采用半初始化和概率扰动策略改进的遗传算法

Improved genetic algorithm using semi-initialization and probabilistic disturbance strategy

针对遗传算法在函数寻优过程中收敛速度慢、易陷入局部最优解的问题,提出一种采用半初始化和概率扰动策略改进的遗传算法DIAGA。首先,通过引入概率扰动策略增加了算法迭代后期的种群多样性,采用半初始化从根本上改变了算法在全局最优解比较过程中的局限性;然后利用马尔可夫链理论证明了DIAGA的收敛性;最后,对六个标准测试函数进行仿真测试。仿真实验结果表明,DIAGA有效摆脱了局部收敛,在搜索精度、收敛速度上具有明显优势,就多维测试函数而言,寻优精度提高了约29%。

Aimed at the problem that the genetic algorithm is slow in convergence and easy to fall into the local optimal solution in the process of function optimization,this paper proposed a genetic algorithm called DIAGA with semi-initialization and probability perturbation strategy. Firstly,it introduced the probabilistic perturbation strategy to increase the population diversity in the late iteration of the algorithm,and used the semi-initialization fundamentally to change the limitation of the algorithm in the comparison process of the global optimal solution. Then this paper used the Markov chain theory to prove the convergence of DIAGA. Finally,it performed simulation experiments on six standard test functions. Simulation results show that the DIAGA algorithm is more effective in getting rid of local convergence and has more obvious advantages in search accuracy and convergence speed. For multidimensional test functions,the optimization accuracy has increased by about 29%.