Development of solar air heaters & thermal energy storage system for drying applications in food processing industries

In the present work, the author has designed and developed all types of solar air heaters called porous and nonporous collectors. The developed solar air heaters were subjected to different air mass flow rates in order to standardize the flow per unit area of the collector. Much attention was given to investigate the performance of the solar air heaters fitted with baffles. The output obtained from the experiments on pilot models, helped the installation of solar air heating system for industrial drying applications also. Apart from these, various types of solar dryers, for small and medium scale drying applications, were also built up. The feasibility of ‘latent heat thermal energy storage system’ based on Phase Change Material was also undertaken. The application of solar greenhouse for drying industrial effluent was analyzed in the present study and a solar greenhouse was developed. The effectiveness of Computational Fluid Dynamics (CFD) in the field of solar air heaters was also analyzed. The thesis is divided into eight chapters.

Studies on pulp propogation in single mode optical fibres

Studies on pulse propagation in single mode optical fibers have attracted interest from a wide area of science and technology as they have laid down the foundation for an in-depth understanding of the underlying physical principles, especially in the field of optical telecommunications. The foremost among them is discovery of the optical soliton which is considered to be one of the most significant events of the twentieth century owing to its fantastic ability to propagate undistorted over long distances and to remain unaflected after collision with each other. To exploit the important propertia of optical solitons, innovative mathematical models which take into account proper physical properties of the single mode optical fibers demand special attention. This thesis contains a theoretical analysis of the studies on soliton pulse propagation in single mode optical fibers.

Speaker identification using models for phonemes

Motivation for Speaker recognition work is presented in the first part of the thesis. An exhaustive survey of past work in this field is also presented. A low cost system not including complex computation has been chosen for implementation. Towards achieving this a PC based system is designed and developed. A front end analog to digital convertor (12 bit) is built and interfaced to a PC. Software to control the ADC and to perform various analytical functions including feature vector evaluation is developed. It is shown that a fixed set of phrases incorporating evenly balanced phonemes is aptly suited for the speaker recognition work at hand. A set of phrases are chosen for recognition. Two new methods are adopted for the feature evaluation. Some new measurements involving a symmetry check method for pitch period detection and ACE‘ are used as featured. Arguments are provided to show the need for a new model for speech production. Starting from heuristic, a knowledge based (KB) speech production model is presented. In this model, a KB provides impulses to a voice producing mechanism and constant correction is applied via a feedback path. It is this correction that differs from speaker to speaker. Methods of defining measurable parameters for use as features are described. Algorithms for speaker recognition are developed and implemented. Two methods are presented. The first is based on the model postulated. Here the entropy on the utterance of a phoneme is evaluated. The transitions of voiced regions are used as speaker dependent features. The second method presented uses features found in other works, but evaluated differently. A knock—out scheme is used to provide the weightage values for the selection of features. Results of implementation are presented which show on an average of 80% recognition. It is also shown that if there are long gaps between sessions, the performance deteriorates and is speaker dependent. Cross recognition percentages are also presented and this in the worst case rises to 30% while the best case is 0%. Suggestions for further work are given in the concluding chapter.

Effects of new nonlinear couplings in relativistic effective field theory

We extend the relativistic mean field theory model of Sugahara and Toki by adding new couplings suggested by modern effective field theories. An improved set of parameters is developed with the goal to test the ability of the models based on effective field theory to describe the properties of finite nuclei and, at the same time, to be consistent with the trends of Dirac-Brueckner-Hartree-Fock calculations at densities away from the saturation region. We compare our calculations with other relativistic nuclear force parameters for various nuclear phenomena.

