• Semi-automatic estimations of criteria weights via multivariate statistical methods. • MAR feasibility depends on intrinsic suitability, water demand and availability. • Multiple realizations of criteria weights are used to conduct variability analysis • The results were in good agreement with those of previous studies at the demo site. • This approach is directly implemented in software for broad usage by practitioners. Managed aquifer recharge (MAR) provides a nature-based solution to water scarcity issues, contributing to the design of effective water management policies. The novelty of this study lies in the integration of multivariate statistical methods within the framework of multicriteria decision analysis to provide nonsubjective, semi-automatic estimations of MAR feasibility based on three thematic layers, particularly intrinsic (hydrogeological, topographical, meteorological) features, water availability and demand for MAR. The concept of MAR typology is used to define the MAR problem and select a set of criteria for each thematic layer that are relevant for the Sado River Basin (southern Portugal). The hierarchical clustering algorithm (HCA) is used to partition the study area into distinct subregions based on selected criteria. For each subregion and thematic layer, a set of weight coefficients for the criteria is generated via products of the cumulative variance of the principal scores with the corresponding eigenvector matrix, which are then used to generate multiple realizations of the thematic maps. The results reveal that the majority of the study area (60%) has mean feasibility scores less than 0.54, whereas the remaining 40% (highest values) vary within a smaller interval (0.54–0.67). The present results are consistent with the previous study, with both identifying the same high-suitability regions. Additionally, it allows exploration of the variability in criteria weights across multiple realizations, providing a preliminary indication of how this variability can influence the results. This approach can easily be implemented in software and ultimately automated, supporting the unsupervised streamlining of the decision-making process.

The symmetrical phase-shift modulation of the three-to-two-level neutral-point-clamped dual-active-bridge (3/2LNPC-DAB) DC-DC converter has more controllable control variables than the conventional two-level DAB converter. The increase in control variables leads to an increase in the number of operating modes and inequality constraints, significantly complicating the modulation optimization process. To enhance the efficiency of the converter, this paper proposes an optimized quadruple phase-shift (OQPS) modulation considering both conduction losses and switching losses. Five effective operating modes that have the potential to achieve zero-voltage-switching (ZVS) for all switching devices are screened based on the ZVS inequality constraints. The inductor peak current is used as an optimization objective to minimize the conduction losses while satisfying the ZVS inequality constraints. The boundary of the feasible region satisfied by the optimal solution is derived by numerical optimization. Based on this, an improved Lagrange function and Karush-Kuhn-Tucker (KKT) conditions are proposed to solve the analytical solution of OQPS modulation. Finally, the proposed OQPS modulation is applied to a 1.6kW prototype to verify the improvement of efficiency. The designed prototype can achieve 96.9% peak efficiency at 300V input and 150V output voltages. Compared with previous modulations, the proposed OQPS modulation can achieve the highest efficiency in the full power range.

Safe multi-agent reinforcement learning based on adversarial strategy and control barrier function feasible set

In this paper, the problem of sub-stabilisation for constrained strongly unidirectionally connected fully actuated systems (UC-FASs) is investigated. The sub-stabilising control law is constructed via a step-by-step design procedure. The resulting closed-loop system consists of linear subsystems with desired eigenstructures. By analysing the response of this closed-loop system, the feasibility set and the corresponding region of exponential attraction (RoEA) are derived under state and input constraints. The effectiveness of the proposed control scheme is demonstrated through a complicated numerical example and simulations of trajectory tracking for a wheeled mobile robot (WMR).

