Research On Walking Of Hydraulically Driven Quadruped Robot Based On Virtual Model And Impedance Control

Walking robot with foot has strong movement flexibility and environmental adaptability and can relatively easily enter complicated environments such as earthquake scenes,mountains,and jungles.As an important branch of walking robots,quadruped robots have broad applications in many fields such as military,commerce,entertainment,etc,thus receive widespread attention from the majority of scholars and have been greatly developed in recent years.In particular,the launch of the Big Dog robot by Boston Dynamics Corporation pushed the theoretical research of the quadruped robot to a new level.Control technology of quadruped robot is an emerging area,in which solid technical reserves is very important,and there is still a big gap between China and the developed countries such as the United States.The existing quadruped robots in China have relatively weak ability to resist disturbance walking dynamically on uneven pavement.Therefore,based on the existing problems,this paper proposes and constructs a layered control framework for the four-legged robot to improve its robustness.This paper refers to the classic quadruped robot design structures at domestic and abroad and builds a quadruped robot model.The kinematics of the robot is analyzed using the D-H convention to obtain the corresponding relationship between the foot trajectory and the joint angle.The Lagrange equation is used to analyze the dynamics of a single leg of the robot,which lays a theoretical foundation for the inverse dynamics compensation required by applying flexibility control.In the study and simulation of the foot-end trajectory planning,it is found that the trajectory obtained by polynomial interpolation has poor flexibility,and is unable to autonomously modified according to environmental changes.Therefore,in this paper,a non-linear oscillator is used to construct a pattern generator network to generate a robot foot trajectory,and a coupling item is added to the network to make the trajectory adaptive.For the study of control methods,this paper first uses the position control algorithm to control the dynamic walking of the quadruped robot.In the simulation experiment,it was found that the robot can walk dynamically on a flat road,but has weak ability to overcome external disturbances when walking on uneven roads and is easy to fall.The main reason is that the robot based on simple position control does not have flexibility on the foot,and the instantaneous contact force between the foot and the ground impair the stability of the body seriously.In order to enhance the flexibility of the robot in the contact with the environment,this paper uses force-based impedance control to achieve the robot's foot-end flexibility control and validates through the single-leg test.At the same time,the foot trajectory can be effectively tracked when the leg is swinging.However,simply using impedance control cannot control the posture of the robot body,thus the virtual model control method is applied to control the posture of the robot body to form a hierarchical control framework for the robot.In order to verify the effectiveness of the hierarchical control framework proposed in this study,a virtual prototype model of the robot was established,and a dynamic walking experiment with the self-disturbance from head and tail and external disturbance of uneven road surface was performed.The simulation results show that the robot based on the layered control framework of this paper can effectively control robot to overcome the disturbance generated by the outside world and itself,and achieve stable dynamic walking.