824 / 2019-04-29 20:13:44

A new method of adaptive multistable stochastic resonance and its application in fault diagnosis of electric power equipment

Adaptive stochastic resonance,Multistable stochastic resonance,Multiple parameter optimization,Power equipment fault diagnosis,Weak feature extraction

Extracting weak characteristics of early mechanical faults of wind turbines has always been a key problem in fault diagnosis. The existing weak feature extraction methods mostly detect fault features from the perspective of noise elimination. However, for the weak characteristic signal of heavy noise pollution situation, if we continuous use the perspective of noise elimination, it will cause the characteristic signal lose. Aiming at the shortcomings of the original noise filtering methods, this paper establishes a noise-aided analysis of early fault model based on the theory of tristable model stochastic resonance. But the existing stochastic resonance method has some shortcomings. Some scholars proposes an improved SR that can further increase the output signal-to-noise ratio and solve such as double well potential particle motion saturation problem. In this paper, owing to the performance of stochastic resonance methods is mostly decided by its system parameters, the influence of system parameters on the model is discussed, and applies particle swarm optimization to achieve the optimal output.