In this thesis, shaft transducerless vector control of Interior Permanent Magnet (IPM) motor with speed and position estimation using saliency based high frequency signal injection and fundamental model based flux observer methods will be investigated. The magnetic saliency characteristic of a 2.2-kW IPM motor will be experimentally extracted by means of high frequency signal injection. High frequency signal injection method will be used to estimate the speed and position at zero and low speed based on the magnetic saliency of the IPM motor. At high speed, fundamental model based flux observer method will be utilized for speed and position estimation. Seamless transition between the two estimation methods will be provided. Using the estimated speed and position information, the motor will be closed loop vector controlled and the drive motion performance over wide speed and load range will be investigated. The IPM motor drive and the estimation/control algorithms will be modeled and their performance will be demonstrated by detailed computer simulations. A three-phase voltage source inverter and a motor test bench will be built, and the estimation/control algorithms will be implemented on a DSP based motor control platform. The IPM motor drive system will be tested in the laboratory and the theory and simulation results will be verified by the experiments.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This research is inserted in the Biomathematics Research Group of the Programa de Pós-Graduação em Matemática of the Universidade Federal de Santa Maria-RS. It consists in the study of a discrete model for a prey-predator-insecticide system based on the Coupled Map Lattice as basic tool for its formulation. Due to the serious problems that crop pests represent to agriculture, we aproached the Integrated Pest Management (IPM), which is a pest control system that seeks to preserve and enhance the natural mortality factors of pests by integrated use of control methods selected based on technical, economic, ecological and sociological parameters. The objective of this work is to build a model as simple as possible to study the effects of different pests management strategies. These strategies are divided into different treatments which take into account the number of measurements of pest population, establishing control levels (CL) and different control measures (pesticide and biological control). For comparison and to point where the treatment is more efficient to compute the total density of the pest population without any control measure and when each treatment is applied over a certain period of time. The difference between these values is converted into a decreasing percentage of the population of pests. The same is done to obtain the percentage decrease in the number of treated sites. It was found that the sooner you apply the insecticide in the system, the more effective the treatment. Furthermore, it was observed that the presence of predators (biological control) decreases the amount of treated sites and, in some instances, is less effective in controlling pest. / Esta pesquisa está inserida na Linha de Pesquisa Biomatemática, do Programa de Pós-Graduação em Matemática da Universidade Federal de Santa Maria-RS. Configura-se como um estudo sobre um modelo discreto para um sistema praga-predador-inseticida tendo como ferramenta básica para sua formulação a Rede de Mapas Acoplados. Devido aos sérios problemas que pragas de lavoura representam para agricultura, abordou-se o Manejo Integrado de Pragas (MIP), que é um sistema de controle de pragas que busca preservar e aumentar os fatores de mortalidade natural de pragas pelo uso integrado de métodos de controle selecionados com base em parâmetros técnicos, econômicos, ecológicos e sociológicos. O objetivo deste trabalho é construir um modelo tão simples quanto possível, para estudar as consequências de diferentes estratégias de manejos de pragas. Essas estratégias são divididas em tratamentos diferentes nos quais leva-se em consideração o número de medições da população de pragas, estabelecimento de níveis de controle (NC) e diferentes medidas de controle aplicadas (inseticida e controle biológico). Para comparar e apontar qual dos tratamentos é mais eficiente, computou-se a densidade total da população de pragas sem nenhuma medida de controle e quando cada um dos tratamentos é aplicado ao longo de um determinado período de tempo. A diferença entre esses valores é convertida em um percentual de decrescimento da população de pragas. O mesmo é feito para se obter o percentual de decrescimento do número de sítios tratados. Foi possível constatar que quanto mais cedo se aplica o inseticida no sistema, mais efetivo é o tratamento. Além disso, observou-se que a presença dos predadores (controle biológico) diminui a quantidade de sítios tratados e, em alguns momentos, é menos eficiente no controle da praga.

