HSPG: An open-loop testing framework for autonomous driving based on proactive generation of hazardous scenario.

Electricity forward contracts are key instruments for managing price volatility in liberalized power markets, where non-storability and real-time balancing create complex price dynamics. These contracts differ from traditional derivatives as they are defined over delivery periods, leading to overlapping maturities and interdependent forward curves. This structure, combined with low liquidity and sparse data in long-term horizons, poses challenges for accurate forecasting. This work proposes a novel probabilistic forecasting framework for electricity forward curves, addressing multivariate dependencies, seasonality, and data sparsity. The approach involves three steps: (i) forward curve estimation with seasonal adjustment, arbitrage-free constraints, and a non-parametric smoothing error model; (ii) dimensionality reduction, and orthogonalization of elementary errors; (iii) probabilistic forecasts using autoregressive models, bootstrap, and Generalized Autoregressive Score (GAS) models for residuals. The framework supports bidirectional estimation and forecasting. By enhancing forecast accuracy and capturing forward curve dynamics, the method facilitates more informed decision-making for energy market participants. Results confirm the model’s effectiveness in capturing key multivariate structures for portfolio risk management. • A multivariate framework for probabilistic forecasting of electricity for- ward curves. • Smoothing techniques to estimate elementary forward contract series. • Dimensionality reduction to enable efficient scenario generation. • Time-series dynamics that capture seasonality and market uncertainty. • Numerical examples to evaluate forecasting and financial metrics.

To explicitly illustrate the relationship between heliostat field optimization and power generation, a coupled model was established in Simulink. By optimizing the geometric layout of the heliostat field, the solar heat collection efficiency can be significantly improved, thereby increasing the thermal input to the system. The optimized heliostat field design can convert solar energy into thermal energy more efficiently and transfer it to the steam generator through the molten salt loop, thereby driving power generation in the Rankine cycle. In this process, the Rankine cycle is responsible for converting the thermal energy supplied by the molten salt loop into mechanical work and ultimately into electrical power output. At the same time, real meteorological data from a commercial heliostat field were introduced, and annual power generation simulations demonstrated that the integrated modeling of the heliostat field, thermal storage, and power block based on actual meteorological boundary conditions and system parameters can effectively reflect the power generation performance of a commercial tower solar thermal power plant. Meanwhile, research on heliostat field optimization should further evolve from identifying general patterns toward parameter design and overall system performance improvement. For molten-salt tower solar thermal power plants, key design variables such as receiver tower height, receiver dimensions, heliostat dimensions, and heliostat field spacing parameters affect not only the annual average optical efficiency of the heliostat field and the thermal power output of the receiver, but also the annual power generation of the entire plant. By integrating SOLARPILOT 1.5.2 and SAM 2025.4.16, the design variables were systematically analyzed to investigate their effects on the annual average optical efficiency of the heliostat field, the number of heliostats, the receiver output power, and the annual power generation, and the reasonable value ranges of the heliostat field parameters were determined accordingly. The established Rankine cycle power block model was then coupled with the parameter optimization results to carry out a secondary optimization of the initial heliostat field. Through the above study, the aim is to realize a shift from single-objective geometric optimization of the heliostat field to comprehensive optimization oriented toward annual plant power generation performance and scenario adaptability, thereby providing a basis for scheme design and parameter selection of molten-salt tower solar thermal power plants. For external validation, the annual generation predicted for the Delingha 50 MW commercial plant was 142.15 GWh, corresponding to a relative deviation of 2.64% from the published design value of 146 GWh. This indicates that the coupled framework can reasonably capture the integrated response of the heliostat field, thermal storage system, and power block at the plant level. The model is therefore suitable for generation-oriented parameter screening and preliminary design of tower molten-salt CSP plants, while detailed component-level transient design still requires higher-fidelity engineering models.

