Milling cutter wear state recognition based on deep belief network and SVM
To the issue that manually extracted wear indicators cannot entirely express milling wear characteristics,a tool wear recognition model based on improved deep belief network(IDBN)and support vector machine(SVM)is proposed.Firstly,the characteristics of cutting force,vibration and AE signal in time domain,frequency domain and time-frequency domain are extracted.Secondly,improved deep belief network is used to reduce the dimensionality of extracted features.Finally,the improved seagull optimization algorithm was used to realize the tool wear state recognition model by optimizing support vector machine.The experimental results show that after 100 random stratified sampling,the average recognition rate of IDBN-ISOA-SVM for tool wear is more than 99%.Compared with other algorithms,this model can accurately identify the wear state of the milling cutter from three aspects:dimensionality reduction method,optimization algorithm and classification model.
wear state identificationDBNseagull optimization algorithmSVM
万方数据
维普
