Abstract: Given the low capability-demand alignment,the difficulty of optimizing combinations under multiple constraints,and the low efficiency in optimization,a service combination optimization method based on the improved Dandelion Optimizer (DO) algorithm is proposed for cloud manufacturing. After analyzing the capabilities of cloud manufacturing service providers and the quality attributes of services,the Analytic Hierarchy Process (AHP) is utilized to normalize and aggregate various attribu-tes. Subsequently,an improved DO algorithm is employed to solve the service combination problem and obtain the optimal service combination solution.In terms of improving the DO algorithm,diversity of population particles is enhanced by introducing the Tent chaotic mapping. Additionally,the convergence speed of the algorithm is improved by incorporating reverse learning mechanisms,counters,and mutation concepts,thus preventing premature convergence. Finally,through simulation experiments and comparative analysis with classical dandelion optimizer algorithm and other improved algorithms such as particle swarm optimization,genetic algorithm,and polar bear optimization proposed in relevant literature on service combination,the effectiveness and stability of the proposed algorithm in cloud manufacturing service combination optimization are verified.

Key words: cloud manufacturing, service composition, dandelion optimizer algorithm, chaotic mapping, reverse learning