Research On Combat Stochastic Collaborative Control Based On CAS Paradigm

A complex adaptive system is a system in which adaptive agents interact,evolve together and emerge in layers.It is the hot topic and frontier in the field of complexity science.The adaptability of a complex adaptive system is mainly reflected by the fact that agents are able to adapt actively to the changed environment according to the circumstances via cooperating with and other agents.The combat system,based on its multi-agent behaviors and multi-task tasks,is a typical complex adaptive system with its features such as interaction mechanism as well as non-linearity,aggregation and de-aggregation,dynamic nature,variety and so on.The complex adaptive system paradigm is a research model for systematic analysis and design using the complex adaptive system theory and an effective method to study the combat stochastic coordinated control of combat system.Based on this paradigm,this paper explores problems like multi-agent robust consensus and formation control,intelligent coordinated control,information interaction,coordinated task allocation,stochastic coordinated control simulation and so on.The main research contents and contributions are as follows:For the robust consensus control problem of multi-agent systems with time-varying delay,parameter uncertainty and external disturbances,a controller synthesis algorithm is proposed for the realization of robust consensus.First,in consideration of the general model of linear agents with parameter uncertainties and external disturbances,based on each agent's behaviors,a distributed control protocol containing time-varying delay factors is designed.Then,benefited from the matrix theory,a synthesis criterion for delay-dependent robust consensus controllers is derived.Finally,based on a cone complementary linearization algorithm,a synthesis algorithm for delay-dependent controllers' robust consensus is proposed.Through this criterion and algorithm,the dynamics of the multi-agent system is guaranteed to achieve coordinated consensus and meets the satisfied index at the same time,hence to make sure the system's stable operation.For the robust time-varying formation control problem of multi-agent systems with time-varying delay and external disturbances,a controller's synthesis algorithm for robust time-varying formation is proposed.First,based on each agent behaviors,a clear-cut definition of robust time-varying formation of multi-agent systems is determined and a distributed control protocol with time-varying delay factors included is designed.Secondly,by mathematical methods like the Newton-Leibniz formulas with free weight matrices,the Lyapunov-Krasovskii functionals,the Jensen inequalities,etc.,a time-delay dependent robust time-varying formation controller synthesis criterion is derived.Finally,based on a cone complementary linearization algorithm,a controller's time-delay dependent synthesis algorithm for robust time-varying formation is proposed,which realizes the coordination and control of multi-agent systems in the formation performances,and receives the final result of achieving the established tasks in a specific formation form.To solve the problem of intelligent coordinated control of multi-agent systems,a cooperative control algorithm based on the intelligent predatory behaviors of wolf colony is proposed.First,the distributed search algorithm is adopted to optimize the points in the space so as to simulate the individual wolves in nature.Then,referring to the leader strategy and the social division of labor and cooperation mechanism embodied in the wolf colony,an intelligent collaborative algorithm based on an improved wolf colony algorithm is proposed.With the advantage of the wireless sensor network to communicate with each other,the integration of the global update strategy and the step length to accelerate the network,and the introduction of the concept of the individual density to describe the distribution density of the wolf colony,the algorithm solves the problem of the traditional wolf group algorithm with too many parameters.Moreover,with the addition of the way of migration in between,the algorithm increases the wolves' inter-communication,hence the autonomy of the multi-agent system in terms of intelligent coordination control is well improved.In view of the problem that the present communication network and perception network of combat system fails to cooperate autonomously with the combat weapon system,considering the organic integration of sensing and combating,on the basis of wireless mobile self-organized network,the concept of "Awareness Combat Network" is proposed and around which,a plane peer to peer topology and network planning of perceived combat network are designed,and a strategy and method of dynamic routing protocols suitable for coordinated control is put forward.A dynamic firepower aiming data link between sensors and shooters in the same combat echelon is set up,thus comes the organic integration of sensor information and target attack and the realization of a high-speed,real-time,dynamic and reliable delivery of data transmission.In order to solve the problem of random cooperative controlling in combat systems,the multi-agent system structure and coordinated control model is set up.Through the setting up of the categories of models and functions of combat coordinated controlling agents,the fundamental method for the multi-agent system's cooperative task decompositions and task allocations is proposed,and analysis of key factors and complexity of random cooperative controlling in combat systems and the establishment and the process of their cooperation is carried out.In view of the cooperative control method and process under the background of the air-ground joint combat system,the cooperative control model and algorithm featuring with the sensors in the air directing the ground fires' joint strike and the cooperative control model and algorithm of the air-ground joint commanding are set up as the major task,and the method and model of random cooperative controlling in united air-ground combats is verified by simulation,which provides a successful example for random cooperative controlling in combat systems of a new type.