Differential evolution algorithm in solving complex functions with high dimensions has shown some weaknesses, such as low convergence speed and low precision. Therefore a new self-adapting differential evolution algorithm with subarea crossover was proposed. Two operators based on Cauchy distribution random number were adapted and their associated control parameter values were gradually self-adapted in this algorithm. The evolution process was divided into two subareas with different configuration values for operators. At the same time, to enhance the convergent speed,a new mutation strategy was introduced to guide searching direction. Benchmark problems were used to verify this algorithm. The result of simulation indicates that it is improved significantly both in convergence speed and precision of optimization. The algorithm also provides a solution to deal with the constrained optimization problems and the feasibility is also verified by two examples.%差分进化算法处理复杂高维优化问题时存在收敛速度慢和精度不高的缺陷,为此提出了分区交叉差分进化算法.利用柯西分布随机数设计两个动态算子,分别生成缩放因子和交叉因子用于进化中,并对进化进行合理的分区,不同区段根据不同的配置利用算子生成相应的交叉因子.同时为了加快收敛速度,采用了新的变异策略,对寻优的方向加以引导.对经典Benchmark函数进行了仿真测试,结果显示,本算法的收敛速度与优化准确率均有显著提高.同时提供了算法处理约束问题的解决方案,并检验了方案的可行性.
展开▼
