Health condition monitoring of hydraulic system based on ensemble support vector machine

Hydraulic system is a vital transmission system for its high stability and fast reaction as well as high transmission ratio. Whereas, hydraulic systems usually operate in a tough environment and need to be ensure for normal operating, which make it essential to precisely detect the health status of every significant component in a hydraulic system. A novel health condition monitoring method for hydraulic system is proposed in this paper based on ensemble support vector machine. Firstly, statistical features are extract from multiple sensor signals to describe the health condition characteristics of the hydraulic system. Then, the extracted features are selected using Pearson correlation coefficient. Finally, the health condition identification is realized based on ensemble support vector machine with stacking algorithm. The experimental results show that the proposed method for health condition identification of the hydraulic system is better than the other methods.