Abstract Plasmaspheric plume hiss plays a crucial role in shaping Earth's electron radiation belts and influencing magnetosphere–ionosphere energy coupling. However, its generation mechanism remains contested between cyclic‐linear and localized‐nonlinear models. By analyzing over 64,000 high‐resolution plume hiss wave segments from the Van Allen Probes (1 January 2013–31 July 2019), we identify a distinct frequency dependence in their latitudinal distributions of directionality and amplitude. For high‐frequency hiss, bidirectional propagation is sharply confined near the magnetic equator, beyond which poleward‐propagating waves overwhelmingly dominate, and the wave amplitude increases obviously with latitude. These signatures are consistent with a rapid, single‐pass, equatorially confined, nonlinear amplification process. In contrast, low‐frequency hiss exhibits a high prevalence and wide latitudinal extension of bidirectional propagation, with relatively smooth amplitude variations. This pattern supports a generation scenario involving slower growth, potentially linear or nonlinear, that is coupled with wave bounce motion along magnetic field lines.

Dynamic probabilistic risk assessment (PRA) is a computational approach to risk assessment that offers certain advantages over the more conventional PRA methods. Advantages of dynamic PRA include automatic generation of potential accident scenarios and explicit consideration of complex system dynamics; however, dynamic PRA typically requires computationally costly simulations, which have been a significant barrier to its widespread adoption. To improve the computational performance of dynamic PRA, this paper proposes a novel algorithm for sampling-based dynamic PRA, the exploratory nuclear tree sampler (ENTS). ENTS utilizes importance sampling and Monte Carlo tree search to reduce the computational cost and increase the number of unique scenarios generated during a dynamic PRA. To demonstrate the computational improvements offered by ENTS, its performance is assessed using two case studies from the dynamic PRA literature. The computational performance of a comparable Monte Carlo method is assessed and used as a benchmark for comparison. This comparison shows that the ENTS approach requires fewer overall simulations to estimate the system risk and generates more unique accident scenarios than the Monte Carlo approach.

This article presents the development and validation of a digital twin of a galvanic production line and a multi-agent system using the Unity environment as a simulation platform. The primary objective was to create a tool that supports the analysis, testing and optimization of process flows and intralogistics involving autonomous mobile robots. The model replicates the actual structure of the technological line, taking into account the sequence of operations, energy consumption, processing times and transport logic. By applying a modular architecture, input data—such as layout configuration, energy usage and operation sequences—are imported from external configuration files, enabling rapid generation and comparison of various scenarios. A series of simulation experiments was conducted and the results were compared with an idealized mathematical model. The analysis revealed the presence of bottlenecks, shifting throughput limitations, as well as queuing and routing conflicts that cannot be captured by simplified analytical models. Furthermore, it was shown that expanding the process structure increases total energy consumption but improves energy efficiency per produced unit. The developed digital twin enables safe testing of structural and operational changes and provides a foundation for future integration with MES or SCADA systems, as well as the advancement of predictive production process management.

Security is the main attribute when dealing with information exchange. Confidential information theft, data loss, and data manipulation are conceivable results of security events. Different forms of data hiding are Cryptography and Steganography. Cryptography converts information into an unreadable form, and steganography hides the existence of information. The proposed work experiments with Advanced Blowfish Encryption based on an extended round function integrity with Chaotic Image Quantization (ABECIQ) as a security mechanism. ABECIQ aims to introduce a novel security mechanism that combines cryptography and steganography with the key generation scenario using a genetic algorithm. Initially, using a genetic algorithm and real-time clock values, the secret keys are created. The Blowfish algorithm's round function 'F' is modified by adding crossover and mutation functions. The generated ciphertext is embedded in an image using the chaotic-quant technique. The proposed work is analysed using parameters of the Avalanche effect, Entropy values, Execution time, Attack scenario, Correlation coefficient, and Peak Signal-to-Noise Ratio (PSNR) values. The experiments demonstrate that the ABECIQ algorithm achieves PSNR values within the range of 65 to 74 dB while SSIM values are above 0.999, which indicate high imperceptibility. The generated keys also show entropy values which are close to the theoretic maximum of 8 bits per character. In addition, the proposed algorithm shows high throughput thereby indicating improved computational efficiency compared to the existing algorithm. The analysis shows that ABECIQ provides better results than the existing Chaotic, Blowfish Encryption, as well as AES-RDH algorithm. ABECIQ is evaluated with different text files of sizes 4KB and 12KB demonstrating better PSNR, MSE, SSIM, and Correlation Coefficient. In addition, the time complexity for ABECIQ has also been analyzed for embedding process.

