Method for predicting main index of flue-cured tobacco flume

Abstract

The invention discloses a method for predicting the main indexes of flue-cured tobacco smoke, which comprises the following steps: firstly, establishing a model for the main indexes of the flue-cured tobacco smoke; secondly, detecting the main chemical compostions and the main indexes of the smoke of a training flue-cured tobacco sample and a testing flue-cured tobacco sample; thirdly, using a gray function clustering method to cluster the training flue-cured tobacco sample; fourthly, determining the variable of the sample entering the model; fifthly, establishing a gray prediction model between the main indexes and the main chemical compositions of the smoke of each type of the sample; sixthly, establishing a BP neural network for each type of the training sample; and seventhly, performing a test and an adjustment by using the main indexes and the main chemical compositions of the smoke of the testing flue-cured tobacco sample. In the process of detection, the main chemical components of the flue-cured tobacco sample to be tested are clustered; and then the model of the type of the sample is applied to performing prediction so as to obtain the main indexes of the smoke. The method overcomes defects that the prior art needs to detect and operate a large number samples, achieves the prediction of the main indexes of the flue-cured tobacco smoke by using a small amount of the chemical composition samples, lightens the workload caused by smoke panel tests for formulating personnel, and can be widely used in the tobacco industry.

Description

technical field [0001] The invention relates to an evaluation method of flue-cured tobacco, in particular to a method for predicting main indexes of flue gas of flue-cured tobacco by using chemical components of tobacco. Background technique [0002] The traditional evaluation method of tobacco is mainly based on human sensory evaluation, but sensory quality evaluation is easily affected by people's physical condition and tobacco smoking amount, resulting in fluctuations in sensory quality. Simultaneously, testers need a large amount of evaluation tobacco, the workload is heavy, and it is not conducive to health. [0003] The gray forecasting model is a nonlinear extrapolation forecasting method developed in the 1980s. Because of its strong practicability, less data required, flexible and convenient modeling, and high forecasting accuracy, it is widely used in social sciences and natural sciences. fields are widely used. The neural network has the advantages of parallel co...

AI Technical Summary

Problems solved by technology

The gray forecasting model is mainly used for forecasting problems with short time, less data, and little fluctuation. In the case of less data, more accurate forecasting results can be obtained; the artificial neural network model (ANN) can be used in the case of more training samples A better prediction effect can be obtained, but the prediction error is larger when the sample size is insufficient.

Images