Research On Anthropomorphic Dexterous Operation Control Of Nuclear Robot

The research on anthropomorphic dexterous operation control of robots is an important research direction in the field of tri-co robots.The project originates from the National Nuclear Energy Development Project "Research on Key Technologies of Nuclear Emergency Disposal Robot",which is based on the demand of dexterous control of complex operation process in special environment.In view of the constraints of space characteristics,task complexity and damage of electronic devices on the dexterous operation of robots in special environments(especially in the case of nuclear accident emergency),considering the adaptability of control systems to special environments,this paper develops the anthropomorphic dexterity for the disposal of robots in nuclear environments,starting from the efficiency and safety of the dexterous movement of robots.Control research,including:The traditional tele-operation control method limits the lack of dexterity of the nuclear robot in the process of motion,and can't maximize the operation speed and response time of the nuclear robot.This paper designs a differential evolution(DE)algorithm optimized of nuclear robot motion planning method.Firstly,the algorithm analyzes the kinematics and dynamics of the nuclear robot,and obtains the working space of the nuclear robot through Monte Carlo.In addition,the generalized momentum algorithm is used to detect the collision in the workspace to improve the global optimization ability of DE.The motion parameters are used to plan the motion of the joint space.On this basis,based on the mechanism of human arm movement,the movement primitive decomposition method of the nuclear robot to complete the screwing valve is designed,and the motion planning based on dynamic motion primitive theory is proposed.In order to avoid the impact of model uncertainty on the sensitivity of collision detection,a collision detection method based on adaptive residual estimation is proposed to solve the problem that the flexible control accuracy of nuclear robot is required to complete complex contact tasks such as arrival and grasp,and to prevent collision between nuclear robot and terminal operating objects.A collision detection method based on adaptive residual estimation is proposed.The generalized momentum residual estimator is designed and the nonlinear factors of the nuclear robot(such as joint friction,velocity and acceleration)are included in the residual uncertainty of the model.Secondly,the paper introduces dynamic thresholds and estimates the model parameters by gradient correction,thus improving the detection accuracy.Finally,the thesis takes the six-degree-of-freedom nuclear robot for anthropomorphic dexterity work as the research object,and carries out the verification experiment of motion planning and collision detection algorithm.At the same time,it compares with the traditional method to evaluate the performance of robot dexterous operation.The experimental results show that the method adopted by this research makes the robot operating speed and response time fast,the collision detection sensitivity is high,and the false alarm probability is low.The method proposed by the subject can improve the efficiency and safety of the robot operation process and realize the dexterity of the nuclear robot operation.