Fault detection method of hydraulic turbine unit based on KPCA-PSO-SVM
LI Zixin1, LIN Haijun1*, XU Xiong2, WEN Lepeng2
1. College of Engineering and Design, Hunan Normal University, Changsha, Hunan 410081, China; 2. School of Electrical and Information Engineering, Hunan University, Changsha, Hunan 410081, China
Abstract:During the operation of the hydro turbine unit, due to the interference of hydraulic, mechanical, electromagnetic and other factors, it is very easy to induce failure, which affects the safe and stable operation of the unit. Based on the extraction of noise signal characteristics of hydro turbine unit operation, this paper proposes a fault detection method for hydro turbine unit noise signal based on kernel principal component analysis(Kernel PCA), improved particle swarm optimization(PSO)and support vector machine(SVM). This method first extracts 13-dimensional features of the time domain and time-frequency domain from the original noise signal collected at the site of the hydropower station, which overcoming the limitation of a single number of features. The eigenvectors extracted by KPCA are then used to reduce the dimensionality of the extracted feature vectors, and then the improved particle swarm algorithm(PSO)is used to optimally seek parameters for the SVM model. The optimized support vector machine(SVM)is used to detect the faults of the extracted features, and the identification of the three common operating states of the hydro turbine unit is completed. The experimental results show that the fault detection method based on KPCA-PSO-SVM proposed in this paper has the classification recognition rate of the operating state of the hydro turbine unit is 96.73%, which is higher than that of SVM algorithm, neural network, KNN, random forest and other methods, and confirms the effectiveness of this method.
黎梓昕,林海军*,徐雄,温乐鹏. 基于KPCA-PSO-SVM的水轮机组故障检测方法[J]. 排灌机械工程学报, 2023, 41(5): 467-474.
LI Zixin,LIN Haijun*,XU Xiong,WEN Lepeng. Fault detection method of hydraulic turbine unit based on KPCA-PSO-SVM. Journal of Drainage and Irrigation Machinery Engin, 2023, 41(5): 467-474.