Aiming at the shortcomings of subjectivity and low efficiency in setting the weight coefficient matrices Q and R of linear quadratic regulator (LQR) for vehicle active suspension systems, a strategy based on snake optimizer (SO) to optimize the weight coefficient matrices of LQR controller was proposed. The LQR controller was designed by analyzing the dynamics of the 1/4 vehicle active suspension system. The objective function L was constructed by weighting the integral ratios of each performance index of active suspension and passive suspension. The minimum value of the function L and the best weight coefficient matrices of the LQR controller in the search space was solved by SO which imitating the living habits of the snake group. In order to verify the effectiveness of this strategy, SO optimized LQR-controlled active suspension was simulated and compared with passive suspension, conventional LQR-controlled active suspension, genetic algorithm optimized LQR-controlled active suspension and particle swarm optimization optimized LQR-controlled active suspension in terms of sprung mass acceleration (SMA), dynamic tire load (DTL) and suspension working space (SWS), in which C-class road and sinusoidal bump road were used as excitations, respectively. The results show that the active suspension controlled by SO-optimized LQR can optimize the root mean square of SMA, DTL and SWS by 59.45%, 37.89% and 42.12% on the C-class road, respectively. The stability time on the sinusoidal bump road is 1.4 s, and the overshoot optimization of SMA, DTL and SWS is 79.22%, 59.22% and 16.33%, respectively, which improves the driving smoothness, road adhesion and operational safety of vehicles.