Abstract:  Fireworks Algorithm（FWA） is one of the latest intelligence algorithms, which is inspired by the explosion process of fireworks. The optimization procedure of FWA includes two regimes: the creation of explosion sparks, which can realize the local and global search process; the creation of Gaussian explosion sparks, which enhances the diversity of group and guarantees that the information of the excellent individuals can be used in the next iteration. In this paper, four parameter tests are used to detect the optimized parameters of FWA for solving Job-shop Scheduling Problem（JSP）. The results show the fittest parameters. Finally, the comparison between FWA and other algorithms for JSP benchmark tests implies the validity and robustness of the FWA for solving JSP.

Key words: job-shop scheduling, firework algorithm, explosion sparks, Gaussian explosion sparks

摘要： 烟花算法是一种新型智能优化算法，该算法模拟烟花在空中爆炸产生火花这一过程。烟花算法的求解过程包含两种机制：产生爆炸火花，从而实现算法的局部和全局寻优过程；产生高斯变异火花，从而增加种群的多样性以便将优良个体遗传到下一代。通过设计四个参数实验，分析了主要参数对算法求解能力的影响，找出求解作业车间调度问题的较优参数。最后通过对作业车间调度的标准问题进行仿真对比实验，证明了烟花算法求解作业车间调度问题的有效性和稳定性。

关键词: 作业车间调度问题, 烟花算法, 爆炸火花, 高斯变异火花