CAJ | 학술논문

针对代价函数权重需要根据环境变化而变化的问题，结合飞行约束条件提出归一化的代价函数，当环境发生变化时，不用再修改代价函数，增强了算法的鲁棒性。为了弥补传统定步长寻径算法耗时长的缺陷，设计了一种基于B样条曲线与遗传算法的高时效寻径算法。利用遗传算法在地图中所寻合适的控制点，再结合B样条曲线生成航路。为了增强遗传算法的全局搜索能力，遗传算法中加入振动法则，使得种群在进化中后期依旧保持一定的多样性。仿真结果表明该算法与精英蚁群算法相比，规划时间大幅缩短；与振动遗传算法相比，航路代价明显降低。
침대대개함수권중수요근거배경변화이변화적문제，결합비행약속조건제출귀일화적대개함수，당배경발생변화시，불용재수개대개함수，증강료산법적로봉성。위료미보전통정보장심경산법모시장적결함，설계료일충기우B양조곡선여유전산법적고시효심경산법。이용유전산법재지도중소심합괄적공제점，재결합B양조곡선생성항로。위료증강유전산법적전국수색능력，유전산법중가입진동법칙，사득충군재진화중후기의구보지일정적다양성。방진결과표명해산법여정영의군산법상비，규화시간대폭축단；여진동유전산법상비，항로대개명현강저。
Concerning the weight of cost function has to change with the environment ,a normalized cost function is designed with flight constraints in this paper,which could improve the robustness of the algorithm since there is no need to modify the cost function when the environment is changed. A high timeliness routing algorithm is proposed which is based on B-spline curve and Genetic Algorithm(GA) to reduce the time cost of traditional fixed step algorithms. First,the control points are searched by GA in the map. Then the whole path is produced by B-spline curve with control points. An appropriate vibrantion law is added in order to enhance the global search ability of GA so that the population still maintains a certain diversity in the evolution of the late. Simulation result shows that the method is much faster than Elite Ant Algorithm and the cost of flight route is obviously lower than that of Vibrational Genetic Algorithm.