A multi-body vehicle dynamic model was built using ADAMS, which was then reduced to a 15 degree of freedom(dof) nonlinear model. Using the 2 dof linear model, a PID control strategy was created, and was then integrated and co-simulated with the 15 dof nonlinear model by the interface of ADAMS/Control. A simulation on ice road with single sine steering angle input was run. The control parameters were optimized and analyzed by a combined optimization algorithm with Adaptive Simulated Annealing (ASA) and Nonlinear Programming Quadratic Line search(NLPQL) method. The simulation results show that the handling stability and safety of the vehicle can be enhanced by the control method, which can be adapted to complex road and driving conditions.