In this article, a new fault diagnosis framework for proton exchange membrane fuel cells (PEMFCs) based on electrochemical impedance spectroscopy (EIS) and the deep learning method is proposed. Specifically, this work employs the PEMFC knowledge to drive the training process of the deep learning network, which makes it possible to improve fault diagnosis performance by deep learning algorithms with a limited scale of actual measured EIS data. A pretraining network is developed to predict the equivalent circuit model (ECM) parameters, which could reduce the time consumption of ECM parameter identification. Besides, considering that the ECM parameters are susceptible to significant changes due to nonfault operation, a fine-tuning network is designed to generate robust diagnosis features, which could support the proposed framework working in different environments. Moreover, the complex neural networks are adopted in the proposed framework to extract features from EIS data, which is composed of complex impedances. Finally, a new evaluation metric PScore is proposed to assess the performance of the diagnosis framework from the perspective of practical applications. The experiments are performed to demonstrate the effectiveness of each component in the framework, and the proposed algorithm has significant improvements in fault diagnosis performance and computational efficiency over traditional algorithms.

Proton Exchange Membrane Fuel Cell (PEMFC) is considered as a clean alternative energy that has been widely used in many fields. However, the large-scale commercialization of PEMFC is still limited by its durability performance. Prognostics and Health Management (PHM) is considered as a helpful solution to improve the durability of PEMFC with degradation prediction and health-based control and maintenance methods. The long-term PEMFC prognostic task, as a primary component in PHM, has received extensive attention from both academia and industry. Recently, the combination of deep neural networks and PEMFC prognostics has expressed a broad research perspective. The deep-learning-based methods, such as Convolutional Neural Network (CNN), Echo State Network (ESN), and Recurrent Neural Network (RNN) family methods, have been well-studied by many researchers and have shown satisfactory performance in degradation prediction. However, the long-term prediction performance is still limited by the model structure and accumulative errors. This paper focuses on the degradation of the PEMFC and a Transformer-based PEMFC prognostic framework is proposed to predict the long-term degradation of the PEMFC system. The Transformer model is applied for the first time to predict the degradation of the PEMFC, and a series-attention mechanism is proposed to replace the self-attention mechanism in the Vanilla Transformer which could improve the health indicator (HI) prediction performance. Finally, the PEMFC dynamic durability test data is utilized to evaluate the performance of the proposed framework in both multi-step-ahead prediction and long-term prognostic conditions, and the experimental results illustrate the feasibility and effectiveness of the proposed method.

The capability of UAVs for efficient autonomous navigation and obstacle avoidance in complex and unknown environments is critical for applications in agricultural irrigation, disaster relief and logistics. In this paper, we propose the DPRL (Distributed Privileged Reinforcement Learning) navigation algorithm, an end-to-end policy designed to address the challenge of high-speed autonomous UAV navigation under partially observable environmental conditions. Our approach combines deep reinforcement learning with privileged learning to overcome the impact of observation data corruption caused by partial observability. We leverage an asymmetric Actor–Critic architecture to provide the agent with privileged information during training, which enhances the model’s perceptual capabilities. Additionally, we present a multi-agent exploration strategy across diverse environments to accelerate experience collection, which in turn expedites model convergence. We conducted extensive simulations across various scenarios, benchmarking our DPRL algorithm against state-of-the-art navigation algorithms. The results consistently demonstrate the superior performance of our algorithm in terms of flight efficiency, robustness and overall success rate.

In the field of computer vision, DEtection TRansformer (DETR) has received significant recognition for its ability to streamline the design process of object detectors through the concept of set prediction. However, its exceptional performance comes at the cost of a high parameter count and significant computational requirements. Moreover, its ability to detect small objects is compromised, making it less suitable for analyzing high-altitude Unmanned Aerial Vehicle (UAV) images. This paper proposes UAV-DETR, a DETR architecture specifically designed for detecting UAV images captured at high altitudes, which achieves a trade-off between parameter count and precision. UAV-DETR is built in two steps: first, inverted residual structures are used to preserve low-dimensional image features, followed by a carefully designed cascaded linear attention mechanism to mitigate parameter redundancy. Through observation and analysis of the attention diffusion issue in the encoder, a cross-channel dynamic sampling mechanism is proposed, which effectively expands the model's receptive field while maintaining accuracy. In addition, the loss function is redesigned by incorporating the Wasserstein distance, which is insensitive to bounding boxes, in order to significantly enhance the convergence speed of the model. Extensive experimental results on two major benchmarks, i.e. VisDrone and UAVDT, validate the simplicity and efficiency of our model. Specifically, on the VisDrone2021 public test set, UAV-DETR exhibits superior performance with only 14 million parameters compared to YOLOv8$_{m}$, reducing the model's parameter count and complexity by 44% and 10% respectively, while achieving a 16.6% improvement in accuracy, without any data augmentation or post-processing procedures.