基于速度概率和自适应速度值的差分进化算法

Improved binary differential evolution algorithm based on velocity probability and adaptive speed value

针对于离散差分进化算法在问题规模较大情况下难以找到全局最优值和收敛速度慢的问题,通过引入速度概率和自适应速度值,提出了一种改进的二进制差分进化算法。通过理论推导,改进的差分进化算法可以有效提高离散差分进化算法对于复杂问题的全局最优值搜索能力和收敛速度。在使用经典0-1背包问题进行的实验中,验证了理论推导的结论正确性以及改进的差分进化算法可行性。

Aiming to discrete differential evaluation algorithm difficult to find the global optimization value and slow in conver- gence rate in large-scale problem, by introducing the velocity probability and adaptive speed value, an improved differential evolu- tion algorithm based on velocity probability and adaptive speed value is proposed. Though theoretical derivation, the correctness of improved discrete binary differential evolution algorithm is proofed. The result of the theoretical derivation is verified by the experiment.