Search Results (3,056)

Electrically driven compressors are a primary energy consumer in natural gas storage facilities. Formulating an optimal gas injection allocation strategy considering their nonlinear characteristics and time-of-use (TOU) electricity prices is crucial. However, single-model optimizations struggle with this due to high dimensionality and strongly coupled variables. To overcome these challenges, we propose a two-stage “instantaneous load allocation—day-ahead scheduling” framework. Stage I employs a hybrid algorithm (ICSA-WOA) to optimize load allocations across various flow rates, generating a lookup table that effectively decouples the underlying physical model. Stage II utilizes this table alongside TOU prices to perform rapid day-ahead scheduling via dynamic programming (DP). Results demonstrate that ICSA-WOA achieves superior comprehensive performance compared to seven classical swarm intelligence algorithms. Furthermore, joint optimization of the pressure ratio and load via ICSA-WOA reduces the total power consumption by 9.7–10.9% relative to traditional fixed-ratio modes. Most significantly, while rigorously ensuring daily injection targets and safety, the proposed method reduces daily electricity costs by 3.3–14.2% compared to single-model approaches, providing a reasonable strategy for economic gas storage operations. Full article

Component obsolescence and supply-chain disruptions increasingly force engineers to spend significant time manually searching and comparing PDF datasheets to identify compatible replacement parts. We propose an AI-powered datasheet assistant based on a Retrieval-Augmented Generation (RAG) pipeline that automatically processes datasheets to accelerate component identification and matching. The core contribution is a summary-driven retrieval mechanism: a Large Language Model (LLM) generates a structured semantic summary of an input datasheet, and the vector embedding of this summary is used to retrieve semantically similar components from a reference database. The system also supports natural language question answering and structured component comparison. Its architecture separates scalable text-only reference indexing from more expensive query-time summarization and reranking. Validation includes a controlled synthetic benchmark and a pilot-scale real-world evaluation on 18 publicly listed microcontroller datasheets grouped into six engineering families. The synthetic benchmark is used to assess pipeline behavior under controlled conditions, while the real-world evaluation measures performance on heterogeneous manufacturer datasheets. In the real-world evaluation, structured summaries generated with Claude Sonnet 4.5 combined with cross-encoder reranking achieved a 72.2% Family Retrieval Rate at $k=1$ (13/18; Wilson 95% CI: 49.1–87.5%). Additional experiments with local LLM summaries indicate that retrieval performance depends strongly on summary quality and model capability, with lightweight local summarizers producing lower first-candidate retrieval performance in this setup. The analysis further reports confidence intervals, no-summary baselines, chunking sensitivity, and an Image Reference Rate metric used as a lexical reference proxy rather than a direct measure of visual grounding. Full article

Cage-based tilapia farming is highly vulnerable to rapid variations in water-quality parameters, particularly dissolved oxygen (DO) fluctuations, which can cause fish stress, fish mortality, and economic losses. In this study, we developed and field-evaluated an Internet of Things (IoT)- and LoRa-based water-quality-monitoring and aeration-actuation system for open-water tilapia cage farming. The system consists of distributed control nodes, a main node, a cloud database, and a mobile application for real-time monitoring of DO, pH, and water temperature, as well as remote and automatic oxygen-pump actuation. An automatic probe-lifting mechanism is integrated into the control node to reduce probe-submersion duration and mitigate the risk of sensor fouling during field operation. Field validation showed that the node equipped with the probe-lifting mechanism achieved better agreement with the reference instruments than the continuously submerged node, particularly for DO measurement, with RMSE values of 0.186 mg/L and 0.683 mg/L, respectively. A communication-performance evaluation showed 100% packet reception up to 1640 m, whereas packet reception was reduced at the longest tested distance of 2290 m, indicating that the field-deployment range should be interpreted cautiously under the tested LoRa configuration. Detection-latency experiments showed sub-second responsiveness, with average delays of 208.6–289.7 ms for single-hop communication and 438.9–529.4 ms for two-hop communication. Expert evaluation and farmer satisfaction assessment indicated positive perceptions of the system’s usability and practical relevance. However, the study has several limitations, including the short field-validation period, limited sensor replication, and a lack of direct fish production outcome measurements, which should be considered when interpreting the findings. Overall, the proposed system provides a practical platform for water-quality monitoring and aeration actuation in cage-based tilapia farming. Full article

