In order to monitor the operation of elevator in real time,this paper puts forward a method of condition monitoring and fault prediction for key components of elevator based on data set.First,the elevator multi-source data set is designed,and the available parameters are determined by combining parameter trend and fault prediction logic.Then recurrent neural networks and bidirectional long short term memory neural networks are introduced,combined with attention mechanisms,to construct the model of condition monitoring and fault prediction,and a prediction system including the function of fault prediction and health evaluation is formed.Through the experiment,the designed condition monitoring and fault prediction system for key components of elevator can play the actual effect.
Data setElevatorCondition monitoringFault predictionKey components
NETL
NSTL
万方数据
