Abstract: To enhance the motion performance and flexibility of quadruped robots, inspired by the biological characteristic of quadruped animals that increase their movement ability by twisting their waists during high-speed movement, this paper addresses the common problem of the lack of lumbar joint structure in existing robots by proposing a new strategy that combines central pattern generator (CPG) control and active lumbar joint control. Firstly, the simulation model of the quadruped robot with a lumbar joint is established in MATLAB/Simulink, and the characteristic of its trot gait is analyzed. Secondly, the CPG control strategy is adopted to design the control network based on the Hopf oscillator to achieve the cooperative control of the four legs of the quadruped robot. Then, a lumbar joint active control strategy is further proposed to combine with the CPG control strategy to optimize the motion control effect of the quadruped robot. The simulation results show that, compared with using only CPG control, when combined with active control of the lumbar joint, the average speed of the quadruped robot can be increased by up to 39.21%, and the optimization rate of unit distance power consumption reaches 10.57%. This effectively verifies the effectiveness and superiority of the proposed strategy in improving motion performance and energy efficiency.