Wear Condition Monitoring Research Of Ball Screw Based On Current Signal

CNC machine tools is the key equipment in manufacturing industry, the ball screw is a key component of CNC machine tools, and its performance directly affects the processing performance of machine tools processing quality and production efficiency, so researching the ball screw wear to ensure the machine processing performance, improve processing quality and production efficiency has important practical significance. In this paper, based on the actual production needs, a ball screw wear condition monitoring method based on current signal is proposed.Firstly, the worn reason of ball screw is analyzed, the performance after the ball screw is worn. The research idea of ball screw wear condition monitoring is determined.The advantages and disadvantages of various wear condition monitoring method in the application are researched and analyzed, aim at the ball screw wear characteristic, the actual situation and easy to realizing the ball screw wear condition monitoring technology engineering applications, the current signal monitoring method is determined as the main method, assist vibration signals and machine tool servo signal monitoring method. The principle of ball screw wear condition monitoring based on current signals is analyzed, a ball screw wear monitoring experiment platform is built.The methods that extracting features from current signal in fast-forward and cutting stages in the time domain and frequency domain are analyzed. Using wavelet packet decomposition decomposes current signal into different frequency bands, the sensitive characteristics for the wear conditions of ball screw from the wavelet packet components is extracted, a characteristic vector that reflecting the wear state of the ball screw is established.Application the support vector machine method to identify the wear condition of the ball screw, the appropriate kernel function is chose and the parameters of kernel function is optimized, the SVM ball screw wear condition identification model is established, the study and training of the model are completed, the identification result of the model is validated.