Optimization design of the key parameters of McPherson suspension systems using generalized multi-dimension adaptive learning particle swarm optimization

Abstract

The performance parameters of suspension systems must be properly matched to ensure the handling and stability performance of a vehicle. Based on real vehicle measured data, a parameterized vehicle dynamic model is built, and the validity of the parameterized vehicle dynamic model is verified by comparing simulation results with real vehicle test results. Seven representative steady-state and transient single evaluation indicators of handling and stability of the vehicle are selected. The key parameters of McPherson suspension system, which significantly affects steady-state and transient handling and stability performance, are selected through a sensitivity analysis. Their contribution rates for each single evaluation indicator are calculated based on 81 simulation tests using the parameterized vehicle dynamic model. A comprehensive evaluation indicator system for the whole vehicle is established. This system contains the seven steady-state and transient single handling and stability evaluation indicators that are obtained using a quadratic response surface fitting for the selected key parameters. The comprehensive evaluation indicator system is used to show whether a vehicle has good steady-state and desirable transient responses. Moreover, a generalized multi-dimension adaptive learning particle swarm optimization is proposed to search for the global optimum of the comprehensive evaluation indicator system across the search space with rapid convergence. Optimization results show that a comprehensive handling and stability performance are improved, and simulation results of the parameterized vehicle dynamic model that is modified in accordance with the optimization results verify the improvement of the steady-state steering driving behavior and transient yaw response of the vehicle. In conclusion, the comprehensive evaluation indicator system is feasible, and the generalized multi-dimension adaptive learning particle swarm optimization is effective for the optimization design of the key parameters of the McPherson suspension system.