A framework for translating English text into Malayalam using statistical models

A methodology for translating text from English into the Dravidian language, Malayalam using statistical models is discussed in this paper. The translator utilizes a monolingual Malayalam corpus and a bilingual English/Malayalam corpus in the training phase and generates automatically the Malayalam translation of an unseen English sentence. Various techniques to improve the alignment model by incorporating the morphological inputs into the bilingual corpus are discussed. Removing the insignificant alignments from the sentence pairs by this approach has ensured better training results. Pre-processing techniques like suffix separation from the Malayalam corpus and stop word elimination from the bilingual corpus also proved to be effective in producing better alignments. Difficulties in translation process that arise due to the structural difference between the English Malayalam pair is resolved in the decoding phase by applying the order conversion rules. The handcrafted rules designed for the suffix separation process which can be used as a guideline in implementing suffix separation in Malayalam language are also presented in this paper. Experiments conducted on a sample corpus have generated reasonably good Malayalam translations and the results are verified with F measure, BLEU and WER evaluation metrics

Studies on the interaction of two level atoms with quantized electromagnetic field

The present study described about the interaction of a two level atom and squeezed field with time varying frequency. By applying a sinusoidal variation in the frequency of the field, the randomness in population inversion is reduced and the collapses and periodic revivals are regained. Quantum optics is an emerging field in physics which mainly deals with the interaction of atoms with quantised electromagnetic fields. Jaynes-Cummings Model (JCM) is a key model among them, which describes the interaction between a two level atom and a single mode radiation field. Here the study begins with a brief history of light, atom and their interactions. Also discussed the interaction between atoms and electromagnetic fields. The study suggest a method to manipulate the population inversion due to interaction and control the randomness in it, by applying a time dependence on the frequency of the interacting squeezed field.The change in behaviour of the population inversion due to the presence of a phase factor in the applied frequency variation is explained here.This study also describes the interaction between two level atom and electromagnetic field in nonlinear Kerr medium. It deals with atomic and field state evolution in a coupled cavity system. Our results suggest a new method to control and manipulate the population of states in two level atom radiation interaction,which is very essential for quantum information processing.We have also studied the variation of atomic population inversion with time, when a two level atom interacts with light field, where the light field has a sinusoidal frequency variation with a constant phase. In both coherent field and squeezed field cases, the population inversion variation is completely different from the phase zero frequency modulation case. It is observed that in the presence of a non zero phase φ, the population inversion oscillates sinusoidally.Also the collapses and revivals gradually disappears when φ increases from 0 to π/2. When φ = π/2 the evolution of population inversion is identical to the case when a two level atom interacts with a Fock state. Thus, by applying a phase shifted frequency modulation one can induce sinusoidal oscillations of atomic inversion in linear medium, those normally observed in Kerr medium. We noticed that the entanglement between the atom and field can be controlled by varying the period of the field frequency fluctuations. The system has been solved numerically and the behaviour of it for different initial conditions and different susceptibility values are analysed. It is observed that, for weak cavity coupling the effect of susceptibility is minimal. In cases of strong cavity coupling, susceptibility factor modifies the nature in which the probability oscillates with time. Effect of susceptibility on probability of states is closely related to the initial state of the system.

Studies on scattering and thermodynamics of black holes in f(R) theory and Einstein's theory

One of the interesting consequences of Einstein's General Theory of Relativity is the black hole solutions. Until the observation made by Hawking in 1970s, it was believed that black holes are perfectly black. The General Theory of Relativity says that black holes are objects which absorb both matter and radiation crossing the event horizon. The event horizon is a surface through which even light is not able to escape. It acts as a one sided membrane that allows the passage of particles only in one direction i.e. towards the center of black holes. All the particles that are absorbed by black hole increases the mass of the black hole and thus the size of event horizon also increases. Hawking showed in 1970s that when applying quantum mechanical laws to black holes they are not perfectly black but they can emit radiation. Thus the black hole can have temperature known as Hawking temperature. In the thesis we have studied some aspects of black holes in f(R) theory of gravity and Einstein's General Theory of Relativity. The scattering of scalar field in this background space time studied in the first chapter shows that the extended black hole will scatter scalar waves and have a scattering cross section and applying tunneling mechanism we have obtained the Hawking temperature of this black hole. In the following chapter we have investigated the quasinormal properties of the extended black hole. We have studied the electromagnetic and scalar perturbations in this space-time and find that the black hole frequencies are complex and show exponential damping indicating the black hole is stable against the perturbations. In the present study we show that not only the black holes exist in modified gravities but also they have similar properties of black hole space times in General Theory of Relativity. 2 + 1 black holes or three dimensional black holes are simplified examples of more complicated four dimensional black holes. Thus these models of black holes are known as toy models of black holes in four dimensional black holes in General theory of Relativity. We have studied some properties of these types of black holes in Einstein model (General Theory of Relativity). A three dimensional black hole known as MSW is taken for our study. The thermodynamics and spectroscopy of MSW black hole are studied and obtained the area spectrum which is equispaced and different thermo dynamical properties are studied. The Dirac perturbation of this three dimensional black hole is studied and the resulting quasinormal spectrum of this three dimensional black hole is obtained. The different quasinormal frequencies are tabulated in tables and these values show an exponential damping of oscillations indicating the black hole is stable against the mass less Dirac perturbation. In General Theory of Relativity almost all solutions contain singularities. The cosmological solution and different black hole solutions of Einstein's field equation contain singularities. The regular black hole solutions are those which are solutions of Einstein's equation and have no singularity at the origin. These solutions possess event horizon but have no central singularity. Such a solution was first put forward by Bardeen. Hayward proposed a similar regular black hole solution. We have studied the thermodynamics and spectroscopy of Hay-ward regular black holes. We have also obtained the different thermodynamic properties and the area spectrum. The area spectrum is a function of the horizon radius. The entropy-heat capacity curve has a discontinuity at some value of entropy showing a phase transition.

