Adaptive control algorithm for quadruped robots in unknown high-slope terrain

• DOI: 10.1016/j.jer.2024.05.018
• Publication Date: 2024-05-27T16:39:15Z
• Authors: Peixing Li; Boer Yin; Lin Zhang; Yanzheng Zhao

The capability to navigate high-slope terrains is crucial for quadruped robots engaged in field operations. We propose a novel terrain estimation and adaptation strategy tailored facilitating robot locomotion on complex slope terrains. Our approach involves predicting slopes by analyzing foot positions IMU data, subsequently adjusting the robot's body orientation height real-time accommodate varying conditions. To effectively control motion of such challenging terrains, we introduce multimodal algorithm that integrates Model Predictive Control (MPC) with Quadratic Programming (QP) torque control. This combined ensures stable efficient even steep slopes. Furthermore, present state method based Kalman filtering, enabling accurate self-assessment without heavy reliance visual sensors. In simulation validation, our achieved notable performance metrics. It forward speed 0.7 m/s angles up 43∘ demonstrated rotational at 2 rad/s 32∘ slope. When subjected external force interference, exhibited resilience, withstanding constant forces 60Nm torques 35Nm flat ground. Moreover, 30∘ slope, maintained face impulses reaching 64Nm ⋅ s along x y directions its coordinate system 7Nm z-axis. comprehensive experimental validation highlight efficacy robustness adaptive algorithm, paving way enhanced navigating unpredictable characteristics.