Abstract: Vehicle dynamics equilibrium is the key parameter in nonlinear dynamics analysis of vehicle stability. Due to nonlinear features of vehicle dynamics, the traditional analytic method, gradient numerical method and genetic algorithm(GA) can hardly work effectively. A hybrid method is proposed for vehicle dynamics equilibrium search, which combines real-coded genetic algorithm with Quasi-Newton(QN) gradient method. The hybrid method is verified by equilibriums solution search of Duffing equation. Based on the magic tyre formula, a three degrees of freedom(3DOF) nonlinear mechanical vehicle model is founded. Numerical experiments with variable front steering angles are carried out to compare proposed GA+QN with conventional GA on 3DOF vehicle dynamics equilibrium solutions. Numerical results show that the hybrid method can find optimal solutions with higher accuracy and efficiency in solution search for 3DOF vehicle dynamics equilibriums.

Key words: vehicle dynamics equilibrium;genetic algorithm;quasi-Newton gradient method;hybrid method