Adaptive Prediction Method For Tool Wear And Remaining Useful Life Of NC Machining Tool

The condition of the tool during NC machining has a very significant impact on the quality and efficiency of the machining,and the deterioration of the tool is inevitable during the machining process.Operators generally change tools when there is some margin in the tool guide life,but in areas such as aerospace and small batch production mode,which are difficult to process,the tool degradation rate is fast and changeable.The traditional tool change method will cause The occurrence of machining with severely degraded tools has caused unpredictable harm to the machining.In order to avoid the above situation,operators can only adopt a more conservative tool change strategy,which has caused great waste.Under the premise of ensuring the processing quality,to fully tap the value of the tool,it is necessary to establish a reliable tool life evaluation standard and method.The international standard takes the amount of flank wear as a criterion for evaluating the degree of tool degradation.Therefore,the maximum wear band width of the flank is used as the evaluation standard of the tool condition.This paper proposes a self-adaptive prediction method of tool remaining useful life,which can accurately and quickly detect the tool through a microscope to obtain the label.Obtain signals such as cutting force,workpiece vibration,machine tool spindle power,and current,and preprocess the extracted signals.Wavelet packet decomposition is performed on the processed signals to extract features,and the extracted features are analyzed for correlation with tool wear and operating conditions,and features with strong correlation with tool wear and weak correlation with operating conditions are screened.The filtered features are input to Gaussian process regression to predict the maximum wear band width and remaining useful life of the tool flank.In this paper,the tool is detected by the tool in place.The studied microscope autofocus method and camera platform can accurately and quickly obtain the tool wear amount.The camera platform’s angular deviation during the serial measurement of the tool is controlled to 0.1° Within.In the case of no built-up edge,the maximum error of the predicted tool wear value does not exceed 0.03 mm,and the predicted maximum error of the tool remaining useful life does not exceed 187s.