Multi-Information Fusion Online Recognition of Diamond Tool Grinding Direction Based on IPSO-RBF

In order to improve the online recognition accuracy of the grinding direction of single crystal diamond tools and thereby enhance the efficiency of single crystal diamond tool grinding, a method based on multi-information fusion and improved particle swarm optimization radial basis function (IPSO-RBF) neural network is proposed to online recognize the tool’s grinding direction. By collecting the vibration signals and acoustic emission (AE) signals of the tool during the grinding process, feature parameters are extracted through signal feature processing. The feature parameters are used as inputs to the IPSO-RBF neural network model for fusion, and the grinding direction of the tool is recognized online. Compared with the drawbacks of the slow convergence speed and easy trapping in local optima of the back propagation (BP) neural network, the IPSO-RBF neural network has a simple structure and fast training. Simulation results show that compared with the BP neural network and the standard RBF neural network, the IPSO-RBF neural network has higher recognition accuracy, faster convergence speed, shorter simulation time, and effectively improves the recognition accuracy of the tool’s grinding direction.