Organic field-effect transistors and phototransistors based on amorphous materials

A comparison between the charge transport properties in low molecular amorphous thin films of spiro-linked compound and their corresponding parent compound has been demonstrated. The field-effect transistor method is used for extracting physical parameters such as field-effect mobility of charge carriers, ON/OFF ratios, and stability. In addition, phototransistors have been fabricated and demonstrated for the first time by using organic materials. In this case, asymmetrically spiro-linked compounds are used as active materials. The active materials used in this study can be divided into three classes, namely Spiro-linked compounds (symmetrically spiro-linked compounds), the corresponding parent-compounds, and photosensitive spiro-linked compounds (asymmetrically spiro-linked com-pounds). Some of symmetrically spiro-linked compounds used in this study were 2,2',7,7'-Tetrakis-(di-phenylamino)-9,9'-spirobifluorene (Spiro-TAD),2,2',7,7'-Tetrakis-(N,N'-di-p-methylphenylamino)-9,9'-spirobifluorene (Spiro-TTB), 2,2',7,7'-Tetra-(m-tolyl-phenylamino)-9,9'-spirobifluorene (Spiro-TPD), and 2,2Ž,7,7Ž-Tetra-(N-phenyl-1-naphtylamine)-9,9Ž-spirobifluorene (Spiro alpha-NPB). Related parent compounds of the symmetrically spiro-linked compound used in this study were N,N,N',N'-Tetraphenylbenzidine (TAD), N,N,N',N'-Tetrakis(4-methylphenyl)benzidine (TTB), N,N'-Bis(3-methylphenyl)-(1,1'-biphenyl)-4,4'-diamine (TPD), and N,N'-Diphenyl-N,N'-bis(1-naphthyl)-1,1'-biphenyl-4,4'-diamine (alpha-NPB). The photosensitive asymmetrically spiro-linked compounds used in this study were 2,7-bis-(N,N'-diphenylamino)-2',7'-bis(biphenyl-4-yl)-9,9'-spirobifluorene (Spiro-DPSP), and 2,7-bis-(N,N'-diphenylamino)-2',7'-bis(spirobifluorene-2-yl)-9,9'-spirobifluorene (Spiro-DPSP^2). It was found that the field-effect mobilities of charge carriers in thin films of symmetrically spiro-linked compounds and their corresponding parent compounds are in the same order of magnitude (~10^-5 cm^2/Vs). However, the thin films of the parent compounds were easily crystallized after the samples have been exposed in ambient atmosphere and at room temperature for three days. In contrast, the thin films and the transistor characteristics of symmetrically spiro-linked compound did not change significantly after the samples have been stored in ambient atmosphere and at room temperature for several months. Furthermore, temperature dependence of the mobility was analyzed in two models, namely the Arrhenius model and the Gaussian Disorder model. The Arrhenius model tends to give a high value of the prefactor mobility. However, it is difficult to distinguish whether the temperature behaviors of the material under consideration follows the Arrhenius model or the Gaussian Disorder model due to the narrow accessible range of the temperatures. For the first time, phototransistors have been fabricated and demonstrated by using organic materials. In this case, asymmetrically spiro-linked compounds are used as active materials. Intramolecular charge transfer between a bis(diphenylamino)biphenyl unit and a sexiphenyl unit leads to an increase in charge carrier density, providing the amplification effect. The operational responsivity of better than 1 A/W can be obtained for ultraviolet light at 370 nm, making the device interesting for sensor applications. This result offers a new potential application of organic thin film phototransistors as low-light level and low-cost visible blind ultraviolet photodetectors.

