兵工自动化
兵工自動化
병공자동화
Ordnance Industry Automation
2015年
11期
4-8
,共5页
张帅%糜玉林%徐吉辉%孙媛
張帥%糜玉林%徐吉輝%孫媛
장수%미옥림%서길휘%손원
消耗件%联合补充问题%模糊资源约束%自适应遗传算法
消耗件%聯閤補充問題%模糊資源約束%自適應遺傳算法
소모건%연합보충문제%모호자원약속%자괄응유전산법
consumable items%joint replenish problem%fuzzy resource constraint%adaptive genetic algorithm
针对不确定环境下飞机消耗件联合补充的问题,建立模糊资源约束的消耗件联合补充模型。以最小化费用为目标,运用模糊数学规划,将模糊约束模型转换为确定约束模型。为提高模型的求解速度和效果,提出综合考虑种群特征、个体特征和种群进化阶段特征的改进自适应遗传算法,并通过仿真实例,对比分析模糊约束模型和确定约束模型的结果,以及一般遗传算法和改进自适应遗传算法的性能。结果表明:改进自适应遗传算法能保持种群多样性,具有良好的收敛性,算法搜索速度快,寻优能力强,其求解的精度和速度均优于一般遗传算法。
針對不確定環境下飛機消耗件聯閤補充的問題,建立模糊資源約束的消耗件聯閤補充模型。以最小化費用為目標,運用模糊數學規劃,將模糊約束模型轉換為確定約束模型。為提高模型的求解速度和效果,提齣綜閤攷慮種群特徵、箇體特徵和種群進化階段特徵的改進自適應遺傳算法,併通過倣真實例,對比分析模糊約束模型和確定約束模型的結果,以及一般遺傳算法和改進自適應遺傳算法的性能。結果錶明:改進自適應遺傳算法能保持種群多樣性,具有良好的收斂性,算法搜索速度快,尋優能力彊,其求解的精度和速度均優于一般遺傳算法。
침대불학정배경하비궤소모건연합보충적문제,건립모호자원약속적소모건연합보충모형。이최소화비용위목표,운용모호수학규화,장모호약속모형전환위학정약속모형。위제고모형적구해속도화효과,제출종합고필충군특정、개체특정화충군진화계단특정적개진자괄응유전산법,병통과방진실례,대비분석모호약속모형화학정약속모형적결과,이급일반유전산법화개진자괄응유전산법적성능。결과표명:개진자괄응유전산법능보지충군다양성,구유량호적수렴성,산법수색속도쾌,심우능력강,기구해적정도화속도균우우일반유전산법。
To deal with the aircraft consumable items joint replenishment problem with fuzzy resource constraint, the consumable items joint replenishment model with fuzzy resource constraint was constructed. To minimize the cost as the target, the fuzzy constraint model was converted to deterministic constraint model by using of fuzzy mathematical programming. In order to improve the rate and effect of resolving the model, the improved adaptive genetic algorithm which took into account the population characteristics, the individual characteristics, and the population evolution characteristics, was proposed. Through simulation examples, the comparisons between the fuzzy constraint model and deterministic constraint model were made, and also the general genetic algorithm and the improved adaptive genetic algorithm. The results showed that the improved adaptive genetic algorithm could keep the population diversity and had good convergence. The algorithm had fast search and optimization capability, and its precision and speed of solving were superior to general genetic algorithm.