Design And Dynamic Gait Control Of The Rescue Hexapod Robot With Erect Posture And Parallel Mechanism Leg

Legged robots have superior mobility on uneven terrains when compared to wheeled vehicles,therefore they are believed to have a promising prospect for disaster rescuing applications.Among them,the hexapod robot can be a promising mobile disaster rescuing platform due to its good stability and high redundancy for fault tolerance.However,most current hexapod robots still rely on sprawl posture and traditional static gait planning methods,which largely limit their dynamic locomotion performance.In this thesis,a large-sized heavyduty rescue hexapod robot with erect posture and parallel leg mechanism was developed by taking both biological principles and practical requirements for rescuing tasks into considerations.Moreover,a dynamic tripod trotting gait control method based on multi-modal active compliance was proposed to enhance the locomotion speed,energy efficiency and the ability to cover uneven terrains of this robot.To this end,this thesis systematically studied the mechanism design,mechanism performance optimization,dynamic gait control method,control stability modeling and gait energy efficiency optimization of the hexapod robot.The main contents of this thesis are as follows:1)By considering the requirement of disaster rescuing tasks,we proposed a new robot leg mechansim,which adopts the erect posture like mammalian animals with parallel mechanism.After,a new large-sized heavy-duty rescue hexapod robot was designed.In this way,we can take the advantage of the stability of hexapod animals and the dynamic locomotion potential of mammalian leg at the same time.Moreover,the key devices can be protected in a shell altogether,thus the reliability can be enhanced.2)We established the kinematic model and dynamic locomotion performance model of the proposed leg mechanism,and optimized its dimensional parameters based on a complex performance certerion for dynamic trotting gait of the robot.Based on the performance optimization,we developed the rescue hexapod robot prototype,i.e.HexbotIV.Its hardware and software system were also developed.3)We proposed a new control strategy to achieve stable and fast dynamic tripod trotting gait for hexapod robot.By adopting a multi-modal active conpliance control strategy and a gait state machine for hexapod trotting gait,the SLIP dynamics can be realized on the hexapod robot.Simulations and experiments on the real hexapod robot were conducted,and the effectiveness of this control method was verified.4)We established the dynamic stability model of the dynamic tripod trotting gait for hexapod robotic system,which take the dynamics of multiple legs and the leg's impedance control model into consideration.Based on this model,the influences of various control parameters on the gait stability were analyzed.By both simulations and experiments,we studied the dynamic responses of the hexapod robotic system under the disturbances caused by terrain unevenness during dynamic trotting.In this way,the stability of the proposed control strategy against terrain unevenness was verified.5)We analyzed the gait energy efficiency of the rescue hexapod robot.After,we proposed an online optimization method for gait energy efficiency.A gradient estimation model was utilized to avoid the influence of model error and measurement error.As a result,the hexapod robot was able to find the optimized gait parameters automatically during the dynamic tripod trotting,and the energy cost can be significantly reduced.The study of this thesis presents a valuable reference of the design and control of hexapod robots,and lays a fundamental for the practical use of the high-performance robotic platform for disaster rescuing tasks.