Evaluation and Reduction of Energy Consumption of Railway Train Movement on a Straight Track Section with Reduced Freight Wagon Mass

Energies, Journal Year: 2025, Volume and Issue: 18(2), P. 280 - 280

Published: Jan. 10, 2025

This paper presents an evaluation and reduction of energy consumption during railway train movement on a straight track section with reduced freight wagon mass. A theoretical model was developed to simulate based input parameters, including speed, gradient, length, travel time, The results indicate that increases by 18.9% as speed rises 90 km/h gradients increase 2.0‰, while decreases 14.5% descending gradient 1.5‰, which corresponds the expected dynamics trains. These are supported experiments showing MAPE error does not exceed 1.9%, can confirm accuracy model. comprehensive analysis potential in mass also conducted. Using design 2.3% allows for 8–89 kW·h, depending length movement.

Language: Английский

---

Cognitive Method for Synthesising a Fuzzy Controller Mathematical Model Using a Genetic Algorithm for Tuning

Big Data and Cognitive Computing, Journal Year: 2025, Volume and Issue: 9(1), P. 17 - 17

Published: Jan. 20, 2025

In this article, a fuzzy controller mathematical model synthesising method that uses cognitive computing and genetic algorithm for automated tuning adaptation to changing environmental conditions has been developed. The technique consists of 12 stages, including creating the control objects’ coefficients using classical methods. research pays special attention error parameters their derivative fuzzification, which simplifies development logical rules helps increase stability systems. were tuned in computational experiment based on helicopter flight data. results show an integral quality criterion from 85.36 98.19%, confirms efficiency by 12.83%. use made it possible significantly improve turboshaft engines’ gas-generator rotor speed performance, reducing first second types errors 2.06…12.58 times compared traditional

Language: Английский

---

Neural Network System for Predicting Anomalous Data in Applied Sensor Systems

Applied System Innovation, Journal Year: 2024, Volume and Issue: 7(5), P. 88 - 88

Published: Sept. 23, 2024

This article advances the research on intelligent monitoring and control of helicopter turboshaft engines in onboard conditions. The proposed neural network system for anomaly prediction functions as a module within engine expert system. A SARIMAX-based preprocessor model was developed to determine autocorrelation partial training data, accounting dynamic changes external factors, achieving accuracy up 97.9%. modified LSTM-based predictor with Dropout Dense layers predicted sensor tested error margin 0.218% predicting TV3-117 aircraft gas temperature values before compressor turbine during one minute flight. reconstructor restored missing time series replaced outliers synthetic values, 98.73% accuracy. An detector using concept dissonance successfully identified two anomalies: malfunction sharp drop minutes activity, type I II errors below 1.12 1.01% detection under 1.611 s. system’s AUC-ROC value 0.818 confirms its strong ability differentiate between normal anomalous ensuring reliable accurate detection. limitations involve dependency quality data from sensors, affected by malfunctions or noise, LSTM network’s (up 97.9%) varying conditions, model’s high computational demand potentially limiting real-time use resource-constrained environments.

Language: Английский

---

Helicopter Turboshaft Engines’ Gas Generator Rotor R.P.M. Neuro-Fuzzy On-Board Controller Development

Energies, Journal Year: 2024, Volume and Issue: 17(16), P. 4033 - 4033

Published: Aug. 14, 2024

The work is devoted to the helicopter turboshaft engines’ gas generator rotor R.P.M. neuro-fuzzy controller development, which improves control accuracy and increases system’s stability external disturbances adaptability changing operating conditions. Methods have been developed, including improvements automatic system structural diagram made it possible obtain transfer function in bandpass filter form. also improved fuzzy rules base neuron activation mathematical model, significantly accelerated training process. frequency time characteristics analysis showed that effectively controlled engine reduced vibration. for ensuring a guaranteed margin synthesis of an adaptive were studied, achieve high reliability. results developed provided with amplitude phase margins, compensating changes Experimental studies demonstrated quality by 2.31–2.42 times compared previous controllers 5.13–5.65 classic PID controllers. Control errors 1.84–2.0 5.28–5.97 times, respectively, confirming controller’s efficiency adaptability.

Language: Английский

---

Helicopters Turboshaft Engines Neural Network Modeling under Sensor Failure

Journal of Sensor and Actuator Networks, Journal Year: 2024, Volume and Issue: 13(5), P. 66 - 66

Published: Oct. 10, 2024

This article discusses the development of an enhanced monitoring and control system for helicopter turboshaft engines during flight operations, leveraging advanced neural network techniques. The research involves a comprehensive mathematical model that effectively simulates various failure scenarios, including single cascading failure, such as disconnections gas-generator rotor sensors. employs differential equations to incorporate time-varying coefficients account external disturbances, ensuring accurate representation engine behavior under different operational conditions. study validates NARX architecture with backpropagation training algorithm, achieving 99.3% accuracy in fault detection. A comparative analysis genetic algorithms indicates proposed algorithm outperforms others by 4.19% exhibits superior performance metrics, lower loss. Hardware-in-the-loop simulations Matlab Simulink confirm effectiveness model, showing average errors 1.04% 2.58% at 15 °C 24 °C, respectively, high precision (0.987), recall (1.0), F1-score (0.993), AUC 0.874. However, model’s is sensitive environmental conditions, further optimization needed improve computational efficiency generalizability. Future should focus on enhancing adaptability validating real-world scenarios.