Large-scale renewable energy bases are increasingly deployed in arid regions, which offer favorable conditions for wind and PV generation supported by energy storage systems and long-distance transmission lines. However, the planning of such bases is complicated by the high variability of renewable generation, limited flexibility resources, and complex multi-objective trade-offs. To address these issues, this study proposes a capacity planning model for wind-PV-thermal-storage renewable energy bases, minimizing construction and operational costs while accounting for uncertainty and explicitly quantifying the value of flexibility resources. Compared with existing capacity planning models that rely on deterministic formulations or simplified two-stage stochastic representations, the proposed model explicitly embeds intraday operational flexibility and forecast-error costs into life-cycle planning. Operational costs are assessed through sequential production simulations, in which intraday forecast errors are incorporated via deviation costs and flexibility requirements. A hybrid sampling strategy combining Latin hypercube sampling and importance sampling is used for scenario generation, followed by scenario reduction to improve computational efficiency. To solve the optimization model, a nested generalized Benders decomposition framework is developed, decomposing the model into a master problem and multiple production simulation subproblems, which are further divided into mixed-integer and continuous-variable layers to enhance computational tractability and solution accuracy. Case studies demonstrate that the proposed model and algorithm demonstrate the role of flexibility resources, resulting in economically viable and practically implementable capacity under high renewable penetration. By explicitly accounting for intraday forecast deviations, the resulting plans ensure reserve adequacy for over 95% of uncertainty realizations while remaining economically viable and practically implementable. Moreover, the impact of carbon emission penalties on capacity allocation and renewable utilization is quantified, highlighting implications for system design and planning strategies for wind-PV-thermal-storage renewable energy bases.

Adaptive risk inversion with iterative exploration for high-risk AV-VRU interactions.

Bi-level planning of data centers with coupled electricity-heat-computation system using data-driven scenario generation for representing uncertainties

Two-stage scenario generation of hydro-wind-solar complementary system based on improved variational autoencoder and generative adversarial networks model

A new framework for the medium and long-term stochastic scenario generation of wind power output

Scenario generation for testing autonomous escort operations of intelligent tugs based on surrogate modeling and hybrid sampling

Coherent load profile synthesis with conditional diffusion for LV distribution network scenario generation

This paper presents a feed-forward neural network (NN) for hosting capacity analysis (HCA) in real-time using information from a limited number of buses, addressing the challenge of data scarcity and the need for efficient HCA with minimal input data. A limited number of buses in the distribution network is identified, referred to as pilot buses, using an approach that maximizes observability and controllability while minimizing voltage deviations at load buses to ensure network robustness. A large number of loading and electrical generation scenarios are generated using Monte Carlo simulation to train the NN with voltage information as inputs and HC as output. This trained NN can then predict the hosting capacity for distribution networks within seconds with new inputs, offering a highly efficient and practical solution for distribution utilities. The effectiveness of the proposed approach is demonstrated on the IEEE 123 bus distribution network and a rural feeder in New South Wales, Australia. Results show that the model predicts hosting capacity in 0.04 s for the IEEE 123 network and 0.18 s for the real network, with an average error of less than 1%, showcasing its accuracy and reliability. This rapid and precise prediction capability enhances distribution network management and operational efficiency for utilities. • A feed-forward neural network is proposed for quick hosting capacity analysis. • Only voltage information from pilot buses is used to train the neural network. • The method is tested on the IEEE 123 bus system and a rural feeder in Australia. • The proposed method shows promising results and requires fewer inputs.

ABSTRACT To effectively investigate the structural discrepancies and complementary energy characteristics among multiple virtual power plants (VPPs), and to improve the economic efficiency, low‐carbon performance, and operational reliability of the multi‐agent system, this paper proposes a low‐carbon collaborative optimal operation strategy for multiple VPPs based on the asymmetric Nash bargaining solution. First, considering the interests of various VPP operators and users, a cooperative operation framework for multiple VPPs under the green certificate and carbon trading mechanism is constructed using Nash bargaining theory, which facilitates electric–thermal energy transactions among different VPPs. Second, the Latin hypercube scenario generation method and a fast backward reduction technique are employed to handle the uncertainties associated with renewable generation and load power. Third, accounting for the multidimensional contributions of each VPP, an asymmetric Nash bargaining mechanism based on energy contribution rates is designed to allocate the cooperative gains among VPPs. Finally, the alternating direction method of multipliers (ADMM) is adopted to protect the information privacy of each participant. Case study results demonstrate that the proposed strategy can enhance the overall operational benefits and carbon reduction performance of the system, while promoting stronger collaboration and fairness‐oriented benefit allocation among VPPs.

