一种求解机组组合问题的改进免疫算法

An improved immune algorithm for unit commitment

针对传统免疫算法存在的两大缺陷:容易陷入局部最优平衡状态、进化后期搜索停滞不前。提出一种改进的免疫算法,用模糊推理来动态改变交叉、变异概率,同时把模拟退火的思想引入到算法中,采用确定性和模拟退火相结合的方法选择接种个体。实验证明,改进的免疫算法用于求解机组组合问题时,不仅不易陷入局部最优解,而且它的收敛性和效率都有所提高。

Aiming at the two main defects of traditional immune algorithm： one is easy to fall into local optimal equilibrium states, the other is search stagnation at evolutionary late stage. An improved immune algorithm is presented, which dynamically adjusts the ratio of crossover and mutation by fuzzy reasoning. The idea of simulated annealing is introduced into the algorithm, the combinative method of certainty and simulated annealing is used to select the inoculated individual. Experiments prove that when the improved immune algorithm is applied to unit commitment, it not only can＇ t fall into a local optimal solution but also can improve the convergence and efficiency.