| Dealing with several flaws associated with traditional evaluation method of tobacco, such as tobacco inventory of the cigarette industrial enterprise is too large, accurate determination of the quality of tobacco is difficult, and the relationship between the quality of tobacco and tobacco planting ecological environment is unknown et al., as well as the traditional evaluation method of tobacco quality is over dependent on subjective weights of experts, ignoring the data characteristics themselves, this study used the 1429 flue-cured tobacco samples (including B2F, C3F, and X2F) from 27 counties and districts in Sichuan province of the year 2009-2012 as materials. And the current situation of Sichuan tobacco quality were analyzed. The individual quality evaluation index of the single sample tobacco was calculated, based on fuzzy theory, gray theory, RSR method, TOPSIS method, entropy weight theory and PCA method, and combined with the weight of subjective, objective, and combination, and the efficiency and difference of different calculation models were compared. Then chose the best model to calculate the comprehensive quality evaluation index of the single sample tobacco and the different regions tobacco. The comprehensive quality evaluation results of in different growing areas and different parts of tobacco were compared. At last, the relationship between the soil, weather and terrain factors and tobacco quality based on the tobacco growing areas of Liangshanzhou and Panzhihua were researched. The main results were as follows:1. The analysis results about the current situation of Sichuan tobacco quality was that:(1) Overall performance of appearance quality of Sichuan tobacco were as following:The tobacco’s color was ranging at the "orange" and "lemon yellow" level, about 85.5%; maturity was ranging at the "mature" and "immature" level, about 79.4%; leaf structure was ranging at "loose" level, about 86.3%; identity was ranging at the "medium" and "thick", "thin" level, about 49.8%,47.8%; oil was ranging at the "have" level, about 87.3%; chroma was ranging at "strong" level, about 71.7%. The tobacco appearance quality in Liangshan, Panzhihua, and Yibin, were better than that in other areas, and Liangshanzhou> Panzhihua> Yibin.(2) Overall performance of chemical quality of Sichuan tobacco were as following:The total sugar, reducing sugar content were high; the total nitrogen content was appropriate, but slightly low; the Cl content was particularly low, and not appropriate; the K content was appropriate, but slightly low; the sugar-nicotine was appropriate, but slightly higher; the nicotine content, starch content, nitrogen-nicotine ratio and K-Cl ratio were appropriate. The total sugar, reducing sugar content was extremely high in Liangshan and Panzhihua, and the sugar-nicotine ratio was extremely high in Panzhihua. The nicotine content, nitrogen-nicotine ratio was slightly higher in Yibin and Luzhou than that in other areas. The Cl content was suitable in Luzhou.(3) Overall performance of sensory quality of Sichuan tobacco were as following:The tobacco had good aroma quality, sufficient aroma quantity, moderate physiological strength, moderate and better concentration, slightly better cluster, less offensive odor, slightly comfortable and clean aftertaste, less irritancy, better tenderness and softness, strong sweetness, and slightly dryness. In addition to that tobacco’ physiological strength and concentration were worse, the others indicators in Liangshanzhou, Panzhihua were better than that in Yibin, Luzhou and Guanyuan,. Liangshanzhou were the best.2. Based on fuzzy theory, gray theory, RSR method, TOPSIS method, entropy weight theory and PC A method, and combination with the weight of subjective, objective, and combination, calculated the appearance, chemical, and sensory quality evaluation index of the single sample tobacco. The results show that:(1) In the quality evaluation of single sample tobacco, the fuzzy theory, gray theory, RSR method and Topsis method which were based on the combination weigh (obtained through the multiplication synthetic with objective weights of the entropy method and subjective weights of the PCA method, then the results were normalized) could reduce interference from subjective factors of experts’evaluation, thus these methods could focus on the data characteristics themselves, and improve the credibility of evaluation results. The grey theory model was better than others. Compared with traditional evaluation method of tobacco quality, The trend degree (R2) of computing results of appearance, chemistry and sensory score obtained by the gray theory model were the highest, which were 0.9758,0.8630,0.9827, respectively; the average absolute error (MAE), root mean square error (RMSE) and mean relative error (MRPE) were the lowest, which were 0.0192-0.0727,0.0242-0.0863, and 0.0066-0.0217, respectively.