Reinforcement-learning-based tobacco leaf baking curve recommendation method

Abstract

The invention relates to a reinforcement-learning-based tobacco leaf baking curve recommendation method. The method comprises: a tobacco leaf baking request is obtained; a recommended model of a tobacco leaf baking curve is obtained; whether the tobacco leaf baking request is null is determined; if so, the recommended model is initialized to form an actual tobacco leaf baking curve; if not, whether the tobacco leaf baking request has newly added data is determined; if so, reinforcement learning is carried out on the recommended model to form an actual tobacco leaf baking curve; and if not, anoptimal baking curve is recommended based on the tobacco leaf baking request. According to the invention, different baking curves are recommended based on different baking requests, wherein the data in the baking request are consistent with historical baking curve data; initial training is carried out on the recommended model; reinforcement learning is carried out on the recommended model when thenew data are added; a user inputs related parameters of a to-be-baked tobacco leaf; and an optimal baking curve is recommended. Therefore, the fresh green tobacco leaves are baked to form dried orange-yellow flue-cured tobacco leaves to the greatest extent, so that the quality of the flue-cured tobacco is ensured.

Description

technical field [0001] The invention relates to a flue-cured tobacco curing process, in particular to a tobacco leaf curing curve recommendation method based on reinforcement learning. Background technique [0002] The curvilinear curing chart (that is, the tobacco leaf curing curve) is a technical chart that has been used in the tobacco leaf curing industry. This chart guides the tobacco leaf curing. The existing tobacco leaf curing curve follows the three-stage theory of flue-cured tobacco, that is, through the Curve to control the temperature and humidity in the barn, and control the temperature and humidity in the barn according to the dry bulb temperature, wet bulb temperature and empty time duration required by the curve, so as to bake fresh green tobacco leaves into dry orange-yellow flue-cured tobacco. [0003] The existing tobacco leaf curing curve cannot be fine-tuned according to different situations, that is, the curing process curve is already fixed and cannot ...

AI Technical Summary

Problems solved by technology

[0003] The existing tobacco leaf curing curve cannot be fine-tuned according to different situations, that is, the curing process curve is already fixed and cannot be changed, and it is impossible to push a reasonable curing curve according to the actual situation of the user. If the fixed curing process is always used Curve, it will cause some special tobacco leaves cannot be fully baked into flue-cured tobacco, resulting in waste

Images