一种带有随机变异的动态差分进化算法

Dynamic differential evolution algorithm with random mutation

提出一种带有随机变异的动态差分进化算法。在这个算法中,两种不同的变异策略DE/rand/1和DE/best/1通过线性递减加权组合策略产生新的变异策略,以便动态利用DE/rand/1和DE/best/1的优点,并且引入一种指数递增交叉概率算子、线性递减缩放因子和一种随机变异机制以进一步提高算法的全局寻优能力。通过四个标准测试函数的测试结果表明,该算法是一种收敛速度快、求解精度高、鲁棒性较强,更适合求解高维复杂的全局优化问题。

The dynamic Differential Evolution （DE） algorithm with random mutation was proposed. In this algorithm, the mutation strategies of DE/rand/l and DE/best/1 were combined by linear dec, reased weight convex combination strategy to produce a new mutation strategy so as to dynamically .use the advantages of DE/rand/1 and DE/best/1. In order to improve the global optimization abilily of DE algorithm, exponent increased crossover probability operator, linear decreasing scaling factor amt random mutation mechanism were introdunced. The test results on the four standard test functions show that the new algorithm has fast convergence, high accuracy and more robustness, more suitable to solve high-dimensional complex global optimization problems.