Multi Variety And Small Batch Mechanical Processing Based On Recursive Analysis Process Status Monitoring

In recent years,the emergence of emerging information technologies such as big data,cloud computing,and the Internet of Things has provided new insights for the intelligent improvement of the manufacturing industry.As the main tool for mechanical processing,CNC machine tools can cause huge economic losses and resource waste if they fail and cannot be identified in a short period of time.Mechanical processing status monitoring is an important means to ensure the orderly completion of the processing process,and is a key research topic in the field of intelligent manufacturing.With the increasingly fierce market competition,customer demand is developing towards diversification and personalization.In order to improve the response speed to customer needs,the production mode of enterprises is gradually shifting towards multi variety and small batch production.Multi variety small batch production refers to a production method that produces a large variety of products within a specified production period,but each type of product has a smaller production quantity.According to the characteristics of multi variety and small batch production,this paper proposes a method for monitoring the state of machining process under the multi variety and small batch production mode based on dynamic time warping(DTW)and Cross Recurrence Plots(CRP).Firstly,the principle of phase space reconstruction and the basis of recursive analysis are studied,and the effectiveness of recursive analysis in processing power signals in machining process is verified.Secondly,this paper studies the operation characteristics of multi variety and small batch production mode,and establishes a condition monitoring model based on cross recursive analysis and DBSCAN algorithm:(1)Take a small number of workpieces for trial processing according to the processing plan,collect a small number of trial processing power signals of the processed workpieces,use wavelet decomposition to generate two wavelet characteristics,and then use dynamic time warping and K-means clustering to generate template database;(2)The real-time power signal is input into the dynamic time warping recognizer to identify the workpiece;(3)Based on the recognition results of workpiece types,the real-time processing power of the workpiece is generated into CRP with the same class of trial processing template power,and recursive quantitative analysis is conducted.Based on the recursive quantitative analysis results,the density based spatial clustering of applications with noise(DBSCAN)algorithm is used to monitor the real-time processing status of the workpiece.Then,based on the SVM abnormal state classification recognition model,the abnormal state is identified and classified.Finally,this article designs experiments based on the multi variety and small batch production mode,and verifies the effectiveness of the state monitoring model through experiments.