L’agriculture française est soumise à de fortes pressions. Elle subit une injonction forte à évoluer vers l’agroécologie, à adopter les principes de la production intégrée, à réduire l’usage de pesticides. Dans ces conditions, une transition vers des systèmes agricoles plus respectueux de l’environnement ne doit pas dégrader la compétitivité des exploitations et le revenu des agriculteurs.Le travail présenté ici a pour ambition de contribuer à la réflexion sur les possibilités de changer les modèles d’agriculture. L’objectif est d’évaluer les conséquences économiques et environnementales d’une adoption généralisée des principes de protection intégrée sur une région agricole en fonction de la diversité des situations de production. La Région Bourgogne est retenue dans le cadre de ce travail. Elle dispose de ressources expérimentales en production intégrée grâce à l’investissement historique de la recherche, du développement et de l’enseignement agricole dans cette région. Par ailleurs, la Bourgogne présente un bon compromis entre la diversité de ses situations de production en grandes cultures et le nombre de situations contrastées à considérer pour rendre compte de l’ensemble du territoire. Dans un premier temps, nous avons caractérisé la diversité des situations de production de la Bourgogne et la diversité des systèmes de culture, et nous avons étudié comment les systèmes de culture sont déterminés par les situations de production. Puis nous avons mis en place une méthode pour générer des systèmes de culture optimisés sur le plan économique dans le cadre des contraintes locales et des principes de protection intégrée. Cette méthode mobilise la programmation linéaire sous contrainte, mise en œuvre avec le logiciel GAMS. Elle a été testée pour deux situations de production contrastées de la Bourgogne, correspondant aux zones de plaines sans élevage et aux plateaux argilo-calcaires superficiels. Nous avons évalué les systèmes de culture simulés pour un ensemble de critères de la durabilité économique, sociale et environnementale. Nous les avons comparés aux performances des systèmes actuels dans ces milieux.Les résultats confirment que les systèmes de culture optimisés avec les principes de la protection intégrée sont différents en fonction du contexte de production. La performance économique est plutôt améliorée par rapport aux systèmes actuels dans la situation à fort potentiel, alors qu’elle est dégradée sur les sols de plateaux. Tous les indicateurs environnementaux considérés sont améliorés dans les deux types de milieu. Le travail permet d’identifier certains inconvénients de la protection intégrée, liés à l’augmentation de la charge de travail et à la complexité de gestion des systèmes.Avec les résultats obtenus et les méthodes proposées, le travail alimente les débats sur la durabilité des systèmes agricoles dans le domaine des grandes cultures. / The agriculture in France is under intense pressure. Farmers are asked to change their crop management, to evolve toward agroecology, to follow the principles of Integrated Pest Management (IPM) and to reduce pesticide use. However such an evolution toward an alternative agricultural model will be possible only if innovative cropping systems are able to maintain competitive and profitable farms.The aim of our work is to contribute to the debates about this issue, by evaluating ex ante the potential consequences of adopting the principles of IPM over one whole agricultural region, taking into account the diversity of production situations within this region. We considered the Burgundy region for several reasons. This region has many experimental resources and available expertise on IPM. This agricultural area also has few but contrasted production situations. Both economic and environmental issues were considered.First, we studied the diversity of production situations and of cropping systems in the area, based on agricultural data sets, and we showed to what extent cropping systems are determined by the context. Then we developed a method to design fictive cropping systems, based on the optimisation of profitability while fulfilling the principles of IPM, and considering the constraints of the production situation. We used the GAMS software to implement this method based on linear programming. The method was tested on two contrasted production situations, namely the lowlands with high agricultural potential and no livestock, and the shallow soils of the plateau. We evaluated the generated cropping systems for a range of criteria covering different issues of sustainability, and we compared the performances to those of current cropping systems in these areas.Results corroborated that IPM-based cropping systems would be different in contrasted production situations. Profitability would be improved in the lowlands with high agricultural potential, whereas it would be negatively affected in the shallow soils of the uplands. All the environmental indicators that we used would be improved in both situations. Our work identified hindrances for the development of IPM, related for example to the increase in the workload at the farm level, and the increase in the system complexity.Both methods that we developed and the results we obtained should contribute to the current debates about the possible transition of arable cropping toward sustainability.

This thesis describes the design and subsequent implementation of a unit for traffic monitoring using Raspberry PI. First section provides a quick overview of assistance systems, which use a road lane detection techniques. Next there is a description of two diferent methods for road lane detection. Follow the description of monitoring scene. Then the work describe the practical part including the design and realization of supporting electronics, selecting of each components, including the modifying of cameras, mechanical design and creating of unit. Another section is about selection and installation of appropriate software components necessary for running of the unit and the selection of development tools for creating user application. After description of graphical user interafce, there is a description of road lanes detection algorithm. At the end of the thesis is summarized a reliability of unit in real traffic situation. At the appendix there are technical drawings, describing the unit.

No description available.

