Application of Improved Fish Swarm Algorithm in Fault Diagnosis of Rolling Bearing
Aiming at the problem that the vibration signal of rolling bearing fault is weak and difficult to identify,a neural net-work diagnosis method using an improved artificial fish swarm algorithm optimization is proposed.Firstly,the speed dynamic pa-rameters are introduced to improve the fixed search step length of the artificial fish swarm algorithm,and the improved artificial fish swarm algorithm is used to optimize the neural network.Secondly,the least-squares trend analysis is used to eliminate the trend terms of the three kinds of fault vibration signals of rolling bearing inner ring,outer ring and ball.And the mean value,standard deviation and crest factor are selected according to the change trend of the characteristic parameters in the time-frequen-cy domain.These three can obviously reflect the characteristic parameters of different fault types.Finally,the genetic algorithm,particle swarm optimization and other optimized neural networks are used as comparison algorithms for rolling bearing fault diag-nosis.The simulation results show that the proposed in this paper has higher average diagnosis accuracy,less error and higher sta-bility than the comparison algorithm.
Rolling BearingFault DiagnosisArtificial Fish Swarm AlgorithmNeural NetworkCharacteristic Pa-rameters
NETL
NSTL
万方数据
维普
