Research On Gait Control Strategy Of Hexapod Robot

With the pace of people exploring nature accelerating, the operating environment is continuously expanded. There are a number of high risk jobs which need the various robots instead of human to complete under some special circumstance. Multi-legged walking robot has the advantages of utilizing isolated ground to support and realizing the various gaits to walk and switch. Therefore, it can complete some complex jobs in a non-contact manner under the non-structural environment. Hexapod robot is a high representative among multi-legged walking robots. It has redundant mechanical structure, flexible ground through capacity and superior load capacity compared to a biped and quadruped robot, and is more suitable to work in the non-structural terrain. However, with the increase of the freedom degree of the robot joint, traditional control algorithm is difficult to realize the complex control rules of multi-joint coordinated motion. The multi-joint coordination problem can be solved based on the combining method of the bionic algorithm of central pattern generator (CPG) and traditional kinematic analysis when the robot walking. Domestic research for the gait of hexapod robot is mostly limited to the structural terrain, which is the strategy planning for the robot gait on flat road. The research of gait strategy of non-structural terrain is still in the exploratory period, and there is no fully self-control gait strategy.For the above, CPG bionic algorithm is combined with traditional kinematic in the paper. VarderPol equation is adopted to design oscillator and annular CPG neural network of hexapod robot is built, which is for the gait generator of hexapod robot to control root joint of the six legs. The robot move order is changed by phase-interlocked relationship. Kinematic analysis method is adopted to plan the motion trajectory of the hip and knee joint of one leg which do interpolation motion with the root joint in a same period. Support phase and swing phase movement are output by solving forward and inverse kinematics equations for one leg of the robot. Autonomic reflex control is researched based on the two, three and six basic gaits of hexapod robot in non-structural terrain, which aims at three complex terrains of crossing obstacle, crossing ditch and climbing slope.Finally, MATLAB-ADAMS is adopted to build co-simulation platform. A closed-loop control system is built through control system designed by Simulink module and ADAMS feedback information. The basic gait planning of hexapod robot and local independent gait of non-structural terrain is test by simulation platform. The experiment results show the two, three and six basic gaits of hexapod robot can be achieved through the gait strategy of combination of CPG and traditional kinematic. Hexapod robot can walk on non-structural terrain by adding autonomic reflex control strategy.