事件触发机制的多弹分布式微分对策制导律  

Event-triggered-based distributed differential game guidance laws

作　　者：席阿行 蔡远利[1] XI A-xing;CAI Yuan-li(Control Engineering Research Institute,Xi’an Jiaotong University,Xi’an 710049,China)

机构地区：[1]西安交通大学控制工程研究所,西安710049

出　　处：《控制与决策》2024年第11期3745-3754,共10页Control and Decision

基　　金：国家自然科学基金项目(62203349);创新特区项目(23-TQ01-04-ZT-01-011)。

摘　　要：为了减少多弹协同控制系统的计算负担与通信资源,引入事件触发机制,研究分布式微分对策制导律.首先,将分布式微分对策制导问题转化为分布式零和博弈问题,利用自适应动态规划求解其事件触发的耦合Hamilton-Jacobi-Bellman方程,同时设计事件触发机制的触发条件和自适应评价网络权重的更新律,提出一种事件触发机制的分布式微分对策制导律;然后,依据Lyapunov定理证明多弹协同系统的稳定性和评价网络估计权重误差的最终一致有界;进而,设计最小采样时间间隔以确保系统不出现Zeno行为;最后,通过仿真实例验证所提出制导律的有效性.结果表明,与传统的分布式微分对策制导律相比,所提出的制导律对控制器状态更新次数可减少65.36%,极大地减少了控制系统的计算负担与通信资源.To reduce computational loads and save communication resources of cooperative multi-missile systems based on the event-triggered mechanism,the distributed differential game guidance laws are investigated.First,the cooperative multi-missile interception problem is converted into multi-player zero-sum differential games.The proposed guidance laws are derived by addressing the couple event-triggered-based Hamilton-Jacobi-Bellman equation,which is solved by using the adaptive dynamic programming technique.Meanwhile,adaptive weight tuning laws and event-triggered conditions are designed.The proposed guidance laws can guarantee the stability of the nonlinear multi-missile system,which is proved using the Lyapunov method.Then,the estimated weights of the neural network and consensus tracking errors are ensured to be uniformly ultimately bounded.Besides,a minimal inter-sample time is established for avoiding the Zeno behavior.Finally,some experiments are provided to illustrate the effectiveness of the proposed guidance laws.Compared with traditional distributed differential game guidance laws,the proposed guidance law reduces the number of state updates of the cooperative system controller by 65.36%.The computational burden and communication resources of the control system can be greatly reduced.

关 键 词：多弹协同制导 事件触发机制 分布式控制 零和博弈 自适应动态规划 HAMILTON-JACOBI-BELLMAN方程 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]