A hierarchical QP optimization control strategy for quadruped robots facing a variety of structural terrains

Addressing the dynamic stability issue of quadruped robots with a relatively high proportion of leg mass in various structured terrains, this paper proposes a hybrid control strategy. Initially, the gait scheduling framework is constructed using Hopf oscillators within the Central Pattern Generator (CPG) to emulate the bio-inspired characteristics of quadruped robot locomotion. Subsequently, terrain estimation and gait planning are employed to enable the robot to traverse diverse structured terrains. During the stance phase of quadruped robot locomotion, a hierarchical quadratic programming (QP) optimization control is implemented. In the swing phase, a force-based impedance control is adopted to enhance the compliance and smoothness of the robot's leg-lifting motion. Furthermore, by leveraging Lyapunov stability theory, the stability of the control system in the swing phase is rigorously analyzed and proven, thereby ensuring the effectiveness and reliability of the control strategy from a theoretical standpoint. For posture control, dynamic stability is achieved by adjusting the linear and angular accelerations of the robot's center of mass. The proposed control strategy was validated through the MATLAB/Simulink simulation platform. The results demonstrate that, compared with the standard QP optimization control, the proposed strategy can more accurately and stably control the quadruped robot's locomotion on flat ground. Moreover, the strategy enables the robot to achieve stable walking on various structured terrains.