Distributed Model Predictive Control For Multiple Mobile Robots Formation

With the development of intelligent robot technology,more and more multiple mobile robots have been introduced into the production and life.Formation control is the most basic and important form of the cooperative control for multiple mobile robots system and applies in many important applications,such as aircraft formation choreography,intelligent navigation transportation,mobile sensor networks and others.However,it is difficult to control the multiple mobile robots formation system due to coupling between the mobile robots,external disturbances,calculation burden and energy constraint.In this thesis,based on Nash iteration algorithm,DMPC method,ESO-based DMPC method and distributed formation control method with multi-rate sampling are investigated for multiple mobile robots formation system.The main works of this thesis are listed as follows:1.Distributed model predictive control(DMPC)algorithm is proposed based on Nash iteration for multiple mobile robots formation control.According to the kinematics model and local performance index of mobile robot,the distributed formation control problem is transformed into an online iterative optimization problem by using Nash iterative algorithm.The local optimal control decision is derived by searching Nash equilibrium point for the mobile robot.Finally,the simulation results show the effectiveness of the proposed method.2.A distributed control strategy with extended state observer(ESO)is proposed for the formation system with external disturbances.According to the kinematics and the extended state equation of the mobile robot model under disturbance,ESO is introduced to estimate the disturbances.The local controller is designed by Nash iteration based DMPC.The estimated disturbance value is used to feed-forward compensate the designed controller.Finally,high disturbance rejection performance of the proposed method is shown by the simulation results.3.A distributed control strategy under multi-rate sampling situation is investigated by considering the energy constraint,the computational burden and communication traffic.With the control of the proposed method,the performance of formation system can be guaranteed by reducing the sampling frequency reasonably,and system states is predicted by model predictive controller in non-sampling point,the proposed method can also reduce the system energyconsumption,computational burden and communication traffic.