Hyperspectral field measurements to determine dry matter yield and nutritive value of legume-grass swards

Summary: Productivity, botanical composition and forage quality of legume-grass swards are important factors for successful arable farming in both organic and conventional farming systems. As these attributes can vary considerably within a field, a non-destructive method of detection while doing other tasks would facilitate a more targeted management of crops, forage and nutrients in the soil-plant-animal system. This study was undertaken to explore the potential of field spectral measurements for a non destructive prediction of dry matter (DM) yield, legume proportion in the sward, metabolizable energy (ME), ash content, crude protein (CP) and acid detergent fiber (ADF) of legume-grass mixtures. Two experiments were conducted in a greenhouse under controlled conditions which allowed collecting spectral measurements which were free from interferences such as wind, passing clouds and changing angles of solar irradiation. In a second step this initial investigation was evaluated in the field by a two year experiment with the same legume-grass swards. Several techniques for analysis of the hyperspectral data set were examined in this study: four vegetation indices (VIs): simple ratio (SR), normalized difference vegetation index (NDVI), enhanced vegetation index (EVI) and red edge position (REP), two-waveband reflectance ratios, modified partial least squares (MPLS) regression and stepwise multiple linear regression (SMLR). The results showed the potential of field spectroscopy and proved its usefulness for the prediction of DM yield, ash content and CP across a wide range of legume proportion and growth stage. In all investigations prediction accuracy of DM yield, ash content and CP could be improved by legume-specific calibrations which included mixtures and pure swards of perennial ryegrass and of the respective legume species. The comparison between the greenhouse and the field experiments showed that the interaction between spectral reflectance and weather conditions as well as incidence angle of light interfered with an accurate determination of DM yield. Further research is hence needed to improve the validity of spectral measurements in the field. Furthermore, the developed models should be tested on varying sites and vegetation periods to enhance the robustness and portability of the models to other environmental conditions.

