Production line scheduling method based on data and model hybrid drive

The invention discloses a production line scheduling method based on data and model hybrid drive, and the method comprises the steps: constructing a flexible production line scheduling model under random disturbance according to the resource position and layout, the number and type of operation orders, random noise disturbance and other elements in a flexible production line; then, according to a data and model hybrid driving framework, solving the scheduling model; in a data and model hybrid driving framework, data driving refers to a value decomposition-based multi-agent reinforcement learning algorithm, model driving refers to a heuristic assignment rule and experience knowledge, and hybrid driving of the data driving and the model driving is mainly reflected in three aspects: firstly, the scheduling rule is used for guiding agent strategy selection; secondly, designing a reward function based on scheduling scene knowledge; thirdly, training a noise detection and denoising module by using historical experi