【摘要】 针对基本微分进化算法不适合求解多目标优化这一问题,对基本算法进行了改进。具体采用的方法是对选择步骤进行修正,当且仅当满足一定条件时,新产生的个体才取代原个体。同时,引入动态调整机制控制算法参数,使算法摆脱后期易于陷入局部极值点的束缚,以提高该算法求解多目标问题的效率和寻优能力。算法分析和仿真结果表明,改进算法在实现求解多目标优化问题上具有较好的效果和求解效率。更多还原

【Abstract】 Aiming at the problem that the basic differential evolution algorithm is not suitable for solving multi-objective optimization,the basic algorithm has been improved.The concrete method is to modify the selection procedure;when IF and ONLY IF certain condition is satisfied,the new individuals replace the original individuals;and the parameter of dynamic adjustment mechanism control algorithm is introduced,to make the algorithm break away the trammel of local extreme value point,thus the efficiency and capability of seeking optimum for solving multi-objective problem are enhanced.The results of algorithm analysis and simulation indicate that the improved algorithm features better effect and higher efficiency in solving multi-objective optimization.更多还原