【摘要】 提出一种解决多峰连续函数优化的混合算法。该方法定义适当的适应度函数 ,使用遗传算法自动形成极值点邻域种群 ,进行聚类分析 ,分别在单极值区域取最优个体为初始点 ,采用最速下降法进行快速寻优 ,在得到全局最优解的同时获得多个局部极值。计算仿真表明该方法有效且收敛快、精度高。更多还原

【Abstract】 A hybrid algorithm for the optimization of multi\|peak continuous function is put for\|ward. It defines a proper fitness function. Adjacent groups of local optimum are formed auto\|matically by genetic algorithm. After clustering, the excellent populations act as initial points, and the steepest descent algorithm is used to search rapidly. The global optimum and local optimum are obtained simultaneously. Simulation shows that the method is effective. It converges quickly and is of high precision.更多还原