Development And Data Processing Of Cutting Process Monitoring System Based On Multi-sensor Fusion

Intelligentization is the developing trend of manufacturing automation.Artificial intelligence technology is widely applied in almost all aspects of the manufacturing process.The foundation of intelligence is the acquisition and processing of information.The real-time physical information in machine tool processing is one of the basic information for realizing intelligence,and it is also an important step to achieve intellectualization by analyzing this imformation.The real time information in machine tool processing is collected through sensor system.Based on the problems of complex structure,single acquisition signal and huge amount of data in the current in the information monitoring system of machine tool process,a portable monitoring platform for machine tools based on multi-sensor is proposed.And the data processing of real time machine tool physical information collected by the platform is processed,and the prediction model is established.On line monitoring of machine tool status by platform is realized.The main research results are as follows:(1)According to the physical information characteristics of cutting process,some kinds of information which reflect the most intuitive and effective cutting process are screened out.That is,cutting force,cutting current,tool vibration and cutting torque are taken as objects of system information collection.According to the characteristics of each signal,the most suitable sensors are selected to collect information efficiently and to compress the volume of the system as much as possible.The host and slave computer is designed,and the program is compiled after selecting the model of the computer,as to realize multi-channel data acquisition and human-machine interaction.Last the experimental of calibrated sensor is designed.(2)The feasibility analysis of the sensor fusion monitoring system is presented.In the process of cutting process signal acquisition,high precision sensor system is used to analyze measurement error.In order to study the influence of cutting parameter in cutting process,the cutting force,vibration and current are collected by the developed multi-sensor fusion system.Visual data analysis is accomplished by single factor and orthogonal experiments.In order to analyze the problem of tool wear and on-line monitoring of machined surface quality in cutting process,the monitoring system was used to collect the physical information in machining process.(3)The cutting signal was processed in time domain,frequency domain and wavelet analysis,and the principal components analysis method was used to reduce dimension.The genetic algorithm,the particle swarm optimization and the grid search algorithm were respectively used to optimize the support vector machine algorithm.The intelligent prediction models of the tool wear and surface quality based on support vector machine were established.At the same time,the relationship between cutting parameters and real time information of cutting is established,which is used as the basis for setting the threshold of the slave computer in the monitoring system.