Research On Leg Joint Control Strategies Of The Hydraulic Quadruped Robot

Robot is an interdisciplinary technology integrating mechanics,the electricity,control strategies,sensors,computers,artificial intelligences and other disciplines.It is increasingly widely used in modern logistics,the medical treatment,the road transportation,the urban emergency,the energy and mining,the military and national defense,and other fields.The multi-legged bionic robot is an important branch of the robot technology.Among them,the hydraulic quadruped robot has advantages of large loading capacities,the fast moving speed,and the good adaptability to terrain different environments.It has great significances to the application for the complex terrain,the special transportation and other operation fields.Taking the leg joint control in the hydraulic quadruped robot as the background,this paper designs joint controllers and control strategies under the modeling and analysis of the driving system.The whole structure model,the leg structure model and the joint driving system mathematical model in the hydraulic quadruped robot are established.The stability of the joint driving system is analyzed by using the Bode diagram.An integral separation PID controller is designed in the position control system when the foot will drop the ground,and the proposed controller has the good trajectory tracking accuracy and the rapidity.An active disturbance rejection controller with strong disturbance rejection performances is designed in the force control system when the foot drops the ground,the proposed controller can effectively reduce the ground impact and improve the system stability.Aiming at the parameter tuning problem of the proposed integral separation PID controller and the proposed active disturbance rejection controller from the intelligent optimization processing perspective,this paper takes the leg joint of the hydraulic quadruped robot as the research object,and designs leg joint control strategies based on intelligence algorithms.Integral separation PID control strategies based on improved beetle antennae search algorithms is proposed.The integral separation PID controller parameter tuning problem is transformed into the three dimension optimization problem solved by the intelligence algorithm,which can effectively solve the controller parameter tuning problem.Aiming at the problem that the fixed reduced exploration step and the attenuation factor are used in the basic beetle antennae search algorithm to can not have a thorough local searching,the Lévy flight beetle antennae search algorithm is proposed in this paper.Then,aiming at the shortcoming that the basic beetle antennae search algorithm fails to consider differences between individuals and dynamic information in the searching process,which can lead to the population diversity loss,a beetle antennae search algorithm based on the elite selection mechanism and the neighbor mobility strategy is proposed.Active disturbance rejection control strategies based on improved salp swarm algorithms are proposed.The active disturbance rejection controller parameter tuning problem is transformed into the sixteen dimensional optimization problem solved by intelligence algorithms.The leg joint force control strategy of the hydraulic quadruped robot is designed from the intelligent optimization perspective.To overcome the shortcoming that the strong randomness of the search step is ignored in the iterative optimization,which leads to the over jumping of each individual and the incomplete local searching,the self-growing Lévy flight salp swarm algorithm is proposed in this paper.The salp swarm algorithm uses differences of the searching range as the moving step,which will lead to the algorithm can not accurately carry out the local-global searching stage.Then,the balance and detection factor of the original algorithm is an exponential function with exponential disaster at the end of the iteration,which leads to the accuracy reduction of the salp swarm algorithm at the end of the iteration.This paper proposes a modified salp swarm algorithm based on the perturbation weight.Tested functions are selected for numerical testing experiments,and proposed algorithms performances are analyzed from the algorithm calculation accuracy,the algorithm iteration speed and the algorithm optimization path graph.Finally,the position control strategy experiment based on improved beetle antennae search algorithms,and the force control strategy experiment based on the improved salp swarm algorithms are completed on the hydraulic quadruped robot experimental platform,and experimental results show the effectiveness of the proposed control strategy in this paper.