| Citrus is one of the main fruits in our country,which plays a very important role in promoting the economic development of our country.However,the economic losses of citrus caused by pests and diseases can not be neglected every year,especially the damage caused by bactrocera dorsalis to citrus economy is very serious.It is difficult to detect early infection of citrus by the change of citrus epidermis during the invasion and incubation period of the bactrocera dorsalis,but the early residual infection of citrus products can easily infect the adjacent healthy citrus,reduce the quality of citrus and further increase economic losses.Therefore,early detection during the invasion and incubation period of bactrocera dorsalis is of great significance to strengthen the quality control of citrus.In this paper,a bionic olfaction detection system was established,and the sensing properties of three quartz crystal microbalance(QCM)sensors made of ethyl cellulose,methyl-?-cyclodextrin and polyethylene glycol-2000 for d-limonene,linalool and decanal were studied.The bionic olfaction detection system was used to detect citrus with different storage time and infection time of bactrocera dorsalis,and the identification and discrimination were carried out with pattern recognition algorithm to verify the feasibility and reliability of the bionic olfaction detection system.The main conclusions of this study are as follows:1)Establishment of bionic olfactory detection system.Three kinds of QCM sensors were prepared by using ethyl cellulose,methyl-?-cyclodextrin and polyethylene glycol 2000 as materials,and the quartz oscillator was modified by dropping coating method.The QCM sensor array was constructed as the core component of the bionic olfactory detection system.The hardware part of the bionic olfaction detection system designs,the gas transmission pipeline and the detection room.In the software part,LabVIEW software is used as the development platform,and the man-machine interactive control interface of the detection system is compiled.The main functions of the design include system parameter setting,key control,pipeline environment status display and gas transmission pipeline control.2)The sensing properties of ethyl cellulose sensor,methyl-?-cyclodextrin sensor and polyethylene glycol-2000 sensor to d-limonene,linalool and decanal were detected by bionic olfaction detection system.The results show that the sensors with a thickness of 8000 Hz(frequency deviation before and after coating)have good sensitivity,repeatability,selectivity and reusability under 1%relative humidity.3)The bionic olfactory detection system detects citrus with different storage time and infection time of bactrocera dorsalis,and distinguishes citrus with pattern recognition algorithm.The steady state values of three sensors are used as characteristic data to form "characteristic fingerprints",and pattern recognition is carried out for different storage time,different infection time of bactrocera dorsalis,different time(including storage time and infection time).The correct recognition rates of principal component analysis can reach 100%,100%,90.62%,and the correct recognition rates of linear discriminant analysis can reach 100%,100%,94.79%.According to the volatile components of citrus determined by GC-MS,the relative contents of 3-oriental,a-pinene,d-limonene,terpene olefin,linalool and decyl aldehyde changed nonmonotonically with the increase of storage time and pest infection time,which further verified the feasibility of biomimetic olfaction detection system for detection of bactrocera dorsalis infection in citrus.In conclusion,the developed bionic olfactory detection system has certain feasibility in early detection of citrus infected by bactrocera dorsalis,which can provide reference and technical support for the follow-up non-destructive detection of citrus infected by bactrocera dorsalis. |
PDF Full Text Request