Research On The Movement Control Of Wheel-legged Bionic Robot

The wheel-legged robot combines the leg robot with the wheel robot.It has the advantages of wheeled robots and flexible humanoid or biomimetic characteristics of legged robots,and has great potential application value in service,transportation and military.The wheel-legged robot has become a hot field in robotic system and technical research.This paper takes the movement of a kind of electro-hydraulic hybrid driving two-wheel self-balancing wheel-legged robot as the core,and the research work of mobile control is carried out in the following aspects.System modeling of wheel-legged robots.Firstly,the kinematic model between the joint angle and the joint position coordinates is established,according to the robot kinematics model and the joint mechanism of the prototype,the relationship between the hydraulic cylinder displacement and the joint angle change is set up.In order to create the overall motion control scheme of the robot,starting from the basic principle of balance control,the equivalent system modeling of the wheel-legged robot is carried out by decoupling and equivalent realization,and the equivalent centroid model and model mapping relationship are obtained,and the Kalman filter state estimation model of the robot hip joint and the equivalent centroid is generated based on structural design of onelegged robot.The transmission scheme of hips and knees based on IMU and joint displacement sensor,this part lays the foundation for the model needed for mobile control.The kinematics and dynamics modeling.Using jumping height as an indicator,the positive and opposite calculation relationship between the foot end position and joint angles are established and linear fitting of hydraulic cylinder strobe is carried out,taking the joint angle as independent variable.Therefore,complete decoupling of robot gesture is realized.Dynamical model based on Lagrange-method is structured and kinetic equation of landing and flying periods is integrated.Also,linear fitting relation is established between the joint torque and joint angle.Based on this,law of function of the joint torque and the influence of joint torque brought by ground shock.Linear Variable Parameter(LPV)dynamics system model of robot establishment and control algorithm design.According to the idea of equivalent centroid,the dynamics of the robot can be simplified into a complex model of nonlinear coupled time-varying parameters in the drive wheel control system,this system is modeled by LPV theory and technology,then the tensor's High-Order Singular Value Decomposition(HOSVD)performs multi-cell transformation and the system is described in the form of vertex system.Based on the specific control objectives,the H-infinity mixed D-stable gain scheduling control algorithm is designed,and the Linear Matrix Inequality(LMI)is used to obtain the controller to ensure the drive wheel control system's response and tracking robustness.The Center Of Mass(COM)control strategy.For the equivalent self-balancing robot system,the corresponding centroid control strategy and control rate are designed respectively on the dynamic performance and static stability balance.For the realization of dynamic acceleration and deceleration,the preview control is used to plan the trajectory of the COM.As for static balance,this paper proposes an algorithm that searches for joint angles to optimize the adjustment time to achieve optimal control of centroid control.Simulation and experiment of wheel-legged robot movement control.Matlab numerical simulation and physics engine simulation combined with V-rep and Matlab are carried out for the contents above.The algorithms were verified and the parameters are set.This paper set up the actual physical prototype as well as control system,and the friction identification and compensation experiments of the drive wheel motor are executed to ensure the accuracy of torque control;combined with the above theory and simulation results,the experiments of lifting,Acceleration and trajectory tracking are finished to verify the rationality of the scheme and algorithm.