Research On Remote Control Of Humanoid Robot Based On Networ

Nowadays,with the continuous progress of science and technology and the vigorous development of robot industry,the research and development of humanoid robot is popular all over the world because humanoid robot has the shape similar to human and flexible operation performance.The rapid development of network not only brings convenience to our life,but also the teleoperation robot based on network has become one of the research hotspots in the world.In some special environments,humanoid robot can not complete complex tasks independently.In order to ensure the working ability of robot in complex environment,this paper will study the control method of humanoid robot based on network.The research contents of this paper are as follows:Firstly,the teleoperation control platform of humanoid robot is built.The hardware part uses force feedback joystick,computer,camera and humanoid robot;In the software,the client program,server program and head and lower limb teleoperation command control protocol are designed.According to the task requirements,the arm model is established and the real-time teleoperation program is designed.Secondly,in order to improve the operation ability of the operator in the teleoperation system,the virtual force model is established by using the artificial potential field method,and the force felt by the operator in theory in the operation process is calculated.Taking the teleoperation handle as the force feedback device in the teleoperation system,the realization method of force telepresence is given,and the experiment of the operator manipulating the handle to remotely manipulate the humanoid robot arm to grasp objects is carried out.The experimental results show that the theoretical value of the force felt by the operator calculated by the virtual force model is basically consistent with the actual force felt by the operator during the experiment.Then,in order to improve the efficiency of teleoperation,dual port RAM is introduced to solve the problem of asynchronous sampling period at the master-slave end,and the synchronous control of the master-slave end is realized.Using UDP protocol to test and analyze the network delay between the two places,it is concluded that the probability density of network delay in any time period can be described by the translation gamma curve,and the prediction method of the distribution parameters of the translation gamma curve of delay probability density is proposed.Experiments show that the variable teleoperation cycle control method based on time delay distribution prediction effectively improves the operation efficiency compared with the current general fixed teleoperation cycle control method based on the maximum allowable delay.Finally,in order to improve the operation performance of teleoperation system,based on supervisory control,the optimal task allocation of master control and slave autonomous control is studied.Based on the physical programming method,an optimal control method combining master control and slave autonomous control is proposed,the preference function of the system is constructed,and the constraints are given.The experimental results show that the combination of master control and slave control can maximize the operating performance of the humanoid robot teleoperation system.