Multi-Objective Optimization of Ultrasonic Surface Rolling Process Parameters for TC4 Titanium Alloy with IWOA–RBF and MOGWO Algorithms

A structured optimization approach was applied to ultrasonic surface rolling process (USRP) parameters, aiming to enhance the material surface characteristics of TC4 titanium alloy. To overcome the premature convergence and limited exploration capability of the standard Whale Optimization Algorithm (WOA), three enhancement strategies were introduced, including population initialization based on an optimal point set, a sinusoidal nonlinear convergence factor, and an adaptive inertia-based position update strategy. By optimizing the structural parameters of the RBF neural network with the improved WOA, an IWOA-RBF predictive model for surface performance evaluation was developed and rigorously validated in terms of prediction accuracy. Using the developed IWOA-RBF model, a multi-criteria decision-making framework integrating the CRITIC weighting method and the TOPSIS ranking approach was constructed to evaluate surface quality. This framework was further combined with a multi-objective Grey Wolf Optimization (MOGWO) algorithm to perform Pareto-based optimization and determine the optimal USRP parameter set. Experimental validation showed that the optimized parameters resulted in a significant reduction in surface roughness, while enhancing both surface hardness and residual compressive stress. The results confirm the robustness and effectiveness of the proposed IWOA-RBF and MOGWO optimization framework, providing a reliable strategy for high-precision parameter optimization and coordinated enhancement of surface properties in the TC4 titanium alloy USRP.