改进联盟竞赛算法及其在生产排产中的应用

Improved league championship algorithm and its application in production scheduling

针对基本联盟竞赛算法易陷入局部极小值、收敛速度慢的缺点,提出了一种改进联盟竞赛算法(improved league championship algorithm,ILCA)。该算法可自适应地修正可调参数,平衡了全局收敛和局部收敛能力。同时,设计了全新的参赛队比赛赛程,提升了参赛个体的竞争力,并且引入了降级机制,保证了整个联赛的多样性。采用ILCA算法对标准测试函数进行寻优。结果表明,ILCA的全局搜索性能、收敛速率都明显地优于其他算法,将该算法应用于车间生产排产模型的参数估计,取得了良好的效果。

An improved league championship algorithm （ ILCA） is proposed to avoid the drawbacks of basic LCA, such as easily being stuck to local minima, and slow convergence speed. The parameters of the algorithm are adjus？ted adaptively to balance the global and local convergence capability. A novel match schedule for the sport teams is designed for individual promotion of competitive power. Degradation mechanism is introduced to ensure the diversity of the entire league. Using ILCA to optimize the benchmark function, it is convinced that ILCA is superior to other compared algorithms in the global searching performance and convergence speed. The proposed algorithm is finally applied to parameter estimation of workshop production scheduling model and achieves good results.