CAJ | 학술논문

针对多作战智能体编组协同作战中任务分配的动态优化问题，提出一种基于滚动时域策略的多编组任务分配动态优化方法。以任务执行效率为目标函数，建立了满足个体任务时窗和编组资源损耗约束的问题模型。给出与突发事件特征对应的预测窗口、滚动窗口和滚动驱动机制。设计了一种改进快速模拟退火对优化子问题予以求解，给出解方案表达、邻域解生成、冲突消解等步骤，采用高温随机贪婪搜索、回火技术、禁忌设计和精英保留策略，避免算法陷入局部最优，提高算法的计算效率。案例的仿真计算表明，所建模型和求解方法可以对多编组任务分配计划进行在线优化，并使任务的执行效率始终维持较高水平。
침대다작전지능체편조협동작전중임무분배적동태우화문제，제출일충기우곤동시역책략적다편조임무분배동태우화방법。이임무집행효솔위목표함수，건립료만족개체임무시창화편조자원손모약속적문제모형。급출여돌발사건특정대응적예측창구、곤동창구화곤동구동궤제。설계료일충개진쾌속모의퇴화대우화자문제여이구해，급출해방안표체、린역해생성、충돌소해등보취，채용고온수궤탐람수색、회화기술、금기설계화정영보류책략，피면산법함입국부최우，제고산법적계산효솔。안례적방진계산표명，소건모형화구해방법가이대다편조임무분배계화진행재선우화，병사임무적집행효솔시종유지교고수평。
The aim of this study is to provide a solution based on rolling horizon to dynamical optimization of aerial multiple grouping task allocation.According to the time window attribute of partial tasks and resource ca-pability wastage character of groups,the mathematical model with the value of task executed efficiency maximi-zation objective is built,which satisfy with former constraints.The strategy elements are given,including the prediction time window,rolling time window and driven mechanism.An improved very fast simulated annealing algorithm is developed to solve the sub-optimized problem.Solution coding,neighborhoods creating and conflict solving methods are proposed,then the random greedy search measure at high temperature,re-annealing mech-anism,tabu policy and elitist reserved policy are applied,so that the IVFSA can improve computing efficiency and avoid local optimum solutions.At last,the superiority and applicability of this approach are illuminated by the simulation of assumption,which ensure the stability of task executed efficiency at high level.