The longitudinal acceleration control plays a very important role in the design of active safety-system of vehicle, which includes driving and braking control systems. In this paper, two different control laws are designed in the engine and brake controller. A fuzzy gain sheduling scheme of proportional-integral-derivative (PID) control law is used in the throttle control, and the parameters of the fuzzy inference system are optimized by particle swarm optimization algorithms. A hybrid feed-forward & feedback control law is applied in the brake control. The vehicle model is constructed for simulation in Matlab/Simulink to test the strategy of throttle controller, brake controller and the combination performance of the two controllers, respectively. The results show that the rapid response and high-precision performance have been achieved in the vehicle longitudinal acceleration control.