Research On Consensus Of Multiple Underactuated Systems

Consensus control of multi-agent systems has become a hotspot in the field of control.It has been widely used in military,industrial,and agricultural fields.However,the research on the consensus control of underactuated system is still not involved.In this paper,the dynamic characteristics of underactuated TORA(Translational Oscillators with Rotating Actuator)system are analyzed,and dynamic model are established.With the knowledge of graph theory,Lyapunov stability theory and matrix theory,the design of consensus control protocol and the related stability analysis are studied.The main contents are as follows:Firstly,the underactuated TORA system is selected as the controlled object,and its dynamics modeling is established.In the presence of uncertainties and external disturbances in the system,the system model is transformed into an affine nonlinear form.The system contains unknown nonlinear function terms and external disturbances,the processing of these two parts is the basis for the consensus control.Secondly,the consensus control based on adaptive fuzzy estimation algorithm is studied.For the existence of the unknown nonlinear function in the affine nonlinear system,the adaptive fuzzy estimation method is used to estimate it in the design process of consensus control law,and the compensation term is designed to compensate the external disturbance of the system.For the case of leaderless topology and leader-following topology,the consensus control protocol is designed separately.The effectiveness of the consensus control protocol is verified by the multiple TORA systems.Furthermore,a consensus control law based on the ESO(Extended State Observer)method is designed.On the premise that the relative velocity information of the system is unmeasurable,the ESO is used to estimate the unknown nonlinear functions and external disturbances in the system.Because the topology may be time-varying,the two cases of fixed topology and switching topology are discussed,and the corresponding consensus control law is designed.The multiple TORA systems are used for simulation experiments to verify the effectiveness of the designed consensus control law.Finally,the finite time consensus control problem based on RBF(Radial Basis Function)neural network is studied.The unknown nonlinear function in affine nonlinear system is approximated by RBF neural network based on the approximation principle.The effectiveness of the finite time consensus control protocol is verified by the multiple TORA systems.