During centralizer hydroforming, internal pressure and axial feed critically influence the forming outcome. Insufficient feed causes excessive thinning and cracking, while excessive feed causes thickening and wrinkling. Achieving uniform wall thickness necessitates careful design of the pressure and feed curves. Using max/min wall thickness as objectives and key control points on these curves as variables, the study integrated Non-dominated Sorting Genetic Algorithm (NSGA-II), Multi-Objective Particle Swarm Optimization (MOPSO), Neighborhood Cultivation Genetic Algorithm (NCGA), and Archive-based Micro Genetic Algorithm (AMGA) with LS-DYNA to automatically optimize loading paths. The results demonstrate the following: ① NSGA-II, NCGA, and AMGA successfully generated optimized paths; ② NSGA-II and AMGA produced larger sets of higher-quality Pareto solutions; ③ AMGA required more iterations for satisfactory Pareto sets; ④ MOPSO exhibited a tendency towards premature convergence, yielding inferior results; ⑤ Multi-objective optimization efficiently generated diverse Pareto solutions, expanding the design space for process design.

The hydroforming performance of trailing arms is governed by the coupled effects of feed parameters, pressure schedules and frictional characteristics. Improper parameter matching readily induces typical forming defects such as wrinkling, cracking and uneven wall thickness. To address this issue, a multi-objective optimization method for hydroforming is proposed in this study. Taking the maximum wall thickness, minimum wall thickness and die-to-workpiece gap of the tubular blank as optimization objectives, and the internal pressure and right-side axial feed velocity as design variables, an integrated numerical simulation framework combining the Archive-based Micro Genetic Algorithm (AMGA) and LS-DYNA is established to analyze the hydroforming process. By adaptively adjusting the key control points of internal pressure and axial feed loading curves, the developed method expands the solution space and realizes the automatic optimization of loading paths. The results reveal that the maximum wall thinning rate of the tubular component drops from 20.4% to 14.8%. Meanwhile, the wall thickness uniformity is improved and forming defects are effectively suppressed while the thickening rate remains stable. Furthermore, a complete round of optimization calculation involving thousands of finite element solutions can yield a complete set of Pareto non-dominated solutions. In this paper, the AMGA multi-objective optimization algorithm is adopted to acquire the optimal loading paths, and physical prototype experiments are carried out relying on self-developed 2000 T hydroforming equipment. Comparisons between measured and simulated wall thickness values of the tubular component show that the maximum relative error is controlled within 7.46%, which verifies the reliable engineering applicability of the proposed optimization scheme and provides new insight into the process optimization for forming similar structural components.