Research On Mobile Assembly Planning And Control Of Mobile Manipulator Based On Attractive Region In Environment

As an important part of robot manipulation,assembly has many important appli-cations in the fields of national defense and civil,such as in-space assembly of space station and space telescope,shell assembly,mouse assembly,hard disk assembly and so on.For some of the tasks,such as maintenance between different deep-sea submersibles and spacecraft docking,the target is moving in the process of manipulation.In addition,for the assembly of objects with large volume and small mass,such as aircraft assembly and foam-product assembly,the assembly range will exceed the working space when the robot base is fixed.Such tasks put forward new requirements for robot assembly strategy and control.However,the current research mostly focus on the assembly of fixed base,and there are few research focuses on the mobile-assembly problem when the target and the robot move simultaneously.In this thesis,the mobile-assembly problem of mobile manipulator is investigated.The key problem including that the control for mobile platform should take consider-ation of velocity constraints and time-varying parameters,the precision of the system is far lower than the required precision of the assembly,the target coordinate frame is non-inertial and the control for mobile manipulator should take consideration of veloc-ity constraints and compliance.Based on the studies for Attractive Region in Environ-ment,combined with model-based control,neural-network control,barrier Lyapunov function and ARIE-based assembly strategy,we design the strategy and control of mo-bile assembly,and realizing high precision mobile assembly with low precision mobile manipulator,specified as follows:1.In view of the safety problem caused by the high speed of the mobile plat-form and the influence of time-varying parameters,the neural-network control of the nonholonomic wheeled mobile robot considering the velocity constraint and parameter uncertainty is proposed.Firstly,based on the dynamic model,the control of wheeled mobile robot considering velocity constraint utilizing the barrier Lyapunov function is proposed,the designed controller is proved to guarantee the tracking system asymptot-ically stable.Then,the controller considering velocity constraint and parameter uncer-tainty of nonholonomic wheeled mobile robot is proposed utilizing neural network,and the control system has been proved to be semi-global uniformly ultimately bounded.The controller can make the robot system learn the system parameter online while the velocity constraint being guaranteed,so that it can adapt to the changes of system pa-rameter.The proposed controller is tested on the simulation platform and the hardware platform respectively,and the effectiveness of the proposed method is verified.Com-pared with the Proportional Integral Derivative(PID)control and the adaptive sliding mode control,the superiority of the proposed method is verified.2.For the issue that the system precision is far lower than the required preci-sion of the assembly,through the use of environmental constraints,based on the theory of Attractive-Region-In-Environment(ARIE),combined with the compliance control method,for the space station translocation and docking verification experiment,a high precision sensor parts assembly task and the Video Graphics Array(VGA)line assem-bly task,we design the constraints and state-independent input to build ARIE,and de-sign the assembly strategy and control parameters which can achieve the requirement of different ARIEs.The proposed strategy and controller parameters are tested on the hardware platform to verify the effectiveness of the proposed method.3.To solve the mobile assembly problem of the mobile manipulator,based on the above two studies,the mobile assembly strategy of the mobile manipulator and the compliant control method satisfying the velocity constraint are proposed.Firstly,the assembly strategy based on the ARIE is extended to the mobile assembly strategy.To solve the problem that the target coordinate frame is non-inertial frame,the input force compensation method is adopted to compensate the influence of the acceleration and deceleration of the coordinate frame on the target object,so that the mobile assembly strategy can be designed based on the assembly strategy of the inertial frame.To solve the problem of mobile manipulator control,using a combination of redundant manipu-lator and mobile platform,the model-based control in research content 1 is applied to the control of mobile platform,on this basis,according to the requirement of the mo-bile assembly strategy,the desired force and constraint of the end effector is designed,and the required torque of each joint of robot is calculated according to the dynamic model of the robot.The proposed mobile assembly strategy and control are tested on a simulation platform to verify the effectiveness of the proposed method.