分布式MAS 在飞行冲突解脱中的应用研究

在自由飞行的环境下，为解决飞行冲突探测与解脱（conflict detection and resolution，CDR）问题，提出一种基于高度层、航向和速度调配的综合解脱方法，并将多 agent 系统(multi-agent system, MAS) 的分布式技术与启发式算法相结合，进行问题求解. 首先设计了分布式MAS框架结构，然后建立了飞行冲突探测模型，高度层调配模型及航向、速度调配模型，最后，综合运用了基于合同网协议的分布式算法和自适应遗传算法进行问题求解.仿真实验表明，所设计的MAS框架是可行的，同时分布式算法和自适应遗传算法的综合应用能很快找到基于高度层、航向和速度分配的近似最优解，为CDR问题提供了新的解决思路.

Application of Distributed MAS in Flight Conflict Avoidance

In order to solve the problem of flight conflict detection and resolution (CDR) in the background of free flight, an integrated CDR method based on assignment of flight level, heading and velocity is proposed, and distributed technology of MAS (multi-agent system) and a heuristic algorithm are combined for the algorithm implementation. Firstly, a framework of distributed MAS is designed. Secondly, a conflict detection model, a flight level allocation model and a heading & velocity assignment model are established. Finally, a distributed algorithm based on contract net protocol and an adaptive genetic algorithm are designed to solve the problem. Simulation results show that the MAS framework is feasible, and the combination of the designed distributed algorithm and adaptive genetic algorithm can search the approximate optimal solution rapidly, based on the allocation of flight level, heading and velocity, which provides a new solution to the CDR problem.