一种混合改进的鸡群优化算法

摘要：针对鸡群算法（CSO）易陷入局部最优和出现早熟收敛的缺陷，提出了一种混合改进的鸡群优化算法（OBSA-CSO）。算法通过采用反向学习对种群进行初始化，同时对越界个体进行边界变异操作保证了算法的种群多样性并利于算法的全局搜索；在寻优过程中对母鸡采用新的位置更新公式，并对最优个体采用改进退温函数的模拟退火扰动，通过Metropolis准则进行择优的方式有效的提高了算法的寻优精度和收敛速度。通过对7个测试函数在固定迭代次数和固定寻优精度条件下的实验表明，改进后的算法相较于传统鸡群算法具有较好的寻优精度和收敛速度。

A hybrid improved for Chicken Swarm Optimization Algorithm

Abstract: To resolve the defect the chicken swarm algorithm (CSO) is easy to fall into local optimum and premature convergence, a hybrid improved chicken swarm optimization algorithm(OBSA-CSO) is proposed. Firstly, through opposition learning to initialize the population, and using the boundary mutation deal with the cross-border individuals, which ensures the diversity of the algorithm and facilitates the global search of the algorithm; Secondly, update the hen with a new location update formula and by using simulated annealing with improved annealing model to the best individual, preferred by the way of Metropolis criterion can effectively improve the searching precision and convergence speed; Finally, the experimental results show that the improved algorithm has better search precision and convergence speed compared with the traditional chicken swarm algorithm through experiments on seven classical test functions.