The study explores the possibilities of applying machine learning models to automate the generation and prioritization of software test scenarios within a dynamic continuous integration (CI/CD) environment. The relevance of the work is driven by the need for rapid quality assessment of software systems amidst the exponential growth of critical applications. The scientific novelty of the research lies in the development of a hybrid approach that, for the first time, integrates ensemble Gradient Boosted Decision Tree (GBDT) models for defect prediction with Reinforcement Learning (Q-learning) algorithms for adaptive test prioritization. The prioritization task is formalized as a combinatorial optimization problem with a weight function, and key quality metrics are defined: APFD (Average Percentage of Faults Detected) and its time-aware modification, NAPFD. An original ontological model of factors influencing software quality has been developed, structuring internal (code complexity), external (infrastructure), and human factors (developer qualification). A modified reward function for the reinforcement learning agent is proposed, which balances defect detection, test execution duration, and historical efficiency. Experimental validation was conducted using benchmark datasets from the AEEEM repository, covering real-world defect data from five open-source Eclipse ecosystem projects: JDT Core, Equinox, Lucene, Mylyn, and PDE. Comparative analysis results showed that the integration of GBDT with Q-learning provides an 8.1% improvement in the APFD metric compared to the best baseline method, XGBoost, and a 4.6% improvement over the existing RETECS method. This confirms the high efficiency of the proposed approach for optimizing test runs under limited time budgets. The training time of the hybrid model is 163.2 seconds, which is entirely acceptable for practical implementation in industrial software development pipelines.

Escalating global production and consumption are driving rapid growth in energy demand, increasing pressure on finite natural resources. In response, this study proposes a data-driven framework that integrates deep learning-based electricity demand forecasting with economy-wide input–output material footprint analysis to support long-term energy planning and policymaking. The innovative aspect of this framework is its ability to jointly assess future electricity generation and related material requirements within a single analytical structure. A comparative analysis is conducted for Türkiye, Germany, and Spain, evaluating the material footprint of electricity generation across renewable and fossil-based energy sources under business-as-usual (BAU) and alternative energy development scenarios. The forecasting models demonstrate strong predictive performance, achieving Mean Absolute Percentage Error (MAPE) values of 1.39% for Türkiye, 4.39% for Germany, and 3.90% for Spain, significantly outperforming conventional statistical methods. Scenario-based results indicate that sustainability-oriented pathways (ST and GCA) can reduce material requirements by approximately 20–30% compared to the BAU scenario, particularly for metal-intensive inputs such as iron and refined oil. The findings underscore the importance of integrating material footprint considerations into energy transition strategies and provide practical insights for policymakers seeking to balance energy security with resource sustainability. The study highlights the value of integrated analytical approaches in supporting more resilient and resource-efficient energy systems.

ABSTRACT Nowadays, with the growing penetration of renewable generation, economic dispatch is increasingly important in short‐term power system operation. In this paper, a deep renewable scenario generation model combining Multi‐Scale Decomposition mixer and Wasserstein Generative Adversarial Network with Gradient Penalty is proposed to achieve novel decision‐oriented forecasting, thus realizing effective characterization of renewable temporal dynamics and economic performance. From the perspective of wind and solar generation, the validity of the proposed method is demonstrated on a real‐world dataset with power station at regional level. Experimental results confirm the superiority of model performance through statistical indicators and power system scheduling test, compared with a number of scenario generation and time series forecasting benchmarks.

The foundation of high-quality economic development lies in enterprises' innovation in models, businesses, processes and products, and digital transformation is the only way for enterprise innovation. Then, what drives enterprise digital transformation? And what path does it follow? By using the frequency of digital-related vocabulary in the annual reports of listed companies to measure the degree of digital transformation, this paper adopts the structural equation model to analyze the driving role of R&D in digital transformation and explore the path of digital transformation at the same time. It is found that the fixed asset ratio, digital technology and total asset turnover are the key nodes connecting R&D and the generation of data application scenarios. The transformation path with total asset turnover as the node either has an insignificant effect or shows a negative effect, revealing the problem of the disconnection between digital technology and business in the transformation process.