Die Natur als Erholungs(t)raum

Zur Erholung in die Natur gehen oder doch lieber zur Natursimulation greifen? Intuitiv würden die meisten Menschen der Natur einen größeren Erholungswert zusprechen als einer Natursimulation. Aber ist die Natur tatsächlich erholsamer? In der Naturerholungsforschung (Restorative Environment Research) kommen häufig Natursimulationen zum Einsatz, um die erholsame Wirkung von Natur zu ermitteln. Problematisch ist dabei, dass deren ökologische Validität und Vergleichbarkeit noch nicht empirisch abgesichert ist. Vorliegende Arbeit setzt an dieser methodischen und empirischen Lücke an. Sie überprüft sowohl die ökologische Validität als auch die Vergleichbarkeit von Natursimulationen. Dazu wird die erholsame Wirkung von zwei Natursimulationen im Vergleich zu der physisch-materiellen Natur empirisch untersucht und verglichen. Darüber hinaus werden Aspekte des subjektiven Erlebens und der Bewertung im Naturerholungskontext exploriert. Als bedeutsamer Wirkmechanismus wird die erlebnisbezogene Künstlichkeit/Natürlichkeit angesehen, die sich auf die Erlebnisqualität von Natursimulationen und der physisch-materiellen Natur bezieht: Natursimulationen weisen im Vergleich zur physisch-materiellen Natur eine reduzierte Erlebnisqualität auf (erlebnisbezogene Künstlichkeit), z.B. eine reduzierte Qualität und Quantität der Sinnesansprache. Stellt man einen derartigen Vergleich nicht nur mit der physisch-materiellen Natur, sondern mit unterschiedlichen Natursimulationstypen an, dann zeigen sich auch hier Unterschiede in der erlebnisbezogenen Künstlichkeit. Beispielsweise unterscheidet sich ein Naturfoto von einem Naturfilm durch das Fehlen von auditiven und bewegten Stimuli. Diese erlebnisbezogene Künstlichkeit kann die erholsame Wirkung von Natur - direkt oder indirekt über Bewertungen - hemmen. Als Haupthypothese wird angenommen, dass mit zunehmendem Ausmaß an erlebnisbezogener Künstlichkeit die erholsame Wirkung der Natur abnimmt. Dem kombinierten Feld- und Laborexperiment liegt ein einfaktorielles Vorher-Nachher-Design zugrunde. Den 117 Probanden wurde zunächst eine kognitiv und affektiv belastende Aufgabe vorgelegt, danach folgte die Erholungsphase. Diese bestand aus einem Spaziergang, der entweder in der physisch-materiellen Natur (urbaner Park) oder in einer der beiden audio-visuellen Natursimulationen (videogefilmter vs. computergenerierter Spaziergang durch selbigen urbanen Park) oder auf dem Laufband ohne audio-visuelle Darbietung stattfand. Die erlebnisbezogene Künstlichkeit/Natürlichkeit wurde also wie folgt operationlisiert: die physische Natur steht für die erlebnisbezogene Natürlichkeit. Die beiden Natursimulationen stehen für die erlebnisbezogene Künstlichkeit. Die computergenerierte Version ist im Vergleich zur Videoversion erlebnisbezogen künstlicher, da sie weniger fotorealistisch ist. Die Zuordnung zu einer der vier experimentellen Erholungssettings erfolgte nach dem Zufallsprinzip. Die Effekte von moderater Bewegung wurden in den Natursimulationen durch das Laufen auf dem Laufband kontrolliert. Die Beanspruchungs- bzw. Erholungsreaktionen wurden auf kognitiver (Konzentriertheit, Aufmerksamkeitsleistung) affektiver (3 Befindlichkeitsskalen: Wachheit, Ruhe, gute Stimmung) und physiologischer (Alpha-Amylase) Ebene gemessen, um ein umfassendes Bild der Reaktionen zu erhalten. Insgesamt zeigen die Ergebnisse, dass die beiden Natursimulationen trotz Unterschiede in der erlebnisbezogenen Künstlichkeit/Natürlichkeit zu relativ ähnlichen Erholungsreaktionen führen, wie die physisch-materielle Natur. Eine Ausnahme stellen eine der drei affektiven (Wachheit) und die physiologische Reaktion dar: Probanden der physisch-materiellen Naturbedingung geben an wacher zu sein und weisen - wider erwarten - eine höhere physiologische Erregung auf. Demnach ist die physisch-materielle Natur nicht grundsätzlich erholsamer als die Natursimulationen. Die Hypothese ließ sich somit nicht bestätigen. Vielmehr deuten sich komplexe Erholungsmuster und damit auch unterschiedliche Erholungsqualitäten der Settings an, die einer differenzierten Betrachtung bedürfen. Für die ökologische Validität von Natursimulationen gilt, dass diese nur mit Einschränkung als ökologisch valide bezeichnet werden können, d.h. nur für bestimmte, aber nicht für alle Erholungsreaktionen. Die beiden Natursimulationen führen ebenfalls trotz Unterschiede in der erlebnisbezogenen Künstlichkeit zu ähnlichen Erholungsreaktionen und können somit als gleichwertig behandelt werden. Erstaunlicherweise kommt es hier zu ähnlichen Erholungsreaktionen, obwohl die bestehenden Unterschiede von den Probanden wahrgenommen und die erlebnisbezogen künstlichere computergenerierte Version negativer bewertet wird. Aufgrund der nicht erwartungskonformen Ergebnisse muss das Erklärungskonzept der erlebnisbezogenen Künstlichkeit/Natürlichkeit infrage gestellt werden. Alternative Erklärungskonzepte für die Ergebnisse („Ungewissheit“, mentale räumliche Modelle), die sich andeutenden unterschiedlichen Erholungsqualitäten der Settings, methodische Einschränkungen sowie die praktische Bedeutung der Ergebnisse werden kritisch diskutiert.

