Optimization of tool edge preparation using magnetoelastic abrasives and development of a wear prediction model based on logistic chaotic mapping and sparrow algorithm

Edge preparation enhances tool life, cutting process stability, and surface quality. The tool was treated with magnetoelastic abrasives to optimize edge preparation. An orthogonal experiment was conducted to investigate the effects of abrasive grain size, tool speed, preparation time, disk spacing, and disk speed on the edge preparation amount. Milling experiments on TC4 titanium alloy were conducted to study the effect of shape factor K on tool wear. Preparation parameters have different effects on the rake face, flank face, and shape factor K. The tool cutting force information and the corresponding wear influence value, including the tool wear information, were collected and processed. The time domain, frequency domain, and time–frequency domain characteristics of the wear values and cutting forces were used as the dataset. A tool wear prediction model (LCSSA-BP) based on a sparrow algorithm optimized BP neural network with logistic chaotic mapping is developed and compared with the sparrow algorithm optimized BP neural network prediction model (SSA-BP). The predicted results of the LCSSA-BP model were closer to the experimental wear values than those of the SSA-BP model, SSAE + BPNN model, SVR model, and BPNN model. Additionally, the LCSSA-BP model exhibited smaller root mean square and absolute mean errors, confirming its superiority.