Multi-mode signal fusion and improved residual dense network fault diagnosis of nuclear power plant
2023 CAA Symposium on Fault Detection, Supervision and Safety for Technical Processes (SAFEPROCESS)(2023)
摘要
The primary cooling system is one of the important components of a nuclear power plant, and once the main cooling system fails that, it can lead to serious consequences. In this paper, a method of multimodal signal fusion and improved residual dense network for nuclear power plant fault diagnosis is proposed. In order to obtain the multifaceted information of the fault, firstly, the original signal is changed in the frequency domain, the time-frequency domain is changed, and then the fused information is fed into the residual dense network improved by convolutional attention for learning. The convolutional attention mechanism can improve the convergence speed and accuracy of the network. Finally, the simulation experiment is verified by the simulation data center of Hualong-1 nuclear power plant. The experimental results as well as comparative analysis show that the algorithm has a better recognition effect than the traditional algorithm on the faults of the main cooling system.
更多查看译文
关键词
nuclear power plant fault diagnosis,multimode signal fusion,convolutional attention mechanism,improved residual-dense network,neural network
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要