Phenacoccus peruvianus (Hemiptera: Pseudococcidae) is an invasive mealybug of Neotropical origin, first reported in the Mediterranean Basin in Almeria (Spain) in 1999. In the following years the mealybug spread into other Mediterranean regions and has also been recorded in Portugal and France, as well as in Sicily, Corsica and the Balearic Islands. Phenacoccus peruvianus is a polyphagous species and damages economically important ornamental plants. Since this was a relatively unknown species, during the first years of invasion, the mealybug was managed by the application of chemical treatments with wide-spectrum pesticides. However, the latest European directive on pesticide use reduces or even forbids pesticide applications in a wide range of urban green areas, giving significant priority to biological control (European Parliament and Council 2009). This thesis sets the basis for introducing biological control into a P. peruvianus management program in urban landscapes, focusing on its characterization, sampling, biology and control. In order to facilitate the identification of this and other mealybug species, we characterised 33 mealybug populations infesting crops and ornamental plants in Eastern Spain, using a combination of molecular and morphological techniques. This characterisation led to the identification of ten mealybug species and made routine identification possible through DNA sequencing or the use of derived species-specific molecular tools. The sequences obtained also add to the phylogenetic knowledge of the Pseudococcidae family and provide insight into the invasion history of some species. Phenacoccus peruvianus populations were high in bougainvillea plants during spring and summer, declining to almost undetectable levels in autumn and winter. The mealybug was mainly found in bracts and there were no significant migrations between plant strata. Phenacoccus peruvianus showed a high aggregated distribution on bracts, leaves and twigs. We recommend a binomial sampling of 200 leaves and an action threshold of 55% infested leaves for IPM purposes in urban landscapes. Its most abundant natural enemies were found to be the primary parasitoids Acerophagus n. sp. near coccois and Leptomastix epona Walker (Hymenoptera: Encyrtidae). We also identified several predator species from the Anthocoridae, Coccinellidae, Chamaemyiidae, and Chrysopidae families. Phenacoccus peruvianus populations were lower during the second and third year of the survey, coinciding with an increase in the parasitoid Acerophagus sp. populations, which displaced the native L. epona. Differential female offspring and resource preemption are discussed as the main reasons for this displacement. To obtain further information on the biology of the new parasitoid Acerophagus sp. we determined some traits of its reproductive and feeding strategies. Acerophagus sp. egg load reached its maximum when it was 5 days old with almost 30 mature eggs. Phenacoccus peruvianus second and third nymphal instars and adults were suitable for parasitism and efficient encapsulation was low (10.76 ± 0.31 %). The parasitoid always preferred older instars when different host instars were available. Acerophagus sp. developed as a solitary parasitoid in the second instar and as a gregarious parasitoid in older instars (2¿4 parasitoids per host). Moreover, it reproduced parthenogenetically and all the emerged offspring were females. Immature development lasted between 20 and 22 days at 25°C and 65% HR. Under these conditions, adults lived for longer than 20 days when fed on honey, but fewer than 3 days when fed on naturally occurring sugar sources (host honeydew and Bougainvillea glabra flowers). / Beltrà Ivars, A. (2014). Biology and management of the invasive mealybug Phenacoccus peruvianus (Hemiptera: Pseudococcidae) in urban landscapes [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/37233

Tesis por compendio / Delottococcus aberiae (Hemiptera: Pseudococcidae) es un pseudocóccido invasor originario del África subsahariana que fue detectado causando graves daños en cítricos del este de España en el año 2009. Debido al desconocimiento existente sobre esta especie, la gestión de D. aberiae se ha llevado a cabo mediante el uso de tratamientos químicos autorizados contra este tipo de insectos. Sin embargo, la última Directiva Europea (2009/128/EC) sobre el uso sostenible de productos fitosanitarios estipula que la aplicación de plaguicidas en el ámbito agrícola debe reducirse, promoviendo estrategias de manejo más sostenibles como es la aplicación del control biológico de plagas. Además, cuando una especie invasora llega por primera vez a un territorio es necesario estudiar su biología, comportamiento, daños causados y posibilidades de control. En esta tesis se presentan por primera vez estos estudios de biología y comportamiento de la plaga así como un análisis de los daños producidos por D. aberiae. También se han realizado estudios con el objetivo de implementar la aplicación de un programa de control biológico clásico sobre esta especie en cítricos en España. Para analizar la biología y comportamiento de D. aberiae se han muestreado durante tres años