Research On Adaptive Position/Force Control Of Manipulator Based On Disturbance Observer

When the actuating end of the manipulator contacts with environment of the work in industrial production,the manipulator must have the awareness and response quickly to external collision,in order to ensure compliant control to achieve ideal control effect,and ensure the safety of people and the environment,which can not only guarantee to track the location of working precisely,but also identify the external contact force accurately.In this paper,the compliant control of position and contact force is studied by taking the environmental stiffness and external interference into consideration.It mainly involves the following two aspects:In order to model in the extension of space and estimate the interference for constrained manipulator system,the first thing is to decouple the motion and force at the end.On the basis of traditional dynamic model,the environmental stiffness is introduced.Then,the coordinate transformation is used to realize the extension from free space to constrained space and establish the dynamic model of manipulator-environment unified extension.While a super twisting interference observer is designed for the uncertain interference existing in the extended space for the constrained manipulator.The auxiliary vector is introduced in the observer,and the super twisting algorithm is used to realize the estimation of the uncertain interference.Finally,for the disturbance torque at the end of the constrained manipulator,a global disturbance observer is designed by means of generalized momentum.The observer can estimate the disturbance torque,and at the same time reduce the cost of installing sensors.For the compliant control of position/force of the constrained manipulator,firstly,the controller is designed and the stability is analyzed without interference.Then adaptive position/force controllers based on disturbance upper bound and disturbance observer are designed respectively to ensure the unknown parameters can be corrected in real time by introducing parameter error into the adaptive law,so that the position and force can be tracked and controlled well.Finally,the validity of the proposed algorithm is verified by MATLAB.