强化学习技术在工业产品质检调度中的实践  

Practical of Reinforcement Learning Techniques in Industrial Product Quality Inspection Scheduling

作　　者：任靖辉 REN Jinghui(Yunnan Yongchang Silicon Industry Co.,Ltd.,Kunming,Yunnan 650500,China)

机构地区：[1]云南永昌硅业股份有限公司,云南昆明650500

出　　处：《自动化应用》2024年第6期22-24,共3页Automation Application

摘　　要：针对工业产品质检调度问题,讨论了模拟退火算法和Q-Leaning强化学习算法的实践过程。首先描述了本次研究存在的问题,其次抽象出了质检顺序矩阵和质检时间矩阵并进行问题求解。在实际应用中,选择模拟退火算法还是Q-Learning算法取决于问题的特性和需求。若问题具有全局搜索需求,则模拟退火算法可能更适合;若问题可以建模为强化学习问题,则Q-Learning可能是更好的选择。以Q-Leaning对问题进行实践求解,得到工业产品质检调度甘特图,可为实际工业产品质检排程提供参考。In addressing the problem of industrial product quality inspection scheduling,this paper discusses the practical process of using both the simulated annealing algorithm and the Q-Learning reinforcement learning algorithm.It starts with a description of the problem at hand,and the abstraction of quality inspection order matrix and inspection time matrix for problem solving.In practical applications,the choice between the simulated annealing algorithm and Q-Learning depends on the characteristics and requirements of the problem.If the problem necessitates global search,the simulated annealing algorithm may be more suitable.If the problem can be modeled as a reinforcement learning problem,Q-Learning might be the better choice.This paper presents the practical application of Q-Learning to solve the problem,resulting in a Gantt chart for industrial product quality inspection scheduling,which provides reference for real-world industrial product quality scheduling.

关 键 词：调度算法 Q-Leaning 产品质检 模拟退火算法 

分 类 号：TH186[机械工程—机械制造及自动化]