Research On Recursive Dynamics Modeling And Control Of Robotic Manipulators

The unstructured environment and diversity of the tasks require robotic manipulators have the capability expanding rapidly from low degrees of freedom to high degrees of freedom,from a single arm to dual arms.Meanwhile,some manipulators are equipped with flexible joints to improve safety and flexibility of the human-robot interaction.The study of dynamics modeling and control methods based on the model are the foundation for achieving high dynamic performance of the above robots,and rapid expansion of the new structures of robotic manipulators require higher demands on the convenience and speed of modeling and control.In order to quickly establish the dynamics model and realize high precise and high efficiency control for robotic manipulators with different structures,based on the recursive form,this thesis carries out the research on modeling and control from rigid joints to flexible joints,from a single arm to dual arms,and using the simulation and experiment methods to verify the above algorithms.In order to achieve scalability of the robotic manipulators,a recursive inverse dynamics model Sub-RNEA is established based on the Newton-Euler dynamics with the subsystem form.And based on the chain method and the substitution method,the forward dynamics models are established.Based on the Sub-RNEA,differentiating the subsystem dynamics twice with respect to time and combining with some formulations in the EJNEA,a high order inverse dynamics model Sub-EJNEA based on the subsystem for flexible-joint manipulator is proposed.Based on the analysis of the output results with the physical simulation model and the EJNEA model,correctness of the above models is verified.For the effect of uncertainties of the dynamics model parameters and disturbances in the motion control for rigid-joint manipulators,a composite adaptive control algorithm based on the virtual decomposition control and the robust integral of the sign of the error is proposed.Based on the virtual decomposition control,the adaptive control with subsystem forms is designed,meanwhile,a model-free control algorithm RISE is introduced to copy with the disturbances,and the recursion and robustness of the algorithm are obtained to improve control performance s for the multi-Do F manipulators.The experimental results are given to illustrate the advantage of the algorithm in the trajectory tracking precision and energy consumption.For the problem that the position of the target object and the clamping internal force need to be controlled simultaneously to achieve handling safety for rigid-joint dual-arm cooperation,under the assumption of rigid grasping,the contact force between the robotic manipulator and the target can be decomposed into the component force to control the target motion and the internal force to squeeze the target,and then the control algorithms can be designed independently for motion control and internal force control.A force/position hybrid control algorithm based on the VDC-RISE is proposed,both simulation and experimental results show that the proposed algorithm in the thesis achieve higher position tracking precision while ensuring the internal force tracking.For the task of the desired interaction between the handled target and the external environment,the hybrid force-position control strategy cannot handle the contradiction between the high stiffness required for motion control and the high flexibility required for interaction control,and a dual-loop impedance control algorithm based on VDC is proposed,which can cope with the problem of the desired interaction between the target and the uncertain environment by designing the outer-loop impedance control method combining stiffness adaptation and dynamic damping adjustment.The design of recursive implementation of the inner-loop impedance not only ensures real-time performance of the algorithm,but also avoids dependence of the classic inner-loop impedance on the robot dynamics model.Simulation results show that the proposed dual-loop impedance control algorithm can meet the control requirements of internal force,target position,and target-environment interaction force.For week robustness of the feedback linearization control for flexible-joint manipulators,under the assumption of high order feedback variables are available,an enhanced feedback linearization control algorithm is proposed combined the FL with a superposed feedback,and it improves the robustness facing with uncertainties of the dynamics model and high-order feedback variables.When the high order feedback variables are unavailable,an enhanced RISE+ control algorithm is proposed combined the RISE with the nominal high order feedforward.And expanding the RISE+ control,a force/position hybrid control algorithm based on the RISE+ for dual-arm manipulators with flexible joints is proposed.The simulation results illustrate that control performances of the control algorithms for the single-arm and dual-arm manipulators are improved.