Micro Milling Tool Wear State Monitoring And Process Optimization Via Hidden Markov Model

Micro milling refers to the milling method in which the milling tool is less than one millimeter and the characteristic dimension of the machined workpiece is between one micron and one millimeter.Due to the high machining precision,high machining efficiency and the ability to process complex 3D surfaces,micro milling has been widely adopted to manufacture micro devices in the fields of aerospace,defense industry and biomedical engineering.Tool wear is unavoidable in micro milling.Tool wear will inevitably increase the surface roughness of workpiece and reduce the accuracy of micro milling.Severe wear can cause tool breakage,fracture and chatter,damage to the machined workpiece and machine tool.Compared with traditional milling,the size of micro milling tool decreases rapidly and the spindle speed increases greatly.The wear of micro milling tool is very fast and very difficult to monitor.Tool wear has become one of the main factors that restrict the development and application of micro milling technology.Effective tool wear monitoring is of great significance for improving micro milling accuracy and reducing processing cost.Starting from the cutting force model of micro milling,this paper analyzes the influence of the tool wear on the cutting force and extracts the tool wear feature from the cutting force.Then,the hidden semi Markov model is used to describe the tool wear process of micro milling and the online monitoring of tool wear and tool remaining life is realized.Finally,according to the monitored tool wear information,the Markov decision model is applied to dynamically optimize the micro milling mode to achieve the maximum cutting benefit.The main contents and innovations of the full text are as follows:(1)Extracting cutting condition-independent features that can infer tool wear under variable cutting conditions.Based on the mechanical model of micro milling,the influence of tool wear and cutting conditions on cutting force is researched.Then the feature that can reflect the micro milling tool wear is extracted from the cutting force signal,and the influence of the cutting parameters on the wear feature is eliminated.Finally,the cutting condition-independent wear features are extracted.(2)Considering the correlation between the durations of the tool wear states under the fixed cutting parameters,the tool wear process is built as a hidden semi Markov model with dependent durations.Based on the established wear model,an effective online estimation method for tool wear condition and remaining useful life of micro milling is proposed.(3)A switching hidden semi Markov model is established to describe the tool wear process of micro milling under variable cutting conditions,and the accurate on-line estimation of tool wear and remaining useful life of micro milling under variable cutting conditions is realized.(4)Considering the cutting benefit and the cost of the micro milling tool,the micro milling tool wear is established as a Markov decision model.The Markov decision model is used to dynamically adjust the cutting parameters according to the monitored tool wear so that the micro milling tool can be fully utilized within its limited service life,thus the maximum cutting benefit is obtained.