Model Reduction,Identification And Robust Control Of Multi-Rigid-Body System

A system consisting of different rigid bodies joined by kinematic pairs(hinges)is called multi-rigid-body connection system.Multi-rigid-body connection system is widely applied in aerospace,vehicle engineering,bionic robot,multi-joint manipulator and other industrial scenes because of its ability to accomplish special complex tasks.It is also a hot issue in the field of automatic control and rigid body mechanics.However,some dynamic models of multi-rigid objects are too complex to direct the design of the controller effectively.Therefore,the problem of how to simplify the model scientifically and then to design the control law based on the model is practically meaningful and theoretically important.For meeting the specific technical requirements of National Natural Science Foundation project and the space institute project,the thesis research the problem of model identification,reduction and robust control problems for the multi-rigid-body connection systems whose the number of rigid-body can be simplified to three.A bionic multi-rigid-body robot fish and a three degree-of-freedom helicopter are selected as the specific research objects.Their geometric shape and mechanical structure characteristics are analyzed,and the common robust control problems and common techniques of this kind of multi rigid body connection system are found.For a three degree-of-freedom helicopter which has a few degree-of-freedom,if the degree-of-freedom is reduced,the complete description of the system will be destroyed.A priori information is obtained by establishing a nonlinear model,and the transfer function model of elevation,pitch and travel motions is obtained by using least square identification algorithm.On this basis,the Matlab/Simulink attitude controller module is built.For a bionic nine-rigid-body robot fish which has a lot of degree-of-freedom and more complex dynamic properties,the thesis put forward reasonable model simplification strategy,which reduce the number of rigid-body to three,according to the mechanical structure of the system and physical parameters of each rigid body.The step signal is used as the test signal to identify transfer function of the reduced model for transversal and portrait plane motion,and the robust control law of the swimming movement is designed.At the same time,the stability margin of the system is analyzed,the control effect is verified by Matlab preliminarily,and experimental data of reduced model is compared with swimming simulation data of complete model.In this thesis,a bionic multi-rigid-body robot fish digital simulation platform and a three degree-of-freedom helicopter experimental platform are used to verify the controller's control effect.The experimental results show that the control law based on simplification and identification model is reliable and effective.