权重自适应clonal选择算法及其应用研究  

Adaptive Weight Clonal Selection Algorithm and Its Applications

作　　者：李勇[1] 王昱[2] 

机构地区：[1]沈阳工业大学特种电机与高压电器部省共建重点实验室,辽宁沈阳110870 [2]沈阳航空工业学院自动化学院,辽宁沈阳110136

出　　处：《控制工程》2011年第1期96-99,共4页Control Engineering of China

基　　金：高等学校博士学科点专项科研基金资助项目(20060145025)

摘　　要：在求解两个目标以上的多目标优化问题时,基于Pareto支配的多目标进化算法多数需要较长的求解时间。基于固定权重的聚合函数方法求解速度快,但要确定一个适合待求解问题的合理权重是十分困难的,为了解决这一问题,将clonal选择算法与权重自适应方法相结合,提出了一种适用于多目标优化问题的权重自适应clonal选择算法。并将权重自适应clon-al选择算法应用于以等功率裕量、轧制能耗及带钢打滑概率作为优化的目标函数的三目标优化模型进行冷连轧轧制规程多目标优化。结果表明,与实际应用的轧制规程相比优化后的轧制规程使功率裕量更加均衡,打滑发生的机率变小,同时也降低了总的轧制能耗。与权重自适应遗传算法相比,该算法能更好的达到预期的优化效果。To the problem that Pareto dominate based multi-objective evolutionary algorithms cost too much running time in multi-objective optimization,and the set of reasonable weights in the fixed weight aggregation methods is hard to design,an adaptive weight clonal selection algorithm suitable for multi-objective optimization problems is proposed,combining the clonal selection algorithm with adaptive weight approach.The adaptive weight clonal selection algorithm is applied to the multi-objective model which has the power distribution,rolling energy consumption and slip rate objective functions.The application result shows that the optimized rolling schedules has the characteristics of more balance power distribution,lower rolling energy consumption and smaller slipping rate compared with the practically applied roll schedules,and can achieve better prospective optimal effect compared to the adaptive weight genetic algorithm.

关 键 词：权重自适应 clonal选择算法 冷连轧机 轧制规程 多目标优化 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]