During the deep hole drilling process, the tool condition monitoring is the key technique in the machining process.It is in favor of the security of the machining process. When the tools is wearing,the motor power of the NC Machine will change. With different wear amounts, the change of power is different.According to this phenomenon we propose tool condition monitoring system based on power, In this paper,we first establish the power datas acquisition system. According to the characteristics of the840D system, we can collect collecting power datas from the PLC of the machine control system through D2K-DASK data acquisition card, so we can obtain the power datas in different conditions. Then we analyse the characteristics of the collected power signals, and use wavelet analysis theory to denoise the power signals.And we use Mallat algorithm to do decomposition,reconstruction and feature extraction to the power signals. Finally, according to the feature extraction, we build tool condition monitoring system based on RBF neural network to identify the tool wear state. |