Abstract: In order to get better performance of semi-active suspension and ride comfort of tracked vehicle, Linear Quadratic Regulator (LQR) algorithm for semi-active suspension was studied. Firstly, based on the half vehicle dynamics model with seven degrees of freedom, a Linear Quadratic Regulator algorithm was designed for tracked chassis. And then, taking random road roughness as the excitant, simulation experiments of passive suspension and semi-active suspension under the control of SGHH and LQR were carried out to get the time responses and the frequency responses in different control strategies. The comparison experiment results show that, Linear Quadratic Regulator algorithm can not only reduce the acceleration, speed and deformation of weight bearing wheels to different degrees but improve ride comfort, satisfy the need of stability of shock absorber and weight bearing wheels adhesion performance of tracked chassis.