The durability of proton exchange membrane fuel cells (PEMFC) is an important issue that restricts their large-scale application. To improve reliability during use, this paper proposes a short-term performance degradation prediction model using particle swarm optimization (PSO) to optimize the gate recurrent unit (GRU). After training only data from first 300 h, good accuracy can be achieved. Compared with traditional GRU algorithm, proposed method reduces root mean square error (RMSE) and...

Vehicle detection in severe weather has always been a difficult task the environmental perception of intelligent vehicles. This paper proposes vehicle method based on pseudo-visual search and histogram oriented gradients (HOG)–local binary pattern (LBP) feature fusion. Using radar information, this can directly extract region interest (ROI) vehicles from infrared images by imitating human vision. Unlike traditional methods, mechanism is independent complex image processing interferences,...

Abstract To improve the prediction accuracy of performance degradation trend proton exchange membrane fuel cell (PEMFC), this paper proposes a temporal convolutional network (TCN) model based on genetic algorithm (GA) optimization to predict PEMFC. Firstly, variational mode decomposition and wavelet threshold denoising algorithms are used denoise original data. Then hyperparameters TCN optimized by GA, GA-TCN for predicting PEMFC is constructed. Finally, uses stack experimental dataset...

Accurate and efficient road recognition is very important for the control of mobile robots autonomous vehicles. In this article, a new surface method based on 24-GHz millimeter-wave radar proposed, which has better environmental adaptability compared with machine vision absolute cost advantage lidar. The core our to propose feature fusion prior knowledge data-driven. First, echo signal subjected statistical analysis, thereby confirming distinguishability signals various types. Then, we...

Radar observation variables reflect the precipitation amount of strong convective processes, which accurate forecast is an important difficulty in weather forecasting. Current forecasting methods are mostly based on radar echo extrapolation, has insufficiency input information and ineffectiveness model architecture. This paper presents a Bidirectional Long Short-Term Memory method for attention mechanism residual neural network (ResNet-Attention-BiLSTM). First, this uses ResNet to...

Rapid growth of China’s urban road vehicles, in particular, the increase number gasoline leads to an traffic congestion and problems pertaining air pollution. The establishment emission inventory vehicles is influenced by several factors, like environmental characteristics, vehicle conditions, so on. In order obtain accordance with actual situation different regions, this study proposed a method establishing list regional differences. Comprehensive consideration evaluation various factors...

Virtual test evaluation is an important development direction for automatic driving technology testing and evaluation. The millimeter-wave radar sensor model used in virtual should meet the real-time accuracy requirements of system. current can satisfy but cannot simulate physical characteristics radar. This study proposes a model, which introduces cross-sectional (RCS) area judgment index based on geometric clipping extraction target visibility judgment. Through simulation verification...

Road surface recognition is an important research content of automated driving environment perception, which can provide the basis for decision-making and motion planning driving. To improve accuracy robustness road recognition, this paper proposes a scheme based on fusion machine vision tire noise. Firstly, we extracted LBP features deep learning from images used PCA algorithm to reduce dimension features. Then, noise were by using statistics MFCC algorithm. We fused image acoustic generate...

<title>Abstract</title> Radar observation variables reflect the precipitation amount of strong convective processes, which accurate forecast is an important difficulty in weather forecasting. Current forecasting methods are mostly based on radar echo extrapolation, has insufficiency input information and ineffectiveness model architecture. This paper presents a Bidirectional Long Short-Term Memory method for attention mechanism residual neural network (ResNet-Attention-BiLSTM). First, this uses...

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

To detect possible failures of the pressure sensor in fuel cell engine air supply subsystem, this study proposes a fault identification method based on Random forest. simulate faults sensor, we injected deliberate and constructed dataset it. This includes gradual fixed deviation complete deviation. The peRandom forestormance training accuracy is observed different datasets, aiming to identify types caused by presence polluted or corrupted samples dataset. results indicate that machine...

Abstract In recent years, with the increasing adoption of hybrid vehicles, energy management strategies &amp;#xD;have become a prominent research focus. Accurate Vehicle Speed Prediction (VSP) is critical &amp;#xD;prerequisite for achieving optimal results in predictive strategies. However, &amp;#xD;existing speed prediction algorithms fail to fully leverage vehicle data enhance &amp;#xD;accuracy. Therefore, novel Net (VSPNet) proposed this study. &amp;#xD;Firstly, we constructed combined...

In the development of optimal braking force distribution strategy for a dual-motor-drive electric vehicle (DMDEV) with series cooperative system, three key factors were taken into consideration, i.e. regenerative coefficient between front and rear motor (β), energy recovery at wheels (α3), front-and-rear-axle (λ). First, overall power loss model two surface-mounted permanent magnetic synchronous motors (SMPMSMs) was created based on d-q axis equivalent circuit model. The relationship β...