Manufacturing Service Supply-Demand Optimization with Dual Diversities for Industrial Internet Platforms

Ruiyang Hao,Ying Cheng,Yongping Zhang,Fei Tao

DOI: https://doi.org/10.1016/j.cie.2021.107237

IF: 7.18

2021-01-01

Computers & Industrial Engineering

Abstract:Industrial internet has been put forward to achieve interconnection between physical world and cyber space towards smart manufacturing, where geographically distributed manufacturing resources are centralized to achieve resource sharing and collaboration among manufacturing enterprises in the form of manufacturing services (MSs). However, MS supply-demand types are diversified and the quantity of MS supply and demand is dynamic, which would pose challenges to MS supply-demand matching and the operational stability of industrial internet platforms. A classified hypernetwork is proposed to depict the dual diversities in both types and quantity variation of MS supply-demand, and the supply-demand ratio of manufacturing services is defined as a main index to reflect the carrying capability of platforms. Then the problem of manufacturing service supply-demand optimization (MS-SDO) is modeled to improve the platforms' operational stability and to satisfy different stakeholders in the platforms, in which the impacts of supply-demand ratio on stakeholders' utility and platforms' operational stability are fully analyzed. Afterwards, a knowledge-guided NSGA-II algorithm is designed to address the issue of MS-SDO, where traditional genetic procedures are modified to suit our issue and a knowledge-guided search approach is used to improve the search performance. Finally, numerous experiments are carried out for algorithm validation and issue analysis, with a simulation system constructed. Experimental results indicate that the majority of Pareto solutions by our proposed algorithm can dominate the solutions by the traditional algorithm, and some implications are concluded for the platform management and the service integration.