漳州职业技术学院学报
漳州職業技術學院學報
장주직업기술학원학보
JOURNAL OF ZHANGZHOU TECHNICAL INSTITUTE
2011年
4期
15-18
,共4页
人工神经网络%遗传算法%配合比优化设计
人工神經網絡%遺傳算法%配閤比優化設計
인공신경망락%유전산법%배합비우화설계
artificial neural network%genetic algorithm%optimum design of mix proportion
采用遗传算法对混凝土进行配合比优化设计,以经济成本为目标函数、以人工神经网络混凝土性能预测模型为约束条件,用实数编码表示个体的染色体位串,通过选择、交叉、变异等遗传算子作用,完成遗传操作,以优胜劣汰为原则,最终达到最优解。将优化结果和工程实例的配合比进行对比,可见该方法可降低造价、节约能源,具有较好的工程使用价值。
採用遺傳算法對混凝土進行配閤比優化設計,以經濟成本為目標函數、以人工神經網絡混凝土性能預測模型為約束條件,用實數編碼錶示箇體的染色體位串,通過選擇、交扠、變異等遺傳算子作用,完成遺傳操作,以優勝劣汰為原則,最終達到最優解。將優化結果和工程實例的配閤比進行對比,可見該方法可降低造價、節約能源,具有較好的工程使用價值。
채용유전산법대혼응토진행배합비우화설계,이경제성본위목표함수、이인공신경망락혼응토성능예측모형위약속조건,용실수편마표시개체적염색체위천,통과선택、교차、변이등유전산자작용,완성유전조작,이우성렬태위원칙,최종체도최우해。장우화결과화공정실례적배합비진행대비,가견해방법가강저조개、절약능원,구유교호적공정사용개치。
The paper developed an optimum design method of concrete using genetic algorithm, which tookeconomical cost as object function restriction conditions. Based on the and the artificial neural network prediction models of concrete properties as rule of survival of the fittest, the genetic algorithm was finally obtained by using real number codes for bit strings of chromosome and arithmetic operators such as selection, cross, mutation. It could be seen by comparing the optimum results with the engineering case, that the method developed could decrease the cost and economize the energy consumption, having a good application in engineering.