The Multi-Objective Optimization Design of Hard Point Parameters for Double Wishbone Independent Suspension 2023-01-0127

There are often a large number of design variables and responses in suspension hard point optimization design. The traditional optimization strategy integrating heuristic algorithm and simulation model is not applicable due to its low efficiency. To solve optimization problems with huge number of design variables and responses, a multi-objective optimization framework combined heuristic optimization algorithm with multi-objective decision-making method is developed. Specifically, the multi-objective optimization was performed by dividing the problem into two independent sub-problems of multi-objective optimization and multi-objective decision-making. Further, to reduce the number of sample points required for building a surrogate model, a two-stage multi-objective optimization is proposed. In the first stage, the initial optimal solution is obtained based on the experimental design (DOE), and the influence of each design variable on each response is obtained through sensitivity analysis; the second stage trains the mixed surrogate model around the initial optimal solution, and then performs multi-objective particle swarm optimization design based on the mixed surrogate model, and finally obtains a satisfactory Pare Frontier. Furthermore, to overcome the limitations brought by a single multi-objective decision-making method, a multi-objective decision-making method developed by combining a weight strategy inspired by grey relational analysis and entropy analysis was applied to determine multiple objectives weights from the Pareto frontier. Finally, the pseudo-damage of each connection point of the suspension based on the hard point parameters optimized by the proposed optimization strategy is better than the original suspension, i.e., the pseudo-damage of each connection point is improved by 21.35%, 35.96%, 12.04%, 14.65%, 25.28%, 14.31%, 16.35%, 20.76%, 12.09%, 11.41%, 14.63%, 11.02%, 33.92%, 33.98%, 47.48%, 85.09%, respectively.