To investigate the dynamic characteristics of key parameters in a piston gravity energy-storage system, an experimental system for novel piston gravity energy storage is designed and developed. Firstly, the structure and working principle of the piston gravity energy-storage system are analyzed. Adopting a modular modeling approach, the system is divided into four core modules, and the piston motion, vertical cylinder chamber pressure, hydraulic actuator, and turbine power models are established. Subsequently, a case study simulation is conducted on the piston gravity energy-storage system to model its dynamic characteristics during discharge conditions, analyzing the variation patterns of key parameters such as the chamber pressure, flow rate, and output power within the system. Finally, the experimental system integrates a digital controller with proportional–integral power regulation and an automatic mode switching logic to enable the constant power closed-loop control, with real-time acquisition of the chamber height, pressure, flow rate, and electrical parameters. The dynamic responses of various system parameters are analyzed. Experimental results indicate that under constant power charging and discharging conditions, the height of the upper chamber exhibits a linear trend, the pressure in the lower chamber is inversely proportional to the height of the upper chamber, and the flow rate remains stable with charging and discharging power. Neglecting energy losses of the pump and hydraulic turbine and only considering friction and hydraulic losses, the charge–discharge efficiency of the energy-storage experimental system is 65%. Full article

As data volumes grow and business requirements become increasingly complex, Hybrid Transactional/Analytical Processing (HTAP) technologies, capable of handling both Online Transaction Processing (OLTP) and Online Analytical Processing (OLAP) workloads on a single platform, have gained prominence. HTAP databases typically maintain dual data storage formats and dual query engines: one row-oriented for OLTP, and another column-oriented for OLAP. Query plans, known as hybrid plans, can be segmented and pushed down to execute on these different formats. However, existing HTAP solutions still face challenges in optimizing these hybrid plans, struggling to explore the vast space of potential execution strategies effectively. To address these issues, this study introduces a learning-based query optimizer for row–column dual store HTAP database systems, which automatically generates multiple high-quality query optimizer hints (HINTs) to derive candidate plans. To balance plan generation efficiency with plan quality, a lightweight, learning-based algorithm using Monte Carlo Tree Search (MCTS) for generating hybrid access HINTs is proposed. Moreover, a Transformer-based neural network model coupled with a hybrid plan feature representation method is developed to select the candidate execution plan with the lowest predicted execution time. This work focuses on latency-oriented hybrid-plan selection for analytical queries in a row–column dual-store HTAP architecture; the current evaluation does not cover full mixed OLTP/OLAP workload scheduling, transactional interference, or concurrency control, which are left as future work. Experimental results on AlloyDB Omni, a recent row–column dual-store HTAP database, using the real-world IMDB dataset and JOB benchmark demonstrate that our system reduces execution time by 75.02% compared to the Cost-Based Optimizer (CBO) and by 62.23% compared to the state-of-the-art row-store-based learning query optimizer in this evaluated analytical-query setting. Full article

