Prediction of work hardening and energy consumption for blade surface milling based on TLBO–GEP

The milling of Al7050 alloy blades involves a strong nonlinear mapping between cutting parameters and responses (work hardening and energy consumption).The accuracy is relatively low when using traditional models for processing. To address this, an improved TLBO–GEP algorithm is proposed, which adds a weight domain to Gene Expression Programming (GEP) chromosomes and optimizes it using Teaching–Learning-Based Optimization (TLBO). Benchmark tests verify its superior convergence and predictive performance. TLBO–GEP algorithm uses the experimental data collected from blade surface milling, with spindle speed, feed per tooth and step depth as inputs, work hardening (microhardness) and energy consumption as outputs, and establishes an explicit mathematical model. The experimental results demonstrate that the prediction errors for work hardening and energy consumption are 0.471 and 2.339%, respectively, significantly outperforming the basic GEP. This high-precision, explicit model provides an effective tool for blade milling parameter prediction and process optimization. In actual production, it can be integrated into processing systems to improve surface quality and reduce production energy consumption.