摘要
设计一种新的混合蚁群算法。该算法以一种新的二进制蚁群算法为基础,混合PBIL(population based incremental learning)算法及遗传算法的交叉操作和变异操作,从而大大提高了种群的多样性及收敛速度,改善全局最优解的搜索能力。通过函数优化测试,表明该算法具有良好的收敛速度和稳定性,最后用于有机物毒性的QSAR研究中,取得较好效果。
A kind of new hybrid ant colony algorithm was designed. It uses a new binary ant colony algorithm as the basis, combining with PBIL and the crossover operation and the mutation operation of GA, thus greatly raising its population polymorphism and speedy convergence rate, and improving the searching ability of the overall optimal solution. Functions optimization tests show that the hybrid algorithm has fine convergence rate and stability. Good results are obtained by applying this algorithm to the modeling of QSAR for toxicities of organic substance.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2009年第5期549-552,共4页
Computers and Applied Chemistry
基金
国家863计划(2007AA04Z171)
上海市重点学科建设项目(编号:B504)
上海市自然科学基金(06ZR14027)
关键词
蚁群算法
PBIL
神经网络
QSAR
ant colony algorithm （ACA）, population based incremental learning （PBIL）, neural network, QSAR