Determination of yield and quality parameters of energy crops applying laboratory and field spectroscopy

A real-time analysis of renewable energy sources, such as arable crops, is of great importance with regard to an optimised process management, since aspects of ecology and biodiversity are considered in crop production in order to provide a sustainable energy supply by biomass. This study was undertaken to explore the potential of spectroscopic measurement procedures for the prediction of potassium (K), chloride (Cl), and phosphate (P), of dry matter (DM) yield, metabolisable energy (ME), ash and crude fibre contents (ash, CF), crude lipid (EE), nitrate free extracts (NfE) as well as of crude protein (CP) and nitrogen (N), respectively in pretreated samples and undisturbed crops. Three experiments were conducted, one in a laboratory using near infrared reflectance spectroscopy (NIRS) and two field spectroscopic experiments. Laboratory NIRS measurements were conducted to evaluate to what extent a prediction of quality parameters is possible examining press cakes characterised by a wide heterogeneity of their parent material. 210 samples were analysed subsequent to a mechanical dehydration using a screw press. Press cakes serve as solid fuel for thermal conversion. Field spectroscopic measurements were carried out with regard to further technical development using different field grown crops. A one year lasting experiment over a binary mixture of grass and red clover examined the impact of different degrees of sky cover on prediction accuracies of distinct plant parameters. Furthermore, an artificial light source was used in order to evaluate to what extent such a light source is able to minimise cloud effects on prediction accuracies. A three years lasting experiment with maize was conducted in order to evaluate the potential of off-nadir measurements inside a canopy to predict different quality parameters in total biomass and DM yield using one sensor for a potential on-the-go application. This approach implements a measurement of the plants in 50 cm segments, since a sensor adjusted sideways is not able to record the entire plant height. Calibration results obtained by nadir top-of-canopy reflectance measurements were compared to calibration results obtained by off-nadir measurements. Results of all experiments approve the applicability of spectroscopic measurements for the prediction of distinct biophysical and biochemical parameters in the laboratory and under field conditions, respectively. The estimation of parameters could be conducted to a great extent with high accuracy. An enhanced basis of calibration for the laboratory study and the first field experiment (grass/clover-mixture) yields in improved robustness of calibration models and allows for an extended application of spectroscopic measurement techniques, even under varying conditions. Furthermore, off-nadir measurements inside a canopy yield in higher prediction accuracies, particularly for crops characterised by distinct height increment as observed for maize.

Parameter identification of structural dynamic models by inverse statistical analysis

Auf dem Gebiet der Strukturdynamik sind computergestützte Modellvalidierungstechniken inzwischen weit verbreitet. Dabei werden experimentelle Modaldaten, um ein numerisches Modell für weitere Analysen zu korrigieren. Gleichwohl repräsentiert das validierte Modell nur das dynamische Verhalten der getesteten Struktur. In der Realität gibt es wiederum viele Faktoren, die zwangsläufig zu variierenden Ergebnissen von Modaltests führen werden: Sich verändernde Umgebungsbedingungen während eines Tests, leicht unterschiedliche Testaufbauten, ein Test an einer nominell gleichen aber anderen Struktur (z.B. aus der Serienfertigung), etc. Damit eine stochastische Simulation durchgeführt werden kann, muss eine Reihe von Annahmen für die verwendeten Zufallsvariablengetroffen werden. Folglich bedarf es einer inversen Methode, die es ermöglicht ein stochastisches Modell aus experimentellen Modaldaten zu identifizieren. Die Arbeit beschreibt die Entwicklung eines parameter-basierten Ansatzes, um stochastische Simulationsmodelle auf dem Gebiet der Strukturdynamik zu identifizieren. Die entwickelte Methode beruht auf Sensitivitäten erster Ordnung, mit denen Parametermittelwerte und Kovarianzen des numerischen Modells aus stochastischen experimentellen Modaldaten bestimmt werden können.

