Multiple fault signal monitoring of ship electromechanical system under multi-sensor fusion
In order to improve the efficiency of ship maintenance,a multi-sensor fusion based monitoring method for multiple fault signals in ship electromechanical systems is proposed.Based on the frequency of the signal in the fault state,the wavelet transform method is used to extract the characteristic parameters of the fault signal as input for the ant colony al-gorithm to optimize the BP neural network,achieve multi fault diagnosis,and complete multi-sensor data fusion through DS evidence theory to obtain the fault diagnosis results,realizing the monitoring of multi fault signals in the ship's electromech-anical system.The experimental results show that this method can determine the type of faults in ship electromechanical sys-tems through multi-sensor fusion.Even if one sensor fails,it does not affect the diagnostic effect.The average accuracy of diagnosing multiple faults in ship electromechanical systems is as high as 97.02％,which can achieve more accurate monitor-ing of multiple faults in ship electromechanical systems.
multi-sensor fusionship electromechanical systemfault monitoringwavelet transformant colony algorithmDS evidence theory
NETL
NSTL
万方数据
