Research On Robust Consensus Control Of Cluster Under System Uncertainties And Communication Constraints

Collaborative control of cluster systems is to accomplish more complex tasks through information interaction among members,and has become an important means to improve efficiency in military and civilian fields today.The consensus problem,as a fundamental problem of cluster control,is a popular research topic today.This paper takes the impact of uncertainty in the system model,the presence of external perturbations and having communication constraints on the behavior of cluster systems as the entry point to study cluster robust consensus control methods to make the system achieve behaviors such as consensus tracking and bipartite consensus.The main research work and innovative results are as follows.1.Research the problem of robust consensus tracking of a cluster system with model uncertainty and unknown disturbance.This paper estimate the system uncertainty using measurable variables and compensate for the perturbations by combining the neural network estimation principle and Uncertainty and Disturbance Estimator(UDE).A robust consensus tracking control algorithm is proposed for this system,and the effectiveness of the algorithm and the control effect are verified by simulation.Compared with previous robust cooperative control methods based on neural network estimation or UDE alone,the proposed method balances the computational complexity and the need for system model information integrity.2.Research the discrete-time mean-square consensus problem of clustered systems under communication constraints.Based on an improved interference observer,a binary-valued consensus algorithm for cluster systems under extreme communication conditions is proposed.The system can only sample bit sequence information that characterizes whether the distance of neighboring intelligences exceeds a threshold,and there is channel interference and measurement noise.By using a recursive projection algorithm to estimate the state values of neighboring intelligences,the cluster mean-square consensus condition is derived based on Lyapunov stability theory,so that all states of the system converge to the same value,and the estimation error of the recursive projection algorithm is consistent and eventually bounded.Finally,the control effect is verified by numerical simulation.Compared with existing binary-valued consensus algorithms,this paper effectively improves the anti-interference capability of the system under complex communication conditions by designing an interference estimator applicable to the binary-valued observation condition.3.Research the problem of bipartite consensus of cluster systems with external perturbations.Using a UDE-based robust control strategy,a robust control algorithm for bipartite consensus of cluster systems with external perturbations is proposed,and the effect of different time scales of the designed first-order filter on the bipartite consensus is analyzed.Based on Lyapunov stability theory,the bipartite consensus condition is given.Numerical simulation results show that with this method,the negative interaction weights in the topology divide the cluster members into two opposing parts,whose states eventually converge to two values of opposite numbers of each other,and the UDE-based estimation errors are consistently and eventually bounded.