Research On Virtual Modeland Three-Part Control Of Bionic Hexapod Robot

Traditional wheeled and tracked mobile robots can only run on relatively flat terrain due to their tire and track movement mechanisms,and their range of motion is limited by the size of the site.In contrast,legged mobile robots have multiple degrees of freedom in each leg and can rely on isolated ground support points to gain a foothold,which provides great flexibility and adaptability in the face of complex and rugged irregular terrain.Among them,hexapod robots have received more and more attention because of their better stability,flexibility and adaptability to the environment.Due to the complex structure of the hexapod robot itself,the variable operating environment,and the frequent interaction with the environment during operation,this puts high control requirements on the hexapod robot.To address the motion control problem of the hexapod robot,this paper explores the tripod gait stable walking control method of the hexapod robot from the existing bionic hexapod robot in the group,following the research sequence from local to overall.The main research ideas and contents of the paper are as follows.(1)According to the special structure of the real machine of the bionic hexapod robot,a multi-coordinate system is proposed for describing the different motions of the robot,and the D-H convention method is used to analyze the kinematics of the single leg of the robot and to establish the position-velocity relationship model of the single leg of the robot.A foundation is laid for the analysis and control of the robot's single leg and the whole.(2)According to the motion characteristics of the robot single leg,the single leg motion is divided into the standing phase and the swing phase,and different virtual model control mechanisms are set for the two phases respectively,and the switching of the robot single-leg phase is accomplished by means of designing the trigger mechanism.To address the problem of constructing complex constraint matrices when the virtual model control method is used for the motion control of the hexapod robot,the direct control of the whole method is abandoned,and the robot is regarded as an ordered combination of multiple legs,and the control effect is applied to each leg separately.The simulation proves that the virtual model control can track the set single-leg motion trajectory very well.(3)Based on Raibert's three-parts control for bouncing robots and combined with the study of human walking motion,a three-parts control method suitable for foot-based robots is proposed.The motion control of the footed robot is decomposed into torso attitude control,leg-lift height control,and velocity control,which is further divided into forward velocity control and angular velocity control according to the robot motion mode.At the same time,a state machine is established for coordinating the overall motion of the robot.And the tripod gait controller of the hexapod robot is designed according to the state machine and the robot control algorithm.Finally,simulation and real machine experiments show that the hexapod robot can move smoothly on flat ground with tripod gait,each control part of the three-parts control has good decoupling,and the robot can overcome uneven terrain such as steps and slopes.