Abstract: An improved Flower Pollinaton Algorithm(FPA)was proposed for solving the integrated scheduling problems of machines and Automated Guided Vehicles(AGVs),in which a new mutation operator was proposed based on the principal component analysis to improve the ability of perceiving the solution space and jumping out of local minima.The crossover operator was introducted to improve the exploration capability of FPA,and an effective self-adaptive initialization strategy based on Chromosome Similarity Matrix(CSM)was proposed to promote the diversity of population.As for the integrated scheduling problems,the integrated scheduling mathematical model of machines and AGVs was established,and then a new decoding algorithm and a new solution accepted mechanism were adopted to solve the integrated model.An integrated scheduling experiment platform was built to verify the improved FPA.

Key words: flower pollination algorithm, automated guided vehicle, principal component analysis, chromosome similarity matrix, new solution accepted mechanism, integrated scheduling

摘要： 针对传统智能算法求解自动导引小车与机器集成的车间调度问题效率低且易早熟的缺点,提出一种基于改进花授粉算法的车间调度算法。其中,基于主成分分析法提出新的变异算子以增强算法对解空间的感知能力和跳出局部极值的能力;引入交叉算子以提高花授粉算法的全局搜索能力;基于染色体相似度矩阵提出一种自适应个体初始化生成策略以提高初始种群的多样性。面向该集成调度问题需求,建立了机器与自动导引小车集成的调度数学模型,然后采用新的解码算法和新解接受机制对模型进行求解。最后通过搭建集成调度实验平台验证了所提改进算法的有效性。

关键词: 花授粉算法, 自动导引小车, 主成分分析, 染色体相似度矩阵, 新解接受机制, 集成调度