Design and Application of Fault Diagnosis System of Machine Tool Angle Head Based on Boolean Matrix Filtering and Optimized BP Neural Network

Abstract

A fault diagnosis method based on Boolean matrix filtering and optimizing back propagation (BP) neural network is proposed for angle head of computer numerical control (CNC) machine tools in the paper. The matrix filtering is firstly carried out with the fault case database and the fault cause symptom Boolean matrix according to the fault types and characteristics of machine tool angle head. On the basis of the combination of multiple fault causes obtained from the initial filtering, the Euclidean distance method is used to narrow the results of fault causes filtering. The BP neural network model with weight vector is established and optimized to perform the accurate diagnosis. Finally, the fault diagnosis and management system of angle head of CNC machine tool integrating with Boolean matrix filtering method, Euclidean distance method and BP neural network model is developed and implemented with Python language and Qt development framework.