Workshop scheduling method, device and system based on deep reinforcement learning

Abstract

The invention discloses a workshop scheduling method, device and system based on deep reinforcement learning, and belongs to the field of job workshop scheduling, and the method comprises the steps: S1, determining a workshop simulation environment according to a target function of a workshop scheduling problem; s2, acquiring an interaction triple lt by using a deep reinforcement learning DRL agent and a workshop simulation environment; s, a, rgt; ; s3, taking the plurality of triads as a training data set to train an agent Actor and an agent Critic; the agent Actor is used for selecting a processing workpiece, and the agent Critic is used for evaluating an expected reward value of a current processing state; and S4, inheriting the network parameters of the trained agent Actor by using the execution agent, and controlling the execution agent to make a decision on the scheduling instance in the workshop processing procedure online so as to determine the next to-be-processed workpiece. The method is a scientific decision based on data driving, is high in decision efficiency, can accurately distribute priorities for the workpieces to be machined, is suitable for various machining scenes, and is high in compatibility.

Description

technical field [0001] The invention belongs to the field of workshop scheduling, and more particularly, relates to a workshop scheduling method, device and system based on deep reinforcement learning. Background technique [0002] The Job Shop Scheduling Problem (JSSP) is a typical scheduling problem in intelligent manufacturing, that is, by rationally arranging the processing sequence of workpieces on each machine to achieve a predetermined objective function, such as minimizing the maximum completion time and minimizing the drag. period etc. A good workshop scheduling method can help enterprises improve production efficiency and resource utilization, so the problem of workshop scheduling has received extensive attention. With the popularity of the Internet, the manufacturing industry has gradually transformed into a multi-variety and small-batch production model, and online orders from customers have gradually occupied the core of some companies' business. This type of ...

AI Technical Summary

Problems solved by technology

[0004] In view of the above defects or improvement needs of the prior art, the present invention provides a workshop scheduling method, device and system based on deep reinforcement learning. The purpose is to obtain training data through the interaction between the agent and the processing environment, and optimize Algorithms and training data are used to train the agent; and then the agent is controlled to directly inherit the scheduling knowledge of the agent in the offline training, so as to make decisions on new scheduling instances in the processing environment; thereby solving the problem of the efficiency of the dynamic online scheduling method in the existing job shop Technical issues with low and poor compatibility

Images