Migration optimization method of decision-making model for industrial process optimization based on knowledge distillation

Abstract

This application relates to the migration optimization method of industrial process optimization decision-making model based on knowledge distillation, which uses knowledge distillation technology to transfer the knowledge in complex models to simple models, and establishes a teacher-student network. The teacher network is defined as a complex model with powerful capabilities and performance, student networks are defined as simple models that express more compactly. Through knowledge distillation, the student network model can approach or exceed the teacher network model as much as possible, so as to obtain similar prediction effects with less complexity, and realize the knowledge transfer from the complex model teacher network to the simple model student network.

Description

technical field [0001] The present application relates to the field of artificial intelligence, in particular to a knowledge distillation-based migration optimization method for industrial process optimization decision models. Background technique [0002] In recent years, with the development of deep learning and computing power, neural network models have been widely used in image classification, object recognition, fault diagnosis and other fields. Generally, when solving optimization decision-making problems, people tend to design more complex convolutional neural networks to collect more data in order to obtain better results. However, as the complexity of the model increases, the number of model parameters increases, the scale of the model and the floating-point numbers required for calculation become larger and larger, which imposes higher requirements on hardware resources (such as memory and CPU), which is not conducive to the model Deploy and use on devices with l...

AI Technical Summary

Problems solved by technology

However, the optimization decision-making knowledge reasoning model embedded in domain rules has superior reasoning performance, but it is relatively complex, which makes it unfavorable for real-time online optimization decision-making in industrial processes, and cannot be effectively applied in industrial process production sites

Images