Language: Английский

---

Method of Helicopter Turboshaft Engines’ Protection During Surge in Starting Mode

Energies, Journal Year: 2025, Volume and Issue: 18(1), P. 168 - 168

Published: Jan. 3, 2025

This article proposes a mathematical model for protecting helicopter turboshaft engines from surges, starting with fuel metering supply and maintaining stable compressor operation. The includes several stages: first, is supplied according to specified program; second, an unstable operation signal determined based on the gas temperature in front of turbine generator rotor speed derivatives ratio; at third stage, when ratios’ threshold value exceeded, stopped, ignition system turned on. Then, restored reduced consumption, corrected, followed by return regular neural network implementing this method consists layers, including calculation, comparison threshold, correction consumption speed. input data are A instability generated if ratio exceed value, which leads adjustment regulation 28…32%. backpropagation algorithm hyperparameter optimization via Bayesian was used train network. computational experiments result TV3-117 engine semi-naturalistic simulation stand showed that proposed effectively prevents surge stabilizing pressure, vibration, reduces 29.7% under start-up conditions. Neural quality metrics such as accuracy (0.995), precision (0.989), recall (1.0), F1-score (0.995) indicate high efficiency method.

Language: Английский

---

A Deep Learning Framework for High-Frequency Signal Forecasting Based on Graph and Temporal-Macro Fusion

Applied Sciences, Journal Year: 2025, Volume and Issue: 15(9), P. 4605 - 4605

Published: April 22, 2025

With the increase in trading frequency and growing complexity of data structures, traditional quantitative strategies have gradually encountered bottlenecks modeling capacity, real-time responsiveness, multi-dimensional information integration. To address these limitations, a high-frequency signal generation framework is proposed, which integrates graph neural networks, cross-scale Transformer architectures, macro factor modeling. This enables unified structural dependencies, temporal fluctuations, macroeconomic disturbances. In predictive validation experiments, achieved precision 92.4%, recall 91.6%, an F1-score 92.0% on classification tasks. For regression tasks, mean squared error (MSE) absolute (MAE) were reduced to 1.76×10−4 0.96×10−2, respectively. These results significantly outperformed several mainstream models, including LSTM, FinBERT, StockGCN, demonstrating superior stability practical applicability.

Language: Английский

---

Intelligent Method of Identifying the Nonlinear Dynamic Model for Helicopter Turboshaft Engines

Sensors, Journal Year: 2024, Volume and Issue: 24(19), P. 6488 - 6488

Published: Oct. 9, 2024

This research focused on the helicopter turboshaft engine dynamic model, identifying task solving in unsteady and transient modes (engine starting acceleration) based sensor data. It is known that about 85% of engines operate steady-state modes, while only around 15% modes. Therefore, developing multi-mode models account for behavior during these a critical scientific practical task. The model acceleration has been further developed using on-board parameters recorded by sensors (gas-generator rotor r.p.m., free turbine speed, gas temperature front compressor turbine, fuel consumption) to achieve 99.88% accuracy dynamics parameters. An improved Elman recurrent neural network with stack memory was introduced, enhancing robustness increasing performance 2.7 times compared traditional networks. A theorem proposed proven, demonstrating total execution time N Push Pop operations does not exceed certain value O(N). training algorithm delay considerations Butterworth filter preprocessing, reducing loss function from 2.5 0.12% over 120 epochs. gradient diagram showed decrease time, indicating model’s approach minimum function, optimal settings ensuring stable training.

Language: Английский

---

Helicopter Turboshaft Engines’ Neural Network System for Monitoring Sensor Failures

Sensors, Journal Year: 2025, Volume and Issue: 25(4), P. 990 - 990

Published: Feb. 7, 2025

An effective neural network system for monitoring sensors in helicopter turboshaft engines has been developed based on a hybrid architecture combining LSTM and GRU. This enables sequential data processing while ensuring high accuracy anomaly detection. Using recurrent layers (LSTM/GRU) is critical dependencies among time series analysis identification, facilitating key information retention from previous states. Modules such as SensorFailClean SensorFailNorm implement adaptive discretization quantisation techniques, enhancing the input quality contributing to more accurate predictions. The demonstrated detection at 99.327% after 200 training epochs, with reduction loss 2.5 0.5%, indicating stability processing. A algorithm incorporating temporal regularization combined optimization method (SGD RMSProp) accelerated convergence, reducing 4 min 13 s achieving an of 0.993. Comparisons alternative methods indicate superior performance proposed approach across metrics, including 0.993 compared 0.981 0.982. Computational experiments confirmed presence highly correlated sensor method's effectiveness fault detection, highlighting system's capability minimize omissions.

Language: Английский

---

Nonlinear Model Predictive Control Method for High-speed Helicopter Power System Based on Integrated Onboard Model

Aerospace Science and Technology, Journal Year: 2025, Volume and Issue: unknown, P. 110093 - 110093

Published: Feb. 1, 2025

Language: Английский

---