A new method based on a WOA-optimized support vector machine to predict the tool wear

Tool wear has been a great impact on machining quality and machining efficiency during cutting. The serious tool wear will even lead to workpiece failure and catastrophic equipment failure. Accurate and effective tool wear monitoring is important to evaluate the degree of tool wear, replace tools in time, and promote the intelligent development of the manufacturing industry. To improve the accuracy of online prediction of tool wear, a new method based on whale optimization algorithm (WOA) optimized support vector machine (SVM) is proposed to predict the tool wear. Specifically, the multi-domain features of cutting force and vibration signals are extracted based on the time domain, frequency domain, and time–frequency domain, and the signal sensitive features closely related to tool wear are selected by the Pearson correlation coefficient method. SVM is applied to predict the evolution of tool wear. WOA is used to improve prediction accuracy by optimizing the internal parameters of SVM. By learning the nonlinear correlation between sensitive features and tool wear, a model for predicting tool wear based on WOA-SVM is constructed to predict the change of tool wear value. The effectiveness and prediction performance of the proposed method are verified by milling experiments. Results show that this method can predict tool wear value based on limited historical data information accurately and effectively. Compared with SVM prediction methods optimized by some common optimization algorithms (particle swarm optimization (PSO) and genetic algorithm (GA)), the prediction accuracy and stability are higher and the generalization is stronger. These findings may be of great significance for the improvement of machining quality and efficiency of parts, the stable operation of manufacturing system, and the intelligent development of manufacturing industry.