The detection of plant diseases is an important part of commercial greenhouse crop production and can enable continued disease and pest control which will ultimately lead to the economical benefit as well as the significant reduction in use of chemical and biochemical treatments. A plant subject to infection typically releases exclusive volatile organic compounds (VOCs) which may be detected by appropriate sensors. A set of experiments were designed, constructed and conducted at University of Warwick in which the state-of-the-art gas sensors namely Electronic Nose (EN) and Field Asymmetric Ion Mobility Spectrometry (FAIMS) were employed to sample the VOC profiles in order to detect powdery mildew-infected and spider mite-infested tomato plants in a non-destructive manner. The data acquired from the EN and FAIMS devices was analysed using Principal Component Analysis, Linear Discriminant Analysis, Support Vector Machines and Artificial Neural Networks. Both EN and FAIMS proved to be able to distinguish between healthy and infected tomato plants with desirable accuracy when coupled with an appropriate data analysis technique. A review of the literature on plant diseases, destructive and non-destructive plant's disease detection tools as well as VOC sampling procedures and instrumentation will be given throughout this thesis. Moreover, the latter part of this thesis presents the master-slave synchronisation of identical chaotic dynamical systems using the open-plus close-loop (OPCL) control method. The study is mainly concerned with the behaviour of the synchronisation of chaotic dynamical systems in respect to an added bias and in the case of mismatch of parameters of master and slave systems. The link between the external bias and the synchronisation error generated as well as between the value of parameters mismatch and the synchronisation error is examined and discussed. The usability of the newly proposed approaches is assessed by the aid of two applications. The first application demonstrates that a weak bias acting on Nano-mechanical resonator shows the linear correlation with the synchronisation error and, consequently, the bias can be estimated via this error. The second application is related to the synchronisation of the cantilevers commonly found in ENs and Atomic Force Microscopy (AFM). It is suggested to use a novel scheme of coupled master{slave cantilevers and, estimate the difference in cantilever-surface interactions in master oscillator and in slave oscillator via measuring the synchronisation error. The scheme is particularly useful for using the master cantilever as a control and the slave cantilever as a unit device which measures surface properties. The study shows that by calculating the error of synchronisation, a precise measurements can be conducted when two cantilevers leave the synchronous region, that is when they de-synchronise. Thus, this thesis also contributes to the understanding of de-synchronisation of nano-scale chaotic systems (Nano electromechanical Systems) in respect to the addition of an external bias and/or parameters mismatch by outlining the possible applications.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:606164 |
| Date | January 2013 |
| Creators | Ghaffari, Reza |
| Publisher | University of Warwick |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://wrap.warwick.ac.uk/61777/ |
Page generated in 0.0021 seconds