(2) The entropy model and PCA method could make full use of the data characteristics themselves, exclude the interference from subjective factors of experts’evaluation, therefore could efficiently and objectively reflect the evaluation results. Compared with traditional evaluation method of tobacco quality, The trend degree (R2) of computing results of appearance, chemistry and sensory score obtained by the entropy method were the highest, which were 0.9306,0.8216,0.9314, respectively; the MAE, RMSE and MRPE were the lowest, which were 0.0314-0.0382,0.0436-0.0971, and 0.0074-0.0493, respectively.3. The tobacco appearance, chemical, sensory and comprehensive quality evaluation results in different regions and different parts of Sichuan showed that:(1) For the different leaves parts of tobacco, the middle leaves (C3F) scored highest, the upper leaves (B2F) scored secondly and the lower leaves scored lowest (X2F) for the score of appearance, sensory and comprehensive quality index. The score of chemical quality index was highest for the upper leaves, scored secondly for the middle leaves and lowest for the lower leaves.(2) For the different regions of tobacco, the tobacco appearance quality index was that: Liangshanzhou> Panzhihua, Yibin> Guangyuan> Luzhou. The tobacco chemical quality index was that: Luzhou> Yibin> Guangyuan> Liangshanzhou> Panzhihua. The tobacco sensory quality index was that: Liangshanzhou> Panzhihua> Luzhou> Guangyuan> Yibin.The tobacco comprehensive quality index was that:Liangshanzhou> Panzhihua> Luzhou> Yibin> Guangyuan.4. The analysis results of relationship between the soil factors, weather factors, terrain factors and tobacco quality was that:(1) Tobacco appearance, chemical quality that planted on the paddy field were better than that on the dry land. The quality of tobacco planted on upland soil types was better, and for the sensory and comprehensive quality, the results showed that:the purple soil> red soil or yellow-red soil> paddy soil, and the difference was significant (p< 0.05). When soil pH value was at 5.5-7.5, the tobacco quality was the best; when the soil pH value was higher than 7.5, the tobacco quality was significantly worse (p< 0.05). The tobacco quality was the best when soil Org.matt content was at 15-30 g/kg, Ava.N content was at 45-135 mg/kg, Ava.P content was at 10-30 mg/kg. The score of appearance, sensory and comprehensive quality index showed that with the increment of the soil Org.matt, Ava.N, Ava.P content, the variation tendency was a parabola with downward opening, and the variation tendency was increase along with the increment of the soil Ava.K content.(2) Results confirmed a positive correlation between temperatures in different tobacco growth period and tobacco chemical quality index, and the temperatures of root elongation and vigorous growth stages were more significant (p<0.01).The precipitation of root-extended period and the tobacco appearance, sensory, comprehensive quality index was significantly positively correlative (p<0.01). It was significantly positively correlative between the precipitation of fast growing period and the tobacco appearance (p<0.01), comprehensive quality index, and had significant positive correlation (p<0.05). The sunshine duration of the tobacco field period (from April to September) had a positive correlation with each tobacco quality index, and the tobacco appearance, sensory quality were more obvious (p<0.01). The sunshine percentage of tobacco field period had a negative correlation with each tobacco quality index, and was negatively correlative with tobacco appearance, sensory quality (p<0.01), as well as tobacco sensory quality index (p<0.05).(3) Tobacco quality was better when planted on the gentle slope zone (slope in 5-15°), followed by that planted on flat area (slope in 0-5°). The tobacco appearance, sensory and comprehensive quality was better when planted on the sunny slope (south), semi-sunny slope (southwest, southeast) than that planted on the shady slope (north), semi-shady slope (northwest, northeast). The tobacco appearance, sensory and comprehensive quality was relatively poor in the low altitude zone (<1500 m) and high altitude zone (>2500 m), and the former was worse. |
PDF Full Text Request