Search

Tribolium castaneum (Herbst) has been one of the most destructive insects for cereals and products in the world. A rapid and accurate detection can be greatly contributed to the stored grain in recent years. In this study, an electronic nose detection device was developed to detect the pests in stored wheat grains using the feature optimization of the gas sensor array. The volatile odor information of Herbst in wheat was also collected with different population density gradients. Firstly, the response signal of the gas sensor array was analyzed and preprocessed after collection. Then, the feature extraction and selection were performed on the response curve of the sensor array, in order to obtain the feature information of the sample. A regression model was established to predict the population density of Herbst in wheat. The Relative Change value (RC), Relative Integral value (RI), and Mean Differential value (MD) of each sensor were extracted as the original feature matrix (10×3), according to the response curves collected by 10 gas sensors. A Genetic Algorithm (GA) was adopted as the feature selection to optimize the original multi-feature matrix, where the recognition accuracy was used as the evaluation index. Finally, Partial Least Squares Regression (PLSR), Principal Component Regression (PCR), and Support Vector Regression (SVR) were used for the regression prediction. The results show that the PLSR model achieved the best prediction. The Principal Component Analysis (PCA) and Cluster Analysis (CA) were used to analyze the discrimination and aggregation between different types of samples. It was found that the electronic nose device effectively distinguished the wheat samples from those infected by pests. The redundant information was removed from the optimized feature matrix, according to the recognition accuracy and PCA data. The recognition accuracy of the feature matrix increased from 82.85% to 97.14% after GA optimization. The number of feature variables was reduced by 60% from 30 to 12, and the number of sensors was reduced to 8. The PLSR, PCR, and SVR were used to establish the prediction model before and after the optimization of the matrix. The prediction set correlation coefficient (r) and root mean square error (RMSE) of the PLSR model were 0.828 and 11.293, respectively. By contrast, the prediction set r and RMSE of the PCR model were 0.764 and 13.859, respectively. The prediction set r and RMSE of the SVR model were 0.804 and 12.976, respectively. The results showed that the PLSR was the best model for the prediction of the population density of Herbst. Consequently, the feasibility and effectiveness of the sensor array multi-feature optimization can be verified for the electronic nose detection of stored grain pests. The finding can also provide a strong reference for the rapid detection of grain pests. [ABSTRACT FROM AUTHOR]

Soil nitrogen as an essential nutrient element is one of the most important indexes to measure soil fertility for crop growth and development. In this research, a new detection was proposed to quickly accurately determine the soil total nitrogen (STN) content using pyrolysis and electronic nose. Ten types of gas sensors were used to construct the sensor arrays. A response test was carried out under the different concentrations of methane, vinyl chloride, and ammonia standard gas. The test results showed that there were significant differences in responses of the sensor array to the types and the concentration, where the response intensity increased with the increase of the standard gas concentration. The sensor array also presented a high specificity and cross-sensitivity during data detection. Furthermore, the pyrolysis gas was obtained from the soil samples using the muffle furnace, further to detect the response curve using the gas sensor array. After that, a 121×10×7 feature space (121 soil samples, 10 number of sensors, and 7 eigenvalues) was constructed to extract the mean (Vmean), variance (Vvav), the maximum gradient (Vmgv), the maximum (Vmax), response area (Vrav), the eighth of the second transient (V8), and mean differential coefficient (Vmdc) of the response curve. A genetic algorithm and neural network model (GA-BP) feature optimization was used to reduce the eigenvalue to 33 dimensions, forming a new feature space of 121×33. More importantly, there was no redundant effect of the constructed sensor array on the new detection. Specifically, the sensors of TGS826, TGS2603, TGS2611, and TGS2600 contributed the most to the construction of the new feature space. The Vmean, Vmgv, Vrav, V8 and Vmdc were the important features to represent the internal relationship between the detection and STN content. The prediction model of feature space and STN content was then established using a back propagation neural network (BPNN), partial least squares regression (PLSR), and a combination of a back propagation neural network and partial least squares regression (PLSR-BPNN). The coefficient of determination (R²), root mean square error (RMSE), and residual prediction deviation (RPD) were used as the indicators of the model. As such, the R² of PLSR, BPNN and PLSR-BPNN models were 0.91, 0.81, and 0.93, respectively, where the RMSE were 0.25, 0.37, and 0.22, while the RPD were 3.24, 2.19, and 3.79, respectively. The predicted performance of the test sets demonstrated that the R² values of the three models were all greater than 0.81, and the RMSE<0.37, indicating that all the models presented the better prediction ability of STN content. However, both PLSR and PLSR-BPNN models presented a much better ability of quantitative prediction than that of the BPNN, from the perspective of RPD indicators. The R² of the PLSR-BPNN model increased by 2.90%, the RPD increased by 16.94%, and the RMSE was reduced by 14.48%, compared with the PLSR model. Therefore, the PLSR-BPNN prediction model can be expected to effectively improve the prediction accuracy of the PLSR model for the better generalization ability of the BPNN model, indicating a reliable relationship model for the STN measurement. The reason was that there was a certain degree of linear and nonlinear correlation between the STN content and characteristic space of the electronic nose. The PLSR-BPNN model greatly contributed to the strength of the nonlinear and linear relationship between the PLSR and BPNN model. Consequently, a more accurate PLSR-BPNN model was established to accurately predict the STN content. There was also a high correlation between the soil pyrolysis gas and STN content. [ABSTRACT FROM AUTHOR]