Autonomous laboratories hold great promise for accelerating materials discovery but often inherit hidden limits because experimental boundaries have been predefined by human intuition or literature precedents. Such a priori constraints risk excluding feasible regions, particularly since synthesizable conditions can shift with hardware or environmental factors. We present SPACESHIP, an AI framework integrated with automated experimental hardware for adaptive exploration of chemical spaces free from literature- or expert-derived feasibility constraints. Through an AI-based prediction, robotic synthesis, real-time characterization, and model update, SPACESHIP combines probabilistic models with an Autopilot acquisition strategy that dynamically switches between models to refine synthesizable regions using both successful and failed experiments. Applied to gold nanoparticle (NP) and nanorod (NR) synthesis, this AI-robotics system achieved 90% accuracy in only 23 experiments, compared with 512 required for the ground truth. It uncovered distinct growth regimes across optical property classes and expanded synthesizable regions by factors of 8 (NPs) and 4 (NRs) beyond literature maps, adapting to hardware-specific conditions rather than relying on fixed, external constraints. By merging machine learning with autonomous experimentation, SPACESHIP addresses the long-standing reproducibility gap in science by diagnosing and adapting to system-specific synthesizable boundaries that shift across laboratories and environments, rather than assuming one universal map.

Transitional weather poses a significant challenge to the secure operation of remote distribution networks, as both high penetration of renewable energy and scarce scheduling resources are prevalent in such networks. In this context, this study proposes a two-stage day-ahead and intraday scheduling method for remote distribution networks that considers the impact of transitional weather and the feasible operation region (FOR) of grid-forming energy storage systems (ESSs) and renewable energy. First, both inertia support and reactive power support capabilities for grid-forming (GFM) energy storage systems are considered in remote distribution networks, where the FOR of energy storage systems is modeled. Second, a two-stage optimal scheduling model is constructed by integrating the complementary use of wind turbines (WTs), photovoltaics (PVs), and energy storage systems. In particular, the FOR of renewable energy is included to ensure the safe and reliable performance of the scheduling scheme of remote distribution networks. Both historical wind and solar meteorological data are then fitted and modeled using the least squares method. Based on the impact of transitional weather and the probability distribution of wind and solar outputs, operational scenarios of renewable energy are developed. Finally, the effectiveness of the proposed scheduling method is validated through simulation analysis based on the modified IEEE 33-bus system under different cases.

Abstract Neural mass models (NMMs) are often used to help understand the circuitry that underpins observed brain dynamics in basic and clinical research. A key step is to fuse models with data so that model parameter values can be inferred for a given data set—a process called model fitting or model calibration. This can shed light on putative physiological mechanisms underlying the observed signals. Calibration is notoriously challenging in biology since models are often non-identifiable, high-dimensional, and nonlinear. Established methods such as dynamic causal modelling (DCM) circumvent some of these issues, for example, by incorporating prior information and employing fast local search methods in the space of feasible parameter values (“parameter space”). However, it is pertinent to better understand the potential limitations of these methods so that we can increase our confidence in the use of models to interpret brain activity, and to develop new approaches as required. Here we use tools from dynamical systems theory to illustrate some of the complexities of model calibration in an archetypal NMM. We use this information to motivate the use of calibration methods that work across large regions of parameter space, rather than focusing on informative priors or localised search methods. We subsequently evaluate the performance of approximate Bayesian computation (ABC) and evolutionary search metaheuristics (ESMs) for mapping feasible sets of parameters for which an NMM can recreate electroencephalographic recordings during an eyes-closed resting state. Our results demonstrate the superiority of ESMs in terms of computational efficiency and accuracy. Furthermore, we elucidate potential reasons why ESMs are able to perform better than ABC, i.e. that they are less susceptible to biases induced by the complexity of underlying cost landscapes. These results highlight the importance of incorporating ESMs in future efforts to model brain dynamics.

As the global response to climate change and energy crises accelerates, the large-scale integration of heterogeneous distributed energy resources (DERs) is rapidly transforming traditional passive distribution networks into active distribution networks. However, the massive quantity and high stochasticity of these underlying devices trigger a severe “curse of dimensionality,” creating significant computational and communication bottlenecks for coordinated system dispatch. To overcome these challenges, the “clustering followed by equivalence” aggregation modeling paradigm has emerged as a critical technical pathway. This paper reviews the state-of-the-art clustering and aggregation methodologies for distribution networks with high DER penetration. The review begins by synthesizing multi-dimensional feature extraction techniques and cutting-edge clustering algorithms that establish the foundation for dimensionality reduction. It then delves into refined aggregation models tailored to heterogeneous resources, including dynamic data-driven equivalence for renewable generation, Minkowski sum-based boundary approximations for energy storage, and thermodynamic alongside Markov chain mapping methods for flexible loads. Building upon these models, the paper comprehensively discusses the practical applications of generalized aggregators, such as microgrids and virtual power plants, in feasible region error evaluation, coordinated network control, multi-agent market games, and privacy-preserving architectures. Finally, the review outlines future research trajectories, emphasizing hybrid data-model-driven architectures for real-time dispatch, distributionally robust optimization (DRO) for enhancing grid resilience and self-healing, and decentralized trading ecosystems to ensure equitable system-level surplus allocation. This review aims to provide a systematic theoretical reference for the coordinated management and aggregated trading of flexibility resources in novel power systems.

