Finite-time Consensus And Formation Control Of Multi-robot Systems

Robot systems have been widely used in the areas of manufacturing,military,national defense,et al.Manipulators and mobile robots are two typical robot systems.It has increasing demand for multi-robot cooperation in many kinds of operations,such as multi robot cooperative assembly,mobile robot logistics,etc.Consensus and formation control of multi-robot systems are important basis for collaboration of the multi-robot systems.However,in the complex environment,the manipulators and mobile robot systems will inevitably be affected by their own constraints and external disturbance,which brings challenges to the consensus control of the multiple manipulators and the formation control of the multiple mobile robots.In addition,the multi-robot systems are required to adjust their states quickly in finite time to respond to the task quickly and reliably in some special tasks.Therefore,it is of great significance to study the finite-time consensus and formation control of multi-robot systems subject to self-constraints and external disturbances.In this thesis,the finite-time consensus control and formation control problems are studied for fully-actuated manipulators and under-actuated mobile robots described by Euler-Lagrange equation.The main work is as follows:(1)The finite-time consensus control method of multi-manipulator systems without velocities is studied.A node-based finite-time consensus control strategy is proposed to achieve finite-time consensus control of the multiple manipulators under undirected topology and directed topology.A super-twisting sliding mode-based observer independent of the system model parameters is designed to estimate the velocity consensus error,which improves the estimation accuracy and speed.A finite-time consensus control strategy based on the super-twisting sliding mode-based observer is proposed to achieve finite-time consensus control of multiple manipulators subject to unmeasurable velocity.(2)The anti-disturbance finite-time consensus control method of the fully-actuated multi-manipulator systems is studied.An integral sliding mode surface with complete robustness is designed.By eliminating the reaching phase of the sliding mode control,it can be guaranteed that any initial state of the system stays on the sliding mode surface,and the disturbance is compensated to improve the anti-interference performance of the system.Under the fixed topology and switched topology,the anti-disturbances finite-time consensus control frameworks including nominal control and discontinuous control are proposed respectively to achieve the finite-time consensus control of multi-manipulator systems.(3)The finite-time formation control method of underactuated mobile robot system is studied.A fast finite-time extended state observer based on linear function of estimation error is proposed to estimate the unmeasurable speed and disturbance,which improves the estimation speed and accuracy.Furthermore,the underactuated mobile robot is modeled as a cascade system containing kinematic and dynamic models.A distributed antidisturbance formation control method based on extended state observer is proposed to realize the finite-time formation control of multiple robots.(4)The finite-time formation control method of underactuated mobile robot with unknown model parameters is studied.A finite-time parameter identification method is proposed by using the idea of memory stack,which make the parameter estimation independent of system state derivative and input gain.A data-driven based adaptive finitetime extended state observer is proposed to realize model-free estimation of unknown velocity information and disturbances.Based on the estimated model parameters,velocities and disturbances,a data-driven based formation control strategy is proposed to achieve multi-robot formation with obstacle avoidance,collision avoidance and connectivity maintenance.(5)The fixed-time formation-containment control method for underactuated mobile robot system is studied.A fixed-time extended state observer is designed to estimate the unknown velocity information and total disturbance signal.The estimation error is strictly convergent by using the high-order sliding mode control method.A two-layer control framework consisting of formation control and containment control is proposed.The leading robots can achieve the formation control with obstacle avoidance in fixed time,and the following robots can achieve containment control in fixed time.The convergence time of the extended state observer and the formation-containment controller is independent of the initial state.The fixed-time formation-containment control method of underactuated mobile robot system is studied.A fixed-time extended state observer is designed to estimate the unknown speed and disturbance,which makes the estimation error converge strictly.A two-layer control framework consisting of formation control and containment control is proposed to achieve fixed-time formation control of the leading robots with obstacle avoidance.The following robot achieves fixed-time containment control.The convergence time of the extended state observer and the formation-containment controller is independent of the initial state.