基于FCSE结合BGWO－LSSVM的风电机组故障诊断OA

With increasing global attention on the development of wind energy resources,ensuring the safe and stable operation of wind turbines has become crucial for the efficient utilization.However,ex-tracting information from complex vibration signals units for fault feature characterization faces many challenges.For this reason,a feature extraction method based on the fractional order idea was proposed to improve the feature extraction capability of wind turbine unit signals under different states.Conside-ring the fractional-order characteristics of signals,the extraction effect of complex signal features was improved by introducing the fractional-order cosine similarity entropy(FCSE)method.Meanwhile,the multi-strategy improved grey wolf optimization algorithm(BGWO)was combined to find the optimal parameter combinations of least squares support vector machine(LSSVM)for efficient feature classifi-cation.Noise with a signal-to-noise ratio of 3 dB was added to the original signal,and FCSE was com-pared and analyzed with two traditional feature entropy to evaluate its anti-noise performance.The si-mulation results show that the feature extraction ability of FCSE on a given dataset is significantly better than that of the other two methods,while the classification accuracy of the proposed fault diagnosis sys-tem on the laboratory dataset and actual vibration signals reaches 100.0%and 97.50%,respectively,proving its validity and reliability in practical applications.