改进自适应微分进化算法求解全局优化问题

针对微分进化(DE: differential evolution)算法在进化后期收敛速度慢，收敛精度低，易陷入局部最优解等缺点。本文通过改进DE的变异方程，并引入一种新的控制参数自适应策略，提出了一种改进自适应微分进化(IADE: improved adaptive differential evolution)算法。进化过程中IADE将根据个体适应值与父代平均适应值之间的关系动态地调整控制参数。同时，采用10个常用于优化算法比较的标准函数对IADE和其它改进DE算法进行对比试验，实验结果表明IADE算法不仅能够显著地提高收敛速度和收敛精度，而且具有非常好的鲁棒性，从而使得该算法能够满足过程优化的实时性、准确性以及稳定性要求。

Improved adaptive differential evolution algorithm for global optimization Title

Differential evolution (DE) algorithm has some disadvantages, such as slow convergence speed, low convergence precision and easy to fall into local optimal solution in the early stages of the evolution. The paper proposed an improved adaptive differential evolution (IADE) algorithm by improving the mutation equation of DE and introducing a new control parameters adaption strategy. In the process of the evolution, the control parameters will be dynamically adjusted by comparing individual fitness with average fitness of the parent population. Meanwhile, the paper chosen the ten standard functions commonly used for the comparison of optimization algorithm to perform the comparative test of IADE and the other improved DE algorithms, and the experimental results show that IADE algorithm not only can significantly improve the convergence speed and convergence precision, but also has very good robustness, so that IADE algorithm can meet the requirements for the real-time, accuracy and stability of process optimization.