【摘要】 本文建立了一种约束优化的演化模型,并构造出求解此模型的多种群空间收缩遗传算法,将信息熵概念引入进化过程,控制各种群寻优搜索时解空间的收缩 本算法用种群的多样性避免遗传进化的早熟现象,并以空间收缩尺度作为停机判据,有效地控制了算法的收敛 利用基于小种群的多种群进化策略,在保证种群多样性的前提下,极大程度地减少了计算量,提高了计算效率 数值算例表明,熵的介入增强了随机搜索类进化算法的寻优目的性,使收敛过程平稳且迅速 算例表明此算法能有效地应用于药物分子对接设计更多还原

【Abstract】 Drug molecular docking design is an ideal approach to compound virtual screening in large databases. So the efficiency of search algorithm becomes a critical problem. An ectropy-based multi-population micro genetic algorithm is presented to find the lowest energy conformation in this paper. The docking problem is modeled by a minimization optimization problem with multiple constraints. An ectropy-based optimization model is constructed to obtain explicit solution of the narrowing coefficients of the searched space for multi-population evolution. Then a new iteration scheme in conjunction with multi-population genetic strategy and an entropy-based searching technique is developed to solve the optimization problems with constraints. The elitist maintaining strategy and efficient convergent rule are used to ensure the global sulution. Application in molecular docking is given to demonstrate the effectiveness of the proposed docking method.更多还原