Qualitative and quantitative pesticide residue analysis in Allium tuberosum using an electronic nose with multivariate analysis.

As the pesticide residue issue in Allium tuberosum becomes more severe, developing a rapid and robust detection method to assess its pesticide residue concentration is crucial. This study proposes a combined method based on the electronic nose (E-nose), Particle Swarm Optimization (PSA) algorithm, and Support Vector Machine (SVM) for both qualitative and quantitative detection of pesticide residues in Allium tuberosum. The results show that the SVM classifier effectively distinguishes between normal and contaminated samples, with a classification accuracy of 100 %. For pesticide type identification, the accuracy of SVM for Chlorpyrifos, Carbendazim, and Procymidone reaches 100 %. In qualitative analysis, the accuracy of SVM for Chlorpyrifos and Carbendazim is 100 %, and for Procymidone, it is 97.22 %. Quantitative results indicate that Support Vector Regression (SVR) performs excellently in predicting Procymidone (Rp = 0.9130) and Carbendazim (Rp = 0.9572). Therefore, the E-nose can serve as an effective tool for monitoring vegetable safety.