四足机器人越障姿态估计算法研究OA

To address the issues of insufficient motion stability and excessive energy consumption during obstacle cross-ing by quadruped robots in complex and rugged terrains,this paper proposes an adaptive posture estimation method based on body perception,combined with model predictive control and whole-body control algorithms,to enhance the motion stability of quadruped robots on uneven terrains.Firstly,by optimizing the foot-end trajectories,the robot can adapt to various terrains,achieving smoother and faster foot-end movements,and enhancing the obstacle-crossing stability in dif-ferent terrain environments.Secondly,a finite state machine is designed to dynamically switch the obstacle-crossing gaits based on the terrain posture estimation results,enabling the robot to consider terrain information during obstacle crossing,optimize its foot-end trajectories,and further improve the obstacle-crossing performance and stability.Finally,to verify the effectiveness of the proposed method,the algorithm is tested on a stair model with obvious ruggedness and unstruc-tured terrains.The results show that the robot's motion energy consumption is reduced by 9.78%during the stair climbing process and by 18.27%during the stair descending process,demonstrating that the proposed method can effectively adapt to rugged terrain environments and perform posture estimation,achieving low-energy and high-stability obstacle-crossing behaviors in such environments.