Alternating Current Optimal Power Flow (ACOPF) model is nonconvex, generally has multiple local optimal solutions (LOSs) distributed across multiple disconnected feasible components of its feasible region. In this paper, a novel, theory based dynamic method, named QGS-TT, is proposed to compute multiple feasible solutions located in different disconnected feasible components. The proposed method incorporates a dynamic transformation that converts the feasible region of ACOPF problem into the union of regular stable equilibrium manifolds (RSEMs) through the construction of a quotient gradient system (QGS). Furthermore, the TRansformation Under StabiliTy-reTraining Equilibria CHaracterization (Trust-Tech, TT) methodology is extended to systematically explore the solution space in a tier-by-tier manner. Starting from an arbitrary initial point, the QGS-TT method employs a QGS-based numerical approach to detect the existence of, and then compute, other feasible components, a process supported by stability boundary theory (properties). Five numerical studies are conducted on the IEEE 2-, 9-, 39-, 118-, and 500-bus systems, each of which exhibits multiple OPF solutions with significantly different objective function values. Regarding the computational performance, it will be shown that it is much faster than an existing method.

The restoration efficiency can be significantly improved through the coordination of transmission and distribution systems. However, this process is hindered by challenges related to information aggregation, model complexity, and the uncertainty introduced by the penetration of renewable generation (RG). For this purpose, this paper proposes a novel distributed load restoration model for integrated transmission and distribution systems (ITDS) using the robust model projection method (RMPM). To achieve the distributed approach, the non convex restoration model for active distribution systems (ADSs) is first reformulated as a second-order cone programming (SOCP) problem using a sequential SOCP algorithm (SSA). The resulting high-dimensional SOCP model for ADSs is then projected into a low-dimensional space via a vertex-searching method (VSM). Next, the uncertainty of distributed RG in ADSs is considered, and the robust feasible region for the ADS restoration model is determined using the column and constraint generation (CCG) algorithm. This robust, convex feasible region is then integrated into the transmission system (TS) restoration model, which is formulated as a two-stage, three-level robust model considering the RG uncertainty. By adopting this restoration scheme, optimal coordination between ADSs and TSs is achieved with limited information exchange, thereby alleviating the communication burden. Moreover, the model's solvability can be ensured due to its low-dimensional, convex structure. Finally, the effectiveness of the proposed method, as well as its superiority over existing approaches, is validated through numerical experiments.

This paper proposes a real-time optimisation (RTO) framework that uses AI-based surrogate models integrated with the advanced regulatory control (ARC). The main contribution is a method that employs a vertically decomposed control architecture, in which the RTO is formulated using steady-state surrogate models for both objectives and constraints. The RTO layer is responsible for guiding advanced regulatory control (ARC) to the optimal point subject to constraints, and the ARC layer coordinates the PID controllers during dynamic behaviour. The case study is an Electric Submersible Pump (ESP) system controlled with an ARC scheme. Results show effective constraint handling, operation within feasible regions, and adaptability to shifts between production and efficiency maximisation—antagonistic goals in ESP operation. The method is computationally efficient, resulting in a 99% reduction in processing time compared to differential equation methods. The RTO also re-optimised operating conditions following unmeasured disturbances. From a practical standpoint, the results indicate that real-time implementation can be more efficient and requires less computational effort.

