Research On Adaptive Control Technology Of Quadruped Robot In Complex Environmen

In the past few years,quadruped robots have become one of the research fields that have received much attention in the field of robotics.With the continuous development of advanced materials,intelligent control,and environmental perception technology,the market research on quadruped robots is gradually deepening.The continuous progress of these technologies has also made it possible for quadruped robots to move from laboratories to daily life.The existing quadruped robots still have shortcomings in their trafficability when facing complex road conditions.In response to the above problems,this project proposes a stable control system for quadruped robots based on model predictive control,aiming to improve the trafficability of robots in complex environments.This article is based on CAN bus communication for the joint motor control of quadruped robots.In response to the parameter optimization problem of the joint motor of quadruped robots,the dynamic model of a single joint control system under motor load is first designed and the control strategy is planned.Through CAN bus communication,the control of the motor is achieved,and the position gain and speed gain of the joint motor are reasonably adjusted and optimized.A control method based on model predictive control is proposed for complex terrain and road surface.First,the kinematics of the quadruped robot is analyzed and a single rigid body dynamic model is established.According to the single moment state of the robot,the state of the next period of time is predicted,and the optimal plantar force is calculated and converted to the joint for output in the form of torque.Building a simulation system based on Webots,constructing complex environments such as gravel,stairs,and slopes in the Webots environment,and verifying model predictive control algorithms in various simulation environments.After completing the simulation experiment,based on the algorithm verification of the Unitree A1 platform,the Unitree A1 type quadruped robot was used to address the problem of physical machine algorithm testing.A1 was equipped with rich secondary development interfaces and connected to the Unitree A1 motion control board.Then,the robot was subjected to high-level and low-level control respectively.Finally,the code was ported and the entity verification and testing were completed.Through experimental analysis,it can be seen that the stability control scheme proposed in this article can smoothly pass through complex road surfaces such as gravel roads,unilateral bridges,and slopes,with a maximum speed of 2.1m/s and a stable up and down 6cm stairs,further meeting the work needs in complex environments.