Workshop scheduling method based on multi-Agent global and local optimization combination

Abstract

The invention relates to a workshop scheduling method based on the combination of multi-Agent global and local optimization, and the method comprises the steps: S1, carrying out the modeling of a workshop dynamic scheduling process through a multi-agent method, and obtaining a plurality of agents; s2, each agent performs independent learning and decision making by combining a roulette probabilistic algorithm according to the knowledge of a local execution task and Q learning in reinforcement learning, and performs local scheduling; and S3, performing global optimization on the learning result of local scheduling of each agent according to the improved differential evolution algorithm, expanding the search space by adopting a decreasing variation factor, and meanwhile, obtaining a global optimization scheduling strategy by adopting a crossover operator which is dynamically adjusted along with the number of iterations. Compared with the prior art, the method has the advantages that dynamic scheduling is carried out on flexible workshop disturbance with small batches, multiple varieties and changeable processes, localized learning and decision making are carried out on Agents by utilizing reinforcement learning, global optimization decision making is carried out by adopting an IDE algorithm, and the efficiency and effectiveness of a scheduling model are improved.

Description

technical field [0001] The invention relates to a workshop scheduling method, in particular to a workshop scheduling method based on the combination of multi-agent global and local optimization. Background technique [0002] Workshop scheduling is a key link in production control, and the results of scheduling directly affect workshop production efficiency. With the rise of production modes such as small batches and multiple varieties, the randomness of customer orders, the variability of processes brought about by production tasks, as well as machine failures on the production site, and the absence of workers, these disturbances make the scheduled plan inefficient or even ineffective. Plans cannot be executed smoothly and can even lead to blockages in the machining process. The traditional static scheduling control method, or the scheduling method based on heuristic search, cannot adapt to this kind of real-time and dynamic manufacturing process to a certain extent. A rea...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is to provide a method based on multi-Agent global and local in order to overcome the shortcomings of the workshop scheduling method in the above-mentioned prior art t

Images