Artificial intelligence (AI) plays an increasingly important role in maritime systems, enabling advanced monitoring, anomaly detection, and decision support. However, the reliability of such systems is challenged by distributional drift, which may significantly degrade model performance over time. While anomaly detection has been extensively studied in the context of data irregularities, considerably less attention has been devoted to detecting anomalies in AI model behaviour itself. In this study, we propose MARLIN-AD (Maritime AI Reliability and Learning Intelligence Network—Anomaly Detection), a dual-layer anomaly detection framework designed to jointly monitor anomalies in data streams and anomalies in model behaviour. The framework integrates data-centric detection methods with model-centric monitoring techniques, including distributional shift detection and prediction stability analysis, within a unified anomaly scoring mechanism. The evaluation is conducted using a fully controlled synthetic data generation process, enabling precise injection of anomalies and systematic simulation of distributional drift across multiple scenarios. Experimental results demonstrate a strong and consistent degradation of model performance under drift conditions. Statistical validation using non-parametric tests, permutation-based inference, and Bayesian bootstrap analysis confirms that the observed degradation is both statistically significant and practically meaningful. In particular, posterior distributions of performance differences indicate a near-zero probability that drifted configurations outperform the baseline model. The results highlight that model degradation under drift exhibits a consistent and structured pattern, reproducible across multiple independent random seeds. Furthermore, the study shows that model-centric monitoring provides the primary signal for detecting degradation—a finding corroborated by ablation analysis—while data-centric monitoring enhances interpretability and root-cause attribution. A pilot validation on publicly available Automatic Identification System (AIS) data from the Danish Maritime Authority confirms the applicability of the data-level component to real operational trajectories. The proposed framework contributes to the development of trustworthy AI systems by enabling comprehensive monitoring of both data integrity and model behaviour in dynamic environments. Full article

Comprehensive mechanized solid backfilling technology exhibits significant advantages in solid waste disposal, “three-under” coal mining, and dynamic disaster control. However, its large-scale application is constrained by low production efficiency, high unit production cost, and high labor intensity. Therefore, industrial upgrading through intelligent technologies is urgently required. In this study, methods including literature review, theoretical analysis, and field measurements are employed to propose three backfilling modes. The configurations of the six core subsystems under each mode are systematically summarized, and the core definition of an intelligent backfilling mine is established. Furthermore, a key technology framework for intelligent backfill mining is developed, based on PLC control and PID algorithms, with a closed-loop architecture centered on “perception–decision–execution.” Engineering applications demonstrate that the surface gangue intelligent pretreatment system achieves functions including automatic vehicle washing, intelligent dust suppression spraying at discharge points, dynamic metering during conveying, and adaptive adjustment of feeding systems. The intelligent surface-to-underground coal gangue vertical feeding system enables full silo alarm and level regulation. The underground jigging intelligent separation system realizes intelligent jigging ratio adjustment, intelligent bed layer measurement and control, and intelligent air volume regulation, with the coal content in gangue discharge maintained below 4%. At the working face, the intelligent solid backfilling system doubles monthly coal output, boosts backfilling efficiency by 50%, and cuts the workforce by 8–10 workers. The intelligent backfilling effectiveness monitoring system operates stably, with a working face weighting factor of 1.12 and precise ground deformation control within Grade I limits. Full article

Automated clinical documentation based on clinician-patient conversations is an emerging application of deep learning, driven by advances in medical speech recognition and natural language processing. Despite technological progress, real-world adoption remains limited. This review analyzes deep learning–based medical speech-to-text systems, focusing on methodologies, evaluation strategies, and barriers to clinical implementation. A systematic review of 31 studies was conducted, covering automatic speech recognition, clinical dialogue processing, and large language model-based documentation pipelines. Speech recognition accuracy varies considerably in noisy, multi-speaker, and spontaneous clinical environments. Downstream tasks such as entity extraction and summarization are highly sensitive to transcription errors and constrained by limited real-world datasets. Most systems lack external clinical validation and are tested in controlled settings. Key challenges include speaker diarization, domain adaptation, privacy protection, and the need for standardized evaluation frameworks. Although LLMs demonstrate strong potential, concerns remain regarding hallucinations and factual reliability, necessitating improved robustness and clinician oversight. Full article

