Convolutional Neural Network-based Fault Identification for Large Cooling Tower Fans
The conventional fault identification method of large cooling tower fan can only calculate the vibration speed fault parameters,and the fault frequency domain is less than 0.Therefore a fault identification method of large cooling tower fan based on convolutional neural network is designed in this paper.First the database of cooling tower fan is con-structed to realize real-time monitoring and storage of fan operation data.The input data constraint,network structure constraint and training process constraint are set based on convolutional neural network to improve its performance and generalization ability.Consequently cooling tower fan fault parameters,such as vibration speed,acceleration,frequency components,etc.,are calculated.Finally the fault frequency domain of cooling tower fan is identified,and the type and de-gree of fault are judged by analyzing the spectrum diagram of vibration signal.The experimental results show that the de-signed fault identification method can achieve fault frequency domain of cooling tower fans higher than the limit of 0,which proves that the proposed method is correct and has better accuracy for the fault identification of cooling tower fans.
万方数据
维普
