Scheduling of stochastic distributed hybrid flow-shop by hybrid estimation of distribution algorithm and proximal policy optimization

作     者：Luo, Lin Yan, Xuesong 

作者机构：China Univ Geosci Sch Comp Sci Wuhan 430078 Peoples R China Minist Educ Engn Res Ctr Nat Resource Informat Management & Di Wuhan 430074 Peoples R China 

出 版 物：《EXPERT SYSTEMS WITH APPLICATIONS》 (Expert Sys Appl)

年 卷 期：2025年第271卷

核心收录：

学科分类：1201[管理学-管理科学与工程(可授管理学、工学学位)] 0808[工学-电气工程] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：National Key Research and Devel-opment Program of China [2022YFB4501402, 2023YFB4606504] Key Research and Development Program Hubei Province, China [2023BAB065] National Natural Science Foundation of China 

主　　题：Stochastic distributed hybrid flow-shop scheduling problem Estimation of distribution algorithm Proximal policy optimization Processing time perturbation 

摘      要：Flexible manufacturing systems face major challenges in improving productivity when dealing with uncertain factors. Consequently, it is crucial to address production scheduling problems that involve these uncertain elements. In this paper, a hybrid optimization algorithm of estimation of distribution algorithm and proximal policy optimization (EDA/PPO) is proposed to solve the stochastic distributed hybrid flow-shop scheduling problem (SDHFSP) with processing time perturbation in order to minimize makespan. A hybrid initialization strategy is developed to ensure population diversity. In the EDA component, a three-dimensional (3-D) probability matrix corresponding to the solution representation is utilized. For the PPO component, 44 state features are selected to characterize the environmental situation, 7 composite rule-based actions are defined for job sequencing, and unique reward associated with scheduling objective is designed. Through extensive experimentation, we analyze the effects of key parameters and determine optimal numerical combinations. Comparative numerical experiments with existing algorithms demonstrate the effectiveness and robustness of the EDA/PPO approach. This study offers valuable insights for production managers addressing stochastic distributed manufacturing with processing time perturbation.