Research On The Control Method Of Spinal Quadruped Robot Bounding Based On ASLIP Model

Quadruped robot has impressive flexibility of locomotion and excellent adaptability to environment,which has made it a hot topic in robot research.With the deepening of research and the improvement of the robot's performance requirements,the research field of quadruped robot has divided into research branch with high-speed locomotion as the goal.Biological studies show that the bounding gait is a typical high-speed symmetrical gait in quadrupeds,besides,there exist the participants of spines during animals' high-speed locomotion.However,the relevant theoretical and locomotion control studies of spinal quadruped robot are rarely reported.Most of the current studies have isolated the spine,and there are few studies on the locomotion control methods of the whole machine,and most of them remain at the level of theoretical analysis.Hence,research on this field is bound to promote the development of bionic engineering and robot locomotion control.In addition,with its characteristics of high-speed,it will also have broad application prospects in areas such as military reconnaissance,rescue and relief,and future life.First of all,start with analyze the locomotion characteristics of the biological cheetah,a seven-rod model of quadruped robot and an ASLIP model for locomotion control are established.The dynamic equation of the model is deduced by the use of the Lagrangian equation;Calculate iteratively the dynamics differential equations,and use the Poincaré mapping method to search the fixed point of the robot seven-bar model's bounding locomotion that based on the ASLIP,the result shows that the fixed point is regionally distributed under the fixed energy level;The comparison of fixed points shows that the locomotion based on the ASLIP model adapts to higher steady horizontal speed than the locomotion based on the SLIP model.And analyzes the characteristics of the ground reaction force,angle of spines,and stability.It provides the model and theory for bounding locomotion of the spinal quadruped robot.Then,based on the generalized force calculation results,the locomotion control methods of horizontal speed,height,pitch angle,touchdown angle,angle of spines,and leg length are studied.The dynamic simulation results verify the effectiveness of the control methods;For the coupling phenomenon of the simulation results,the modified control methods are studied;In order to control the spinal quadruped robot based on the locomotion control methods,the force mapping relationship between the seven-bar model and the quadruped robot is established;A state machine for the bounding locomotion of a spinal quadruped robot is designed to solve the top-level control problem.It provides control methods for bounding locomotion of the spinal quadruped robot.Finally,a spinal quadruped robot co-simulation platform for experiment is built.The results of the locomotion starting,the situ bounding and the acceleration experiments verify the validity of the modified locomotion control methods and the correctness of the force mapping relationship.The effectiveness of the ASLIP model for robot locomotion control is verified based on the reacceleration and steady-state locomotion experiments.Results shows that the locomotion based on the ASLIP model can achieve higher steady-state horizontal speeds compared with the SLIP,which reflects the speed superiority of the ASLIP model.