Overview of Research on Health Assessment and Fault Prediction of Complex Equipment Driven by Big Data

Abstract

Health evaluation is an important content of health management for complex equipment, and fault prediction is a key link to realize condition-based maintenance. Based on the literature review, this paper gives a feasible modeling method: The health status of complex equipment is expressed as time series by constructing forward and backward LSTM model, and the health evaluation model is constructed and trained based on the deep learning platform TensorFlow. The improved particle filter constraint algorithm is used to optimize LSTM neural network which is to generate PF-LSTM prediction model. The research on health evaluation and fault prediction of complex equipment based on machine learning discussed in this paper puts forward new requirements and development direction for machine learning of complex equipment, and is of great significance for practical implementation of lean intelligent manufacturing, improvement of stability and economic benefits of complex equipment, and reduction of operation and maintenance costs.