A Production System Scheduling Method Based on Transfer Reinforcement Learning

Abstract

The invention discloses a production system scheduling method based on migration reinforcement learning, which includes the steps of: (1) defining the state and action of the production system; (2) modeling Q learning of the production system scheduling problem; (3) establishing a production scheduling case library ; (4) Calculate the task similarity between the target task and the source task, and match the knowledge transfer case set; (5) Establish the action mapping relationship between the source task and the target task, and transfer the action knowledge of the source task to the target task; (6) Calculate the similarity between the state of the target task and the state of the cases in the case set, and match the cases of knowledge transfer; (7) map the actions of the selected cases to the actions of the target task, and modify the action selection strategy to realize knowledge transfer; (8) according to the revised After the action selection strategy executes the action, updates the state and Q-value tables. Compared with Q learning, the invention improves the learning performance and effectively solves the real-time problem of online scheduling of production tasks.

Description

technical field [0001] The invention belongs to the field of intelligent robots, and in particular relates to a production system scheduling method based on transfer reinforcement learning. Background technique [0002] With the continuous development of information technology and robot technology, robots are playing an increasingly important role in industrial production. At the same time, with the proposal and development of themes such as "Industry 4.0" and "Made in China 2025", and the current production tends to be more and more small batch, personalized, and the manufacturing cycle is getting shorter and shorter, flexible and intelligent manufacturing is my country's The inevitable trend of manufacturing development. A typical application scenario of flexible intelligent manufacturing is an intelligent manufacturing system composed of intelligent decision-making centers, industrial robots, CNC machine tools, etc. A key issue in the intelligent decision-making center is...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the above technical problems, the present invention provides a production system scheduling method based on transfer reinforcement learning, which can use existing similar cases for knowledge transfer, and overcome the problem of low efficiency in solving complex production scheduling strategies

Images