Tool wear prediction in milling based on a GSA-BP model with a multisensor fusion method

Tool wear damages the surface quality of the workpiece and increases equipment downtime. Tool wear prediction is of great importance for reducing processing costs and improving processing efficiency. This paper applies multisensor fusion technology to predict tool wear. The cutting force, vibration, and acoustic emission signals are collected simultaneously during the milling process. The time domain, frequency domain, and time–frequency domain characteristics of each signal are extracted, reduced, and filtered through correlation analysis. A GSA-BP prediction model is established by a BP neural network in which the weights and thresholds are optimized through the gravitational search algorithm (GSA). The test results show that the prediction results of the GSA-BP model are in good agreement with the actual wear, and the prediction accuracy is higher than that of the traditional BP neural network model.