Controller performance is strongly influenced by its parameters. Estimating these parameters requires an effective estimation approach for obtaining the best possible response. This study proposes a novel methodology for the estimation of controller parameters, utilizing the dream optimization algorithm (DOA) and a new objective function. The proposed method is employed to determine the optimal parameters of various PID controllers used in the automatic voltage regulator (AVR) system. Thus, the suggested objective function consists of transient response metrics and the stability index “integral of time-weighted absolute error ($ITAE$)”. Three different PID controllers are used, which are cascaded PIPD with filter (CPIPDF), cascaded fractional-order PI fractional-order PDF (CFOPIFOPDF), and PIDF. The DOA’s performance is compared with famous and recent optimizers and shows more reliable performance. For example, based on the statistical analysis, the DOA obtained a standard deviation of 0.0042, while the closest competitor obtained 0.0089. Furthermore, the CPIPDF, CFOPIFOPDF, and PIDF controllers are compared under a wide variety of operating conditions. Based on ITAE, the CPIPDF controller achieved lower values than the CFOPIFOPDF and PIDF controllers. Also, the results show that the CPIPDF controller achieves better performance than other published controllers. For instance, the CPIPDF controller improves AVR performance by approximately 45.3% compared to the fireworks whale optimization algorithm-based PIDD² controller in the case of varying load condition impact. Moreover, scenarios that remain insufficiently addressed in the literature, such as communication delays, restricted excitation voltages, and external disturbances, are considered. Full article

Elevators exemplify mechatronics by integrating mechanical, electrical, and software systems. This paper discusses a four-story tabletop elevator model developed to demonstrate mechatronics and automation concepts in engineering education. The system utilized an Arduino MEGA microcontroller, 3D-printed components, an integrated servo motor, and standard electronics to replicate commercial elevator logic. The physical design features a ball screw linear actuator for vertical motion. It replicates dual-door systems with one door on the moving car and fixed doors at each floor that open simultaneously upon arrival. Development included designing the physical model, prototyping control algorithms, and integrating hardware and software. The model successfully demonstrated key functions: automatic dual-door operation, safety interlocks, smooth inter-floor motion, responsive floor-selection buttons with LED feedback, and efficient routing algorithms prioritizing requests based on current direction and location. Performance testing confirmed that the model accurately replicates modern elevator behavior and serves as an effective educational tool. Full article

Nowadays, Python is very popular as the first programming language for novices, including high school students, to learn due to its short code features with rich libraries. Thus, it is important to provide a learning environment supporting studies starting from the fundamentals, since students have no knowledge on how a program runs on a computer. Previously, we have developed a web-based programming learning assistant system (PLAS) to allow the self-study of major programming languages, including Python, by university students. It offers several types of exercise problems that have different learning goals and levels for step-by-step study. Any student answer is automatically marked at the answer interface for quick feedback. However, PLAS has not implemented functions to assist the learning needs of high school-level students. In this paper, we propose a novice-friendly answer interface for a Python programming learning assistant system (PyPLAS) that introduces a code behavior visualization and an AI assistant with learning logs. The visualization allows learners to observe the changes in variable states and the control flow. The assistant provides multi-level hints during learning and reflective feedback after it by analyzing the logs based on engagement, reasoning strategies, learning pace, and tool usage. For evaluation, we implemented the proposed interface using Python Flask for the web platform and Ollama as a locally deployed AI model. A pilot application was conducted with high school students solving introductory Python exercises in PyPLAS. The results showed high task completion, positive questionnaire responses toward embedded visualization and interface usability, and teacher-observed usefulness of the four-dimensional learning analytics for interpreting learner behaviors. These findings provide preliminary evidence for the feasibility and practical value of the proposed interface, while larger controlled studies are required to validate its instructional effectiveness. Full article

In the anti-sway control of variable-length double-pendulum gantry cranes, traditional sliding mode control relies on high switching gains, which can cause chattering. Additionally, the introduction of neural networks presents challenges in tuning high-dimensional parameters. To address these issues, this study proposes an adaptive hierarchical sliding mode control strategy based on an RBF neural network and particle swarm optimization. First, a low-energy-dissipation dynamic model is established without the small-angle assumption. Second, a composite hierarchical sliding surface is designed to achieve multi-objective decoupling, and an RBF neural network is utilized to approximate the system’s unknown dynamics online, thereby reducing switching gains and suppressing chattering. The asymptotic stability of the closed-loop system is proven based on Lyapunov theory. Finally, a particle swarm optimization algorithm is introduced to achieve automated, high-precision matching of high-dimensional controller parameters. Simulation results indicate that the control method designed in this paper can achieve automatic matching of high-dimensional parameters, effectively resolving the chattering issue in sliding mode control. Furthermore, under wide-range parameter perturbations and external multi-source disturbances, the controller exhibits strong robustness and demonstrates excellent positioning and anti-chattering performance. Full article

