State Monitoring System Of CNC Machine Based On Multi-sensor Information Fusion

Equipment manufacturing industry has already become the pillar industry of our country,which can reflect the overall industrial level of our country to a certain extent.CNC equipment is the key equipment in manufacturing industry.In the process of CNC machine processing,the monitoring of working state is a complex process,and accurate on-line monitoring is difficult.CNC machines need to work under different processing conditions,so the state is constantly changing.Industrial statistics show that the downtime caused by machine failure accounts for 10%to 15%of the total downtime of CNC machines.When the machine is equipped with state monitoring system,it can reduce the downtime of failure by 75%,and increase the productivity by 10%-60%,increase the utilization rate of machine tool by more than 30%.Therefore,it is of great practical significance to research and develop an intelligent state monitoring system for CNC machine to prevent scrap of work pieces,equipment damage and ensure the machine’s trouble-free operation.In this paper,the structure and working principle of CNC machine are summarized,and the working state signals and the sensors of collecting signals are analyzed.It is decided that the state of CNC machine can be monitored by multi-sensor and multi-signal fusion.Multi-sensors have the characteristics of operation.They can acquire more effective and valuable information from the observed information thr-ough a variety of sensors so as to achieve the goal of fusion.In this paper,a multi-sensor fusion technology based on wavelet analysis and neural network is applied to machine tool state monitoring system.This paper introduces the theoretical basis of wavelet analysis and neural network.And introduce the composition of the state monitoring system.At last a complete software and hardware system for on-line monitoring of CNC machine tools is establishedWavelet transform decomposition can effectively extract the singularity feature of the signal,and feature vectors can also be effectively distinguished by neural network fusion.After testing the hardware and software system,the multi-sensor fusion technology based on wavelet analysis and neural network can effectively identify the working state of CNC machine tools.