Terminal Sliding Mode Active Disturbance Rejection Control For Modular Robot Manipulators: Theory And Experimental Study

With the rapid development of science and technology,the unmanned work mode plays an increasingly important role in various fields.As the core carrier of unmanned work mode,robots have increasingly become an indispensable tool for human creating and improving society.For the traditional robots,in the face of complex environments,the control accuracy is not high and ability to complete tasks is weak.Meanwhile,the cost and maintenance are high.The traditional robots can no longer meet the needs of human tasks.Therefore,modular robot manipulator(MRM)with low cost and high control performance that can meet various task requirements.The MRM has received extensive attention.The MRM is a nonlinear system that is composed of standardized electromechanical design interface modules such as the data reading and writing module served by bus device,the motion control module,the power supply module,the I/O control module and the sensor measurement module.The configuration of MRM can be freely adjusted according to the missions with different needs.From the perspective of hardware,modular robot manipulator system is an organic mechanical combination such as links and rotary joints with standardized interface properties.From the view of control,each module subsystem is regarded as an autonomous control individual.Hence,the modularity makes the MRM with good flexibility and adaptability.The MRM has received growing attention in the robotic field.The MRM is widely applied in disaster emergency rescue,space mission operation and deep-sea environment detection.In practical applications,complex external disturbance is variable.The model uncertainty caused by inaccurate modeling and external disturbance greatly affect the control performance.Therefore,there is an urgent need to find a suitable control strategy to offset the adverse effects.The expected control strategy can ensure that nonlinear uncertain MRM has high control precision.In actual project,it is often necessary to have better real-time performance that the system can respond immediately.Although the complex control strategy has high control accuracy,its algorithm execution time occupies a large amount of system bandwidth,which is not conducive to engineering application.Therefore,it is necessary to study the dynamic control method for an MRM with modular design idea,low complexity,fast operation speed and compensation ability for the model uncertainty.Aiming at the above problems,this thesis studies the terminal sliding mode active disturbance rejection control method for the MRM.The effectiveness of the proposed algorithm was verified through modular robot experiments for tracking control.The specific research contents are as follows:(1)Introduce the background and significance of this thesis,and then overview the research status and hot issues of MRM.(2)The joint torque feedback(JTF)technology is used to compensate for the load on the joint motor system from other motors,link and end payload.The JTF method not only considers the force/torque acting on each joint and link,but also ponders the coupling force/torque between the joints.In each subsystem model,there are parts of moment of inertia,friction,IDC,sensor feedback and external disturbance term.Eventually,the dynamic model of the n-DOF MRM system is obtained.(3)To solve the problems of large overshoot and long adjustment time caused by integral saturation,a linear adjustment integral sliding mode ADRC algorithm is designed to reduce the steady-state error.Under the condition of large initial error,a linear adjusting integral action through absolute value of difference between sliding surface and boundary layer is weakened to integral action for ensuring the stability of MRM.The extended state observer(ESO)is used to estimate and compensate the friction,IDC and external disturbance.The application of ESO reduces the switching control gain of the sliding mode and effectively weakens the chattering.(4)To overcome the mutual constraints between the system chattering and the arrival time in traditional sliding mode control,in this thesis,a tangent excitation function-based integral terminal sliding mode control is proposed.The proposed NLESO based on tangent excitation function(tansig)is utilized to estimate and compensate the system uncertain disturbances.The proposed NLESO can simplify the traditional extended state observer design parameters for theoretical analysis and practical application.To diminish the arrival time when the system state converges to the equilibrium point in this thesis,the proposed sliding mode control algorithm integrates the adaptive term exponential reaching law for the MRM.The presented algorithm improves the robotic system anti-disturbance ability.Finally,the conclusions and the perspectives of future research are presented at the end of this thesis.