中文摘要：

为了改善人工鱼群算法求解精度较低、容易过早收敛的弱点,提出了一种应用佳点集和反向学习的人工鱼群算法。改进算法在迭代中对当前种群中部分优质个体执行一般动态反向学习,生成它们的反向种群,引导种群向包含全局最优的解空间逼近,以提高算法的平衡和探索能力。当种群的拥挤程度超过阈值λ时,利用佳点集机制对大部分个体重新初始化,以帮助算法脱离局部最优的约束。在六个Benchmark函数上的实验表明,该算法收敛速度快、求解精度高,适合求解函数优化问题。

英文摘要：

Concerning the problem that artificial fish swarm （AFS） algorithm is easy to premature, low solution precision of weakness, this paper proposed an improved AFS algorithm using good-point set （GPS） and opposition-based learning（OBL）. In every iteration, it chose some individuals with better fitting-value execute generic dynamic OBL to generate their opposition search populations, guided search space of the algorithm to approximate the optimum. It was helpful to improve the balance and exploring ability of the AFS. On the other hand, reinitializing most of individuals by GPS to help population avoid the local optimum when the population＇ s degree of jam was exceed A. Some experiments on six classical test functions and the results show that proposed algorithm has high solution precision and rapid convergence speed, suitable to solve the optimization problem on function.