CAJ | 학술논문

针对坦克分队遂行不同类型作战任务时会选取不同准则评估作战目标的特点，选定两类准则目标评估：价值评估准则和威胁评估准则，并确定各准则的评估指标。依据径向基函数（RBF）神经网络的快速性和唯一最佳逼近性及遗传算法（GA）的全局搜索能力和并行处理能力，设计了RBF神经网络遗传算法（RBF-GA），用于评估坦克分队作战目标。实例分析表明，该算法不仅满足目标评估技术指标要求，还具有较高实用价值。
침대탄극분대수행불동류형작전임무시회선취불동준칙평고작전목표적특점，선정량류준칙목표평고：개치평고준칙화위협평고준칙，병학정각준칙적평고지표。의거경향기함수（RBF）신경망락적쾌속성화유일최가핍근성급유전산법（GA）적전국수색능력화병행처리능력，설계료RBF신경망락유전산법（RBF-GA），용우평고탄극분대작전목표。실례분석표명，해산법불부만족목표평고기술지표요구，환구유교고실용개치。
The value evaluation criterion and the threat evaluation criterion ware chosen based on the combat missions that tank unit carried out,and the evaluation indexes ware determined accordingly. The radial basis function neural network genetic algorithm(RBF-GA)is designed based on the rapidity and the optimal approximation ability of RBF neural network and the overall search capability and the parallel processing ability of genetic algorithm to evaluation combat target of tank unit. The example shows that the algorithm not only meets the demands of evaluation technical indicators,but also has a high practical value.