Background/Objectives: The main aim of this article is to introduce a novel tool that allows the automatic scoring of many of the subtests from the conventional subpart of the Behavioural Inattention Test (BIT) and its German adaptation, the Neglect Test (NET). BIT and NET are standard test batteries used in the diagnosis of neglect. Our article has two parts. First, we examine the shortcomings of manual scoring, and secondly, we introduce our computer vision tool and evaluate its diagnostic validity and efficacy. Methods: In Part 1, diagnostic consistency was examined across raters with varying expertise using selected BIT and NET tasks, with repeated assessments conducted under controlled evaluation conditions. In Part 2, a computer vision-based tool (CANDO) was developed to automate scoring using a deterministic computer vision pipeline designed to reproducibly apply scoring criteria across tasks. The performance of CANDO was compared with ground truth across cancellation, line bisection, and copying tasks. Results: Manual scoring showed high overall agreement between and within raters under ideal conditions. However, diagnostic classification still differed across raters and repeated assessments. These inconsistencies were primarily driven by drawing and copying tasks. CANDO achieved very high accuracy for cancellation and line bisection tasks and strong agreement for copying tasks, while reducing variability associated with subjective judgment, time pressure, and oversight. The remaining discrepancies between computer vision and human raters had limited impact on diagnostic outcomes. Conclusions: Manual assessment of unilateral neglect is vulnerable to inconsistencies arising from subjective evaluation and the structural limitations of scoring systems. Computer vision-based automation can reduce diagnostic variability, improve reproducibility, and increase assessment efficiency, while preserving clinically relevant information. The presented framework provides a practical tool to support higher-quality neglect assessment. Full article

In order to predict the wind speed of the excavation roadway, control the frequency conversion operation of the local fan in real-time, and realize the real-time monitoring, collaborative prediction, and frequency conversion control of the ventilation state of the excavation face, the frequency conversion ventilation control system of the excavation face is designed. Based on the theory of frequency conversion control, the genetic-neural network wind speed prediction optimization model was established, and the frequency conversion ventilation control system of the excavation face was designed by using S7-200 SMART PLC. The system test results show that the genetic-neural network optimization model can collaboratively predict wind speed according to the environmental parameters (dust concentration, methane concentration, temperature, and humidity, etc.) of different working conditions. The frequency conversion ventilation control system realizes the real-time monitoring of the environmental parameters of the excavation surface, and also provides two control modes: automatic and manual. Compared with the traditional constant power frequency control fan wind speed, the PID wind speed closed-loop control technology can control the fan wind speed by frequency conversion, so that the actual wind speed of the roadway continues to approach the predicted value stably. The variable frequency ventilation control system can be widely used in different types of mines to realize the adaptive control response of ventilation equipment. Full article

Although PI and PID controllers are mature control laws, their effect on energy-related variables is rarely isolated in a complete electric vehicle traction model when the plant, controller tuning basis and driving conditions are kept unchanged. A full-system MATLAB/Simulink model was developed, comprising a DC motor with PWM H-bridge, reduction gear, vehicle dynamics and a lithium-ion battery with SOC monitoring. Fixed-gain PI and PID configurations were compared under FTP75, with US06 added as a dynamic-cycle assessment. Speed tracking was evaluated using RMSE, MAE, IAE and ITAE, while energy behavior was assessed through SOC depletion, battery voltage, current and braking-command signals. Under FTP75, both controllers achieved nearly identical tracking accuracy, with an overall RMSE of 0.1525 km/h across the active intervals. Despite this kinematic equivalence, PID reduced SOC depletion by 0.980 percentage points over 4.963 km and produced a less intense but more distributed braking command. The additional 600 s US06 simulation did not confirm a general PID advantage: both controllers reached the same maximum speed and showed practically identical tracking accuracy, while PID did not reduce SOC depletion. The results show that the derivative channel changes the control-command pattern, but it does not automatically improve kinematic or energy performance under fixed-gain tuning. Full article