Fault Diagnosis of Wind Turbine Bolts based on ICEEMD-SSA-SVM Model

Abstract

Background: Compared with traditional power generation systems, wind turbines have more units and work in a more harsh environment, and thus have a relatively high failure rate. Among blade faults, the faults of high-strength bolts are often difficult to detect and need to be analyzed with high-precision sensors and other equipment. However, there is still little research on blade faults. Methods: The improved complete ensemble empirical mode decomposition (ICEEMD) model is used to extract the fault features from the time series data, and then combined with the support vector machine optimized by sparrow search algorithm (SSA-SVM) to diagnose the bolt faults of different degrees, so as to achieve the purpose of early warning. Results: The results show that the ICEEMD model used in this paper can extract the bolt fault signals well, and the SSA-SVM model has a shorter optimization time and more accurate classification compared with models such as PSO-SVM. Conclusion: The hybrid model proposed in this paper is important for bolt fault diagnosis of operation monitoring class.