Design And Control Of A High Speed Running And Jumping Underactuated Cheetah-like Quadruped Robot

The current research directions of quadruped robots are mainly divided into heavy load,high adaptability and flexibility,high speed.The representative research institutions are Big Dog and Wild Cat of Boston Dynamics.However,there are few research achievements in better combination with the body shape changes of the bionic object in the process of movement with high-speed.The robots are still equipped with the completely rigid body and there is a certain gap in bionic research.Based on the above research background,in this paper,the cheetah was chosen as bionic object,the mechanism of the spine was added.At the same time,in order that the control complexity of multimechanism could be reduced as much as possible,the robot was set to be jointly driven by the spine and legs,and the whole machine is underactuated.After the configuration was determined,the models of increasing the swing frequency of the legs and optimizing the configurations of the legs and spine were further established to complete the comparison and verification of the robot performance improvement effect.It may provide various research ideas and references for the high-speed research of bio-robot.First of all,the robot configuration was determined by referring to the actual skeleton composition of the biomimetic object cheetah and the movement of the legs and spine in its actual running.Then,the final leg parameters were determined through the trajectory optimization.For the goal that the whole robot is underactuated,the mechanism which was used to transfer the leg drive to the spine was determined,and the whole design of robot was finished.Then the kinematics and dynamics of the robot were analyzed and verified.In kinematic analysis,mainly aiming at solving the foot trajectory when the spine is rigid or deformable,two models were established to complete the verification of theoretical calculation of kinematics.After that,the dynamics analysis of the robot was carried out.The leg drive torque and the drive force of the linear drive in the spine were solved through the dynamic force analysis after the split of the bar group and Lagrange dynamics respectively.The dynamic theoretical results were also verified by the established models.In order that the methods of improving the speed of the robot can be better excavated,the comparisons were carried out by the improvement of leg swing frequency and the optimization of robot configuration.In the process,found that as the crank speed was increased gradually,the legs was disordered and led to the gait failure situation.Then try to adjust the leg angle and the initial phase difference of the active crank respectively to make the failure situation effective,combined with the leg sequence of different gaits,the models of these two adjustment methods were established,respectively the robot speed improvement effects were compared,the conclusions were drawn.In the attempt to improve the speed by configuration optimization,stiffness of leg elastic elements and spine deformation were configured for different gaits,and parameters representing the movement performances including speed,stability,impact and drive torque were compared.Different gaits combined with different configurations,movement performance changes were different.After comprehensive analysis,the most suitable configuration for each gait and the most suitable for the increase of the stride by adding the spine deformation was obtained.This may provide theoretical reference for the biorobot to achieve greater speed through configuration adjustment.