Refined composite hierarchical multiscale Lempel-Ziv complexity: A quantitative diagnostic method of multi-feature fusion for rotating energy devices

Digitalization and intellectualization of energy system require the fault information of energy conversion devices to be captured as accurately as possible from the massive data, so as to realize early fault alert. However, the comprehensiveness of feature extraction, the setting difficulty of model parameters and applicability of various scenarios have shortcomings by conventional methods, leading to their limitations in measured signals with multi-source and multi-feature information. Therefore, this paper exploits a quantitative diagnostic method named refined composite hierarchical multiscale Lempel-Ziv complexity (RCHMLZC). Firstly, the enhanced hierarchical decomposition and multiscale Lempel-Ziv complexity (MLZC) are coupled to develop hierarchical multiscale Lempel-Ziv complexity (HMLZC), which overcomes the drawback of MLZC that cannot quantify the complexity of signals at different frequencies. Secondly, RCHMLZC is proposed to solve the problem that LZC value of HMLZC fluctuates greatly under high scale factors, and then is used to extract the features of vibration signals. Finally, the extracted features are input into the random forests model to realize the efficient recognition of different status signals of rotating energy devices. A total of 14 types of multi-feature fault signals from bearings, shafting and runner are used to verify the reliability and superiority of the proposed method. Compared to the five conventional models, the comprehensive indicators of the proposed method for bearing fault experiments are improved by 1.549%, 4.637%, 14.153%, 20.242% and 22.112%, while the values for the shafting fault experiments are improved by 0.404%, 0.427%, 2.778%, 2.722% and 5.895%. In addition, the proposed method is applied to the analysis of the fault cases of hydraulic turbine, demonstrated the ability to zero miscalculation. It would be a helpful tool to improve energy conversion efficiency and reduce maintenance cost.