Article: A cooperative evolutionary algorithm based on a double deep Q-network for adaptive robust service composition in cloud manufacturing Journal: International Journal of Data Mining and Bioinformatics (IJDMB) 2026 Vol.30 No.7 pp.51 - 83 Abstract: To address the challenge of collaboratively balancing service quality and system robustness arising from the dynamic and uncertain nature of cloud manufacturing environments, a cloud manufacturing service composition optimisation model is constructed in which QoS weights adapt to real-time operating conditions, an uncertainty penalty discourages unstable compositions, and failure-related parameters are treated as bounded uncertain quantities evaluated under worst-case realisations. A coevolutionary approach (DDQCE) that involves a double deep Q-network is used to solve the model. The proposed algorithm adopts a main population-elite archive coevolution framework combined with four targeted local search operators. In addition, a double deep Q-network is integrated to enable intelligent selection of local search operators during the evolutionary process. Experimental results demonstrate that the DDQCE algorithm outperforms representative traditional multi-objective evolutionary algorithm in terms of key performance indicators, including convergence, diversity, and solution set quality. Furthermore, a case study on aerospace impeller manufacturing validates the effectiveness of the algorithm in practical cloud manufacturing scenarios. The results provide theoretical foundations and technical support for the development of intelligent and robust cloud manufacturing service platforms. Inderscience Publishers - linking academia, business and industry through research

Title: A cooperative evolutionary algorithm based on a double deep Q-network for adaptive robust service composition in cloud manufacturing

Authors: Jianjia He; Yan Wang; Xi Cheng; Zhuoyu Zhang

Addresses: Business School, University of Shanghai for Science and Technology, Shanghai 200093, China; School of Intelligent Emergency Management, University of Shanghai for Science and Technology, Shanghai 200093, China; Centre for Super Network Research (China), University of Shanghai for Science and Technology, Shanghai 200093, China ' Business School, University of Shanghai for Science and Technology, Shanghai 200093, China ' Business School, University of Shanghai for Science and Technology, Shanghai 200093, China ' Business School, University of Shanghai for Science and Technology, Shanghai 200093, China

Abstract: To address the challenge of collaboratively balancing service quality and system robustness arising from the dynamic and uncertain nature of cloud manufacturing environments, a cloud manufacturing service composition optimisation model is constructed in which QoS weights adapt to real-time operating conditions, an uncertainty penalty discourages unstable compositions, and failure-related parameters are treated as bounded uncertain quantities evaluated under worst-case realisations. A coevolutionary approach (DDQCE) that involves a double deep Q-network is used to solve the model. The proposed algorithm adopts a main population-elite archive coevolution framework combined with four targeted local search operators. In addition, a double deep Q-network is integrated to enable intelligent selection of local search operators during the evolutionary process. Experimental results demonstrate that the DDQCE algorithm outperforms representative traditional multi-objective evolutionary algorithm in terms of key performance indicators, including convergence, diversity, and solution set quality. Furthermore, a case study on aerospace impeller manufacturing validates the effectiveness of the algorithm in practical cloud manufacturing scenarios. The results provide theoretical foundations and technical support for the development of intelligent and robust cloud manufacturing service platforms.

Keywords: cloud manufacturing; service composition; adaptive; robust optimisation.

DOI: 10.1504/IJDMB.2026.154767

International Journal of Data Mining and Bioinformatics, 2026 Vol.30 No.7, pp.51 - 83

Received: 19 Jan 2026
Accepted: 15 Apr 2026
Published online: 13 Jul 2026 *

Title: A cooperative evolutionary algorithm based on a double deep Q-network for adaptive robust service composition in cloud manufacturing

Keep up-to-date