibility, and terrain adaptability through torso posture and center of gravity optimization, aiming to maintain balance. The strategy’s effectiveness was validated via Adams–Simulink co-simulation. Optimal position and posture adjustment for the torso was then implemented at the six-legged support stage after each step, employing inverse kinematics and a triangular gait. It is found that without pose adjustment, the direction deviation will accumulate and significantly deviate from the trajectory. The introduction of this adjustment can effectively correct the direction deviation and torso posture angle, increase the stability margin, ensure stable straight-line walking, and significantly reduce joint energy consumption. Crawling experiments with the physical prototype further validate the strategy. It rapidly counters instantaneous attitude fluctuations during leg alternation, maintaining a high stability margin and improving locomotion efficiency. Consequently, the robot achieves enhanced directional stability, overall stability, and energy efficiency when traversing terrain. Full article

In the development of tight gas reservoirs in Chuanzhong BJC Gas Field of the Sichuan Basin, running horizontal casing in ultra-long open-hole section faces challenges. These include large friction prediction errors and high casing buckling risks. These challenges significantly impede both the efficiency and safety of field development. Traditional static segmented friction models fail to accurately predict friction coefficients. The reason is that they cannot track dynamic changes in wellbore inclination, azimuth, and dogleg severity in real time. To address this bottleneck, this study develops a technical system termed AI-based dynamic friction inversion-segmented process optimization. Clustering algorithms are used to divide regions. These regions have low, medium, and high friction characteristics. The simulated annealing algorithm dynamically corrects friction coefficients. Meanwhile, the segmented processes of float collars and drilling fluid density are optimized. Verification was conducted on well J108-2H, which features an open-hole section of 4060.9 m and a horizontal-to-vertical ratio (HD/TVD) of 1.88. Results show that this system significantly reduces the mean absolute percentage error of friction coefficient prediction. It also greatly improves the accuracy of casing running feasibility assessment. As a result, the casing in well J108-2H was run smoothly and efficiently. The research results provide an innovative solution for the safe and efficient development of ultra-long open-hole sections in unconventional gas reservoirs. Full article

Background: Redox-responsive prodrug nanoassemblies (NAs) have been extensively utilized in precise cancer therapy. But there is no research shedding light on the impacts of the π–π stacking interactions on the self-assembly capacity of redox-responsive prodrugs and the in vivo delivery fate of NAs. Methods: Three structurally engineered doxorubicin (DOX) prodrugs (FAD, FBD, and FGD) were developed through α-, β-, and γ-positioned disulfide linkages with π-conjugated Fmoc moieties. The NAs were comprehensively characterized for their self-assembly kinetics, redox-responsive drug release profiles, and physicochemical stability. Biological evaluations included cellular uptake efficiency, in vivo pharmacokinetics, and antitumor efficacy in tumor-bearing mouse models. Results: Systematic characterization revealed that π-conjugated disulfide bond positioning dictates prodrug self-assembly and inversely regulates reductive drug release relative to carbon spacer length. The FBD NAs demonstrated optimal redox-responsive release kinetics while maintaining minimal systemic toxicity, achieving 101.7-fold greater tumor accumulation (AUC) than DiR Sol controls. In 4T1 tumor-bearing models, FBD NAs displayed potent antitumor efficacy, yielding a final mean tumor volume of 518.06 ± 54.76 mm³ that was statistically significantly smaller than all comparator groups (p < 0.001 by ANOVA at a 99% confidence interval). Conclusion: These findings demonstrate that strategic incorporation of redox-sensitive disulfide bonds with different π–π stacking interactions in the prodrug structure effectively optimizes the delivery-release balance of DOX in vivo, ensuring both potent antitumor efficacy and reduced systemic toxicity. Full article

Most existing ultra-deep gas wells are characterized by high temperature, high pressure, and high sulfur content. During development, they face serious challenges such as unclear mechanisms of acid gas-induced blowouts and difficulties in wellbore pressure inversion, posing significant challenges to well control operations. To reveal the reasons behind the tendency of acidic gases to trigger blowouts and to clarify the impact of different concentrations of acidic gases on the flow behavior of annular fluids, this study considers the effects of solubility and phase changes on the physical properties of acidic gases. A method replacing critical parameters with pseudo-critical parameters is used to analyze the variation trends of gas density, solubility, and other properties along the well depth. A mathematical model for the annular flow of acidic gas overflow incorporating solubility phase change effects is established. The model is numerically solved using a four-point difference scheme, exploring the essential characteristics of gas flow in the annulus after overflow, and discussing the distribution patterns of physical properties of acidic gases, as well as dynamic parameters such as wellbore pressure and temperature along the well depth. Numerical simulations show that the physical properties of acidic gases change significantly with well depth: the more acidic gas present in the wellbore, the smaller the deviation factor, and the greater the density and viscosity, with parameter changes exceeding 40% near the pseudo-critical point for binary mixtures with 40% H₂S. Compared to pure methane, mixed fluids containing acidic gas experience more than 20% volume expansion near the wellhead for ternary mixtures with 20% CO₂ and 20% H₂S, and the flow velocity increases by more than 10% for mixtures with ≥30% acidic gas content, leading to a higher risk of a sudden pressure drop during well control. This study clarifies the migration patterns of acidic gas overflow in HPHT (high pressure, high temperature) gas wells, providing valuable guidance for optimizing well control design, improving well control emergency plans, and developing well-killing measures. Full article

To address thermal management challenges in CR400BF high-speed EMU electrical cabinets—stemming from heterogeneous component integration, multi-condition dynamic thermal loads, and topological configuration variations—a dual-metric-driven finite element model calibration method is proposed using ANSYS Workbench. A multi-objective optimization function, constructed via the coefficient of determination ($R^2$) and root mean square error ($RMSE$), integrates gradient descent to inversely solve key parameters, achieving precise global–local model matching. This establishes an equivalent model library of 52 components, enabling rapid development of multi-physical-field coupling models for electrical cabinets via parameterization and modularization. The framework supports temperature field analysis, thermal fault prediction, and optimization design for multi-topology cabinets under diverse operating conditions. Validation via simulations and real-vehicle tests demonstrates an average temperature prediction error $\le 10\%$, verifying reliability. A thermal management optimization scheme is further developed, constructing a full-process technical framework spanning model calibration to control for electrical cabinet thermal design. This advances precision thermal management in rail transit systems, enhancing equipment safety and energy efficiency while providing a scalable engineering solution for high-speed train thermal design. Full article