This paper investigates the internal secrecy and external covertness of a mixed-trust autonomous aerial vehicle (AAV) communication system assisted by rate-splitting multiple access (RSMA). In this setting, a semi-trusted user with partial decoding capability poses an internal eavesdropping threat, while multiple distributed wardens attempt to detect the transmission from the AAV to the semi-trusted user, creating an external covertness challenge. To characterize these security aspects, a unified analytical framework is developed. First, the internal eavesdropping capability of the semi-trusted user is quantified by deriving a closed-form expression for its eavesdropping success probability. Based on the outcome of the eavesdropping attempt, tractable expressions for the secrecy outage probability of the legitimate user are obtained. Furthermore, the external covertness performance is analyzed by deriving closed-form false alarm probability, missed detection probability, and detection error probability (DEP) for an individual warden, together with the optimal detection threshold and the corresponding minimum DEP. The cooperative global detection performance with multiple wardens is further characterized under conservative fusion rules. Extensive Monte Carlo simulations validate the analytical results and, through a joint evaluation of secrecy, reliability, and covertness metrics, illustrate the feasible operating regions enabled by RSMA power allocation in comparison with a NOMA baseline. The results provide a comprehensive theoretical basis for the design of secure and covert AAV communication strategies in mixed-trust environments.

Autonomous characterisation of fouling-free feasible operating regions in continuous flow precipitation of ketoprofen nanoparticles

Sepsis is a time-sensitive cause of preventable death worldwide, with disproportionate mortality in low-resource settings (LRS). Many recommendations in international sepsis guidance presume resources unavailable in many facilities and communities. We sought to develop a practical framework that helps health systems embed feasible sepsis actions within broader emergency and essential critical care systems, while highlighting where evidence is limited and where local learning systems are needed. A targeted scoping review of peer-reviewed and grey literature on sepsis epidemiology, emergency care systems, essential emergency and critical care, implementation strategies, and quality improvement (QI) in LRS; and key guideline and policy documents relevant to sepsis and emergency care. We prioritized publications and guidance relevant to LRS, including observational studies, pragmatic implementation reports, consensus statements, and policies addressing emergency care organization, workforce, supply chains, diagnostics, and QI. Task force members abstracted actionable strategies, implementation barriers/enablers, and feasibility considerations across the care continuum (community, transport/prehospital, facility-based acute care, and referral). We also identified domains where guideline certainty is low or indirect for LRS. A Society of Critical Care Medicine-convened multidisciplinary task force iteratively developed the "Sepsis Frame of Survival" using a structured process that included 1) scoping evidence review, 2) a Delphi-style prioritization of candidate framework elements by importance and feasibility, and 3) a structured consensus meeting ("Utstein-style" conference format) to finalize the model and its priority actions. We produced a concise implementation roadmap and a feasible measurement set aligned with resource constraints. The Sepsis Frame of Survival is a pragmatic model to organize sepsis improvement as part of emergency and essential critical care strengthening. It emphasizes high-impact actions that can be implemented with limited resources (triage and early recognition, timely antimicrobials, oxygen and basic supportive care, cautious fluid resuscitation with reassessment, source control and referral, diagnostics/microbiology where feasible, and QI). The framework explicitly distinguishes near-term, feasible changes from longer-term system investments and highlights the need for locally generated evidence to guide quality indicators and resuscitation strategies in LRS.

Chaos evolution optimization with feasible region analysis for stability enhancement in cascaded converters

Enhancing grid balancing services from electric vehicle aggregators under uncertainty: A probability-guaranteed feasible region approach

