Evaluation Model of Flue-cured Tobacco Smoking Quality Based on Artificial Neural Network and Its Implementation

Tobacco is an important economic crop that is widely distributed throughout the world. The smoking quality of flue-cured tobacco is an indicator of the comprehensive characteristics of flue-cured tobacco. However, because the chemical composition of flue-cured tobacco is too complex, different experts also have different preferences, leading to uncertainty and subjectivity in the smoking quality of flue-cured tobacco. In this paper, the correlation analysis method was used to determine the chemical components that affect the smoking quality of flue-cured tobacco. The BP artificial neural network was used to establish the artificial neural network model of chemical composition and smoking quality. Finally, artificial neural network model was used to predict the smoking scores of different groups of flue-cured tobacco samples. The results show that the artificial neural network is an effective and reliable method for evaluating the score of flue-cured tobacco, which has guiding significance for the cultivation of flue-cured tobacco and evaluation of the quality of flue-cured tobacco. Based on this method, the evaluation software of flue-cured tobacco smoking quality using the C++ language can create the evaluation of flue-cured smoking quality model and assessment of smoking quality, and it has good practicability.