Abstract: In order to improve performance of passenger car rollover prevention, the robust control strategy with multi-population genetic optimization is presented. Taking the roll casting, roll steering, and the coupling relationship between vertical motion and roll motion into consideration, a six degrees of freedom model is established on the rollover dynamic theory of passenger car. As a control target, the maximum lateral-load transfer ratio(LTR) is applied to the robust control strategy for passenger car rollover prevention with differential braking. To enhance the anti-disturbance capability, the multi-population genetic methodology is used to optimize the weight function of controller. Some characteristics of the control strategy are discussed by numerical cases, such as the adaptive ability of the strategy in two typical driving conditions including J-Turn and Worst-Case, and the robustness of the strategy under the front wheel steering interference and pavement interference, as well as the sprung mass and speed variation. The results show that the critical velocity can be increased by over 75%, and the strategy is successful with good robustness at various driving conditions, external disturbances and parameter perturbations.

Key words: multi-population genetic optimization, passenger car rollover prevention, robust control, vehicle active safety