Hyperglycaemia is highly prevalent, affecting approximately one-third of critically ill patients receiving enteral nutrition (EN). Persistent hyperglycaemia is associated with an elevated risk of infection, increased mortality and prolonged length of stay, negatively impacting patient prognosis. Therefore, implementing an evidence-based hyperglycaemia management strategy is crucial to improve patient outcomes. This study aimed to reduce the incidence of hyperglycaemia during EN in critically ill patients through evidence-based practice. A non-concurrent control study was conducted using pre- and post-implementation data. The baseline audit group comprised intensive care units patients receiving EN from January to June 2024, compared with a follow-up audit group receiving an evidence-based protocol from July to December 2024. Following evidence-based practice implementation, the adherence rate for 12 of the 25 review indicators exceeded 60%, with 15 indicators demonstrating a significant improvement compared with pre-intervention levels (p < 0.05). Furthermore, the incidence of hyperglycaemia significantly decreased from 51.73% to 35.34% (p = 0.001), whereas the target blood glucose achievement rate significantly increased from 25.76% to 37.96% (p < 0.001). Concurrently, significant improvements were observed in both mean blood glucose levels (9.20 [7.50, 11.50] mmol/L) and mean fasting blood glucose (8.20 [6.80, 10.35] mmol/L) (p < 0.05). Evidence-based practice for hyperglycaemia management during EN in critically ill patients significantly improves medical staff compliance with blood glucose management protocols, leading to reduced hyperglycaemia incidence, improved target blood glucose achievement, optimized glycaemic control and shortened hospital stays. The significant reduction in hyperglycaemia rates and increased provider adherence to the evidence-based intervention suggest that evidence-based hyperglycaemia management practices can be effectively translated to other clinical settings and provide a feasible set of strategies for implementing similar programmes in resource-constrained settings.

Trajectory optimization for UAV-based medical delivery with temporal logic constraints and convex feasible set collision avoidance

Heterogeneous multi-receiver wireless power transfer (WPT) systems exhibit strong electromagnetic coupling and non-uniform load requirements, making it difficult to determine an operating point that is simultaneously feasible for all receivers and efficient at the system level. This study presents an integrated electromagnetic-circuit digital twin modelling approach that enables constraint-consistent optimal operation in single-transmitter, multi-receiver WPT systems with mixed compensation topologies. A design-stage digital twin is constructed by tightly coupling three-dimensional electromagnetic-field simulation with equivalent circuit modelling. Self- and mutual-inductance parameters are extracted from the physical coil configuration and mapped into the circuit model, enabling physics-consistent evaluation of receiver delivered power and total transmission efficiency under interacting receiver conditions. The operating point is determined by enforcing receiver-specific rated power requirements as feasibility constraints while refining system efficiency within the feasible region. The approach is demonstrated on a 1T-3R WPT case study comprising two series-compensated receivers and one parallel-compensated receiver with non-uniform rated powers. The results show that the proposed digital twin modelling identifies an operating condition that satisfies all receiver power constraints while improving overall efficiency. A one-dimensional parameter sweep around the obtained operating point further confirms local robustness and sensitivity characteristics within the simulation-based design-stage framework, supporting the validity of the constraint-consistent optimum for heterogeneous multi-receiver configurations.

A rapid accumulation of plastic waste has created an urgent need for efficient and sustainable recycling technologies. Among various approaches, pyrolysis stands out as promising method of thermochemical recycling of plastic waste; however, the process needs optimization and further research to make it more energy-efficient and sustainable. The conventional approaches for optimization focus on the enhancement of yield, only overlooking efficiency and system-level sustainability. In this study, a machine learning-enabled surrogate-assisted multi-objective artificial intelligence (AI) optimization framework is developed for plastic pyrolysis to maximize product recovery and minimize energy consumption. The model integrates energy return on investment (EROI) and higher heating value (HHV) into process design. A curated dataset of 312 experimental cases covering polyolefins, PET, nylon, and mixed plastics was used to train multiple machine learning algorithms, such as polynomial regression, Gaussian process regression, and Random Forest models. The Random Forest algorithm demonstrated superior predictive robustness across oil yield, HHV, char formation, and EROI. Pareto front analysis using NSGA-II revealed that moderate reaction severities (400–450 °C, 40–70 min) maximize net energy performance while minimizing solid residues. The conditional variational autoencoder as a GenAI model was incorporated to work as a generative proposal engine, which enhances the exploration of chemically feasible operating regions under uncertainty-aware active learning. The integration of techno-economic and life-cycle assessment demonstrates that energy-positive configurations outperform high-yield scenarios, achieving IRR > 15%, energy intensity < 10 MJ kg−1, and CO2 reductions up to 47% relative to incineration. The proposed framework establishes a data-driven methodology for aligning polymer pyrolysis optimization with circular economy and energy sustainability objectives.