Fuzzy Control Methods For Underactuated Truss-like Manipulator

Underactuated robot means its number of actuators is less than the number of degrees of freedom.Underactuated robot has attracted attention of more and more researchers around the world due to its compact structure,light weight,low energy consumption,high reliability and strong environmental adaptability.An rope-driven underactuated robot based on parallelogram truss mechanism is studied in this dissertation,called underactuated rope-driven truss-like manipulator.The ingenious parallelogram truss mechanism is designed to be collapsible,expandable,and shape adaptatible in grasping,as which consequence it has a very broad application prospects.Underactuated truss-like manipulator is a new type of manipulator with complex stucture,thus it is difficult to design controllers by using conventional control methods.The development of fuzzy control theory has formed a relatively complete theoretical system.Fuzzy control method does not rely on precise mathematical model of the controlled object,and has the advantages of antiinterference ability,etc.,which is very suitable for the control problem of the robot arm.The dynamic model is studied from the perspective of control,and the related control problems are deeply explored.The fuzzy control theory is the core method applied to the manipulator in various scenarios.A semi-physical simulation platform is constructed for the actual underactuated truss-like manipulator.The dynamic behavior of the manipulator and the effectiveness of the control methods are verified in the semi-physical simulation experiment.The research in this dissertation will cover the following aspects:Based on the physical structure and mechanism of the underactuated truss-like manipulator,the complex vector combined with Lagrangian equation method and model equivalent method is utilized to establish the dynamic model under non-ideal conditions,including that the joints are non-homogeneous rods,the dimensions of each joint are variable,the direction of the gravitational field is arbitrary,and the joints of the manipulator are frictional and elastic.The underactuated truss-like manipulator is inverted in the gravitational field to become an underactuated system affected by gravity.It is necessary to design a control law to stabilize it in a vertically upward equilibrium position.Because the dynamic model of the underactuated truss-like manipulator is complex and contains a large number of nonlinear terms,and there are problems such as model uncertainty and external disturbance,this dissertation designs Mamdani type fuzzy controller based on the advantage that fuzzy controller does not depend on the precise model of the controlled object.The fuzzy controller has a larger convergence domain than the linear controller.Furthermore,the Type-2 fuzzy controller is designed by expanding the membership function to three dimensions to cover the uncertain information.Simulations are conducted in MATLAB\Simulink,and the results show that the Type-2 fuzzy control has better control effect when there is modeling uncertainty involved.The planar underactuated truss-like manipulator running in the horizontal plane is closer to its conventional state.Under this condition,there is no influence of the gravitational field,and its wide-range motion more obviously exposes the nonlinearity and non-fullrankness in the dynamic model.Based on the above problems,this dissertation proposes a pseudo-inverse method to solve the problem of the matrix dissatisfaction rank,and then utilizes the feedback linearization method to convert the controlled subsystem into a linear system.By combining the LQR method,controller is designed to realize the goal of trajectory tracking control.However,pseudo-inverse method makes some system information lost.To solve this problem,it is proposed to use the coordinate transformation method to process the control force matrix into a form that can use partial feedback linearization method,and then design the controller for the linearized system.The first two methods must be based on a very accurate mathematical model of the controlled object,which is greatly affected by external disturbances.To solve this problem,a fuzzy controller is proposed to design the controller.The numerical simulations verify the effectiveness of the above three methods on the desired trajectory tracking control of underactuated truss-like manipulator duiring opening and closing process.To compare the control effects of the three control strategies,external disturbances are added in the system,and the results show that the fuzzy control has best performance among the three methods.Since there is no correlation with the actual robot arm in the digital simulation,this dissertation builds a semi-physical simulation platform based on MATLAB Simulink Desktop Realtime to verify the control effect of the controller on the actual controlled object.Based on this semi-physical simulation platform,a rope pretightening control scheme for the actual underactuated truss-like manipulator is proposed to protect the rope during the operation of the manipulator to prevent the rope from slipping and damaging the device.The fuzzy controller of the joint angle tracking control of the manipulator arm is verified by the semi-physical simulation experiment.The designed fuzzy control method is effective to track the desired position signal during the gathering and opening process,and the designed fuzzy controller is proved to be correct and useful.