Development of thermo-analytical prediction and classification models for food in thermal devices using a multi sensor system and artificial neural networks

Die thermische Verarbeitung von Lebensmitteln beeinflusst deren Qualität und ernährungsphysiologischen Eigenschaften. Im Haushalt ist die Überwachung der Temperatur innerhalb des Lebensmittels sehr schwierig. Zudem ist das Wissen über optimale Temperatur- und Zeitparameter für die verschiedenen Speisen oft unzureichend. Die optimale Steuerung der thermischen Zubereitung ist maßgeblich abhängig von der Art des Lebensmittels und der äußeren und inneren Temperatureinwirkung während des Garvorgangs. Das Ziel der Arbeiten war die Entwicklung eines automatischen Backofens, der in der Lage ist, die Art des Lebensmittels zu erkennen und die Temperatur im Inneren des Lebensmittels während des Backens zu errechnen. Die für die Temperaturberechnung benötigten Daten wurden mit mehreren Sensoren erfasst. Hierzu kam ein Infrarotthermometer, ein Infrarotabstandssensor, eine Kamera, ein Temperatursensor und ein Lambdasonde innerhalb des Ofens zum Einsatz. Ferner wurden eine Wägezelle, ein Strom- sowie Spannungs-Sensor und ein Temperatursensor außerhalb des Ofens genutzt. Die während der Aufheizphase aufgenommen Datensätze ermöglichten das Training mehrerer künstlicher neuronaler Netze, die die verschiedenen Lebensmittel in die entsprechenden Kategorien einordnen konnten, um so das optimale Backprogram auszuwählen. Zur Abschätzung der thermische Diffusivität der Nahrung, die von der Zusammensetzung (Kohlenhydrate, Fett, Protein, Wasser) abhängt, wurden mehrere künstliche neuronale Netze trainiert. Mit Ausnahme des Fettanteils der Lebensmittel konnten alle Komponenten durch verschiedene KNNs mit einem Maximum von 8 versteckten Neuronen ausreichend genau abgeschätzt werden um auf deren Grundlage die Temperatur im inneren des Lebensmittels zu berechnen. Die durchgeführte Arbeit zeigt, dass mit Hilfe verschiedenster Sensoren zur direkten beziehungsweise indirekten Messung der äußeren Eigenschaften der Lebensmittel sowie KNNs für die Kategorisierung und Abschätzung der Lebensmittelzusammensetzung die automatische Erkennung und Berechnung der inneren Temperatur von verschiedensten Lebensmitteln möglich ist.

Factorial Hidden Markov Models

We present a framework for learning in hidden Markov models with distributed state representations. Within this framework, we derive a learning algorithm based on the Expectation--Maximization (EM) procedure for maximum likelihood estimation. Analogous to the standard Baum-Welch update rules, the M-step of our algorithm is exact and can be solved analytically. However, due to the combinatorial nature of the hidden state representation, the exact E-step is intractable. A simple and tractable mean field approximation is derived. Empirical results on a set of problems suggest that both the mean field approximation and Gibbs sampling are viable alternatives to the computationally expensive exact algorithm.

View-Based Strategies for 3D Object Recognition

A persistent issue of debate in the area of 3D object recognition concerns the nature of the experientially acquired object models in the primate visual system. One prominent proposal in this regard has expounded the use of object centered models, such as representations of the objects' 3D structures in a coordinate frame independent of the viewing parameters [Marr and Nishihara, 1978]. In contrast to this is another proposal which suggests that the viewing parameters encountered during the learning phase might be inextricably linked to subsequent performance on a recognition task [Tarr and Pinker, 1989; Poggio and Edelman, 1990]. The 'object model', according to this idea, is simply a collection of the sample views encountered during training. Given that object centered recognition strategies have the attractive feature of leading to viewpoint independence, they have garnered much of the research effort in the field of computational vision. Furthermore, since human recognition performance seems remarkably robust in the face of imaging variations [Ellis et al., 1989], it has often been implicitly assumed that the visual system employs an object centered strategy. In the present study we examine this assumption more closely. Our experimental results with a class of novel 3D structures strongly suggest the use of a view-based strategy by the human visual system even when it has the opportunity of constructing and using object-centered models. In fact, for our chosen class of objects, the results seem to support a stronger claim: 3D object recognition is 2D view-based.