varias parcelas de cítricos con poblaciones de D. aberiae en la Comunidad Valenciana. En estas parcelas se han recogido muestras y contabilizado todos los pseudocóccidos presentes, su estadío de desarrollo y el estrato y órgano del árbol donde se encontraban. El periodo de daños al fruto fue estudiado en semicampo y campo mediante la infestación artificial con D. aberiae de frutos de distintos tamaños de diámetro. Por último se estudió el comportamiento y las posibilidades de control biológico de D. aberiae muestreando diversas parcelas de cítricos en su zona de origen (Sudáfrica). Los resultados mostraron que la abundancia de las poblaciones de D. aberiae en cítricos es elevada en primavera y verano, reduciéndose a niveles mucho más bajos en otoño e invierno. Además el insecto completa varias generaciones a lo largo del año, estando dos de ellas muy claramente definidas y siendo las que dan lugar a elevadas poblaciones de la plaga. En cuanto a su distribución, D. aberiae se instaló principalmente en la copa del árbol y se observaron migraciones entre los distintos órganos, mostrando una clara preferencia por el fruto en desarrollo. Entre febrero y septiembre parte de las poblaciones de D. aberiae se encontraron en tronco y suelo, existiendo movimientos de subida y de bajada a la copa en función de la fenología de la planta y las condiciones climáticas. La comparativa entre técnicas de muestreo reveló que las trampas de cartón corrugado proporcionan una medida cuantitativa de la abundancia de D. aberiae en las parcelas. Por su parte, las trampas pegajosas, provistas de hembras de D. aberiae, fueron capaces de detectar los principales vuelos de machos. Por otro lado, D. aberiae causó daños directos al fruto (deformación y/o reducción de tamaño) al alimentarse del ovario de la flor o de los primeros estados de desarrollo de éste. Estos daños son posiblemente debidos a su interferencia con el proceso de división celular. Por último, en Sudáfrica, lugar de origen de la plaga, los mayores niveles poblaciones de D. aberiae se encontraron en verano y la tasa de parasitismo fue máxima en otoño. Entre el complejo de enemigos naturales de D. aberiae encontrados destacaron dos especies, Anagyrus sp. nov. 1 (Hymenoptera: Encyrtidae) y Allotropa sp. nov. (Hymenoptera: Platygastridae). Ambos parasitoides podrían tener un papel importante en un programa de control biológico contra D. aberiae en España. Por ahora, Anagyrus sp. nov. 1 parece el mejor candidato por sus mayores tasas de parasitismo en Sudáfrica. / Delottococcus aberiae (Hemiptera: Pseudococcidae) is an invasive mealybug native to sub-Saharan Africa that was detected causing significant damage to citrus fruits in eastern Spain in 2009. Due to the lack of knowledge about this species, the management of D. aberiae has been carried out by the application of authorized insecticides against mealybugs. However, the latest European Directive (2009/128 / EC) on the sustainable use of pesticides stipulates that chemical treatments in agroecosystems must be reduced, promoting more sustainable management strategies such as the application of biological control methodologies. In addition, when an invasive species arrives for the first time in a territory it is necessary to study its biology, behavior, damage caused and control possibilities. This thesis presents for the first time these studies of biology and behavior of the pest as well as a characterization of the damage produced by D. aberiae. The possibilities of implementing a classical biological control program against this species in citrus in Spain have also been studied. To analyze the biology and behavior of D. aberiae, several citrus orchards infested with the mealybug have been sampled for three years in the Valencian Community (eastern Spain). Samples have been collected periodically and the number of mealybugs, their developmental stage as well as the infested stratum and organ where they were present have being recorded. The period of damage to the fruit was studied in semi-field and field conditions by the artificial infestation with D. aberiae of fruits of different diameter. Finally, the behavior and possibilities of biological control of D. aberiae were studied by sampling several citrus orchards in the native area of the mealybug (South Africa). Results showed that the density of D. aberiae populations in citrus orchards is high in spring and summer, decreasing to lower levels in autumn and winter. In addition, the insect completes several generations throughout the year and two of them are clearly defined and result in high population levels. Regarding its distribution, D. aberiae was mostly installed in the canopy of the tree and migrations were observed between different organs, showing a clear preference for the developing fruit. From February to September some mealybugs were found in the trunk and soil, moving upwards or downwards depending on the phenology of the plant and the climatic conditions. The comparison between sampling techniques revealed that corrugated cardboard band traps provide a quantitative measurement of D. aberiae density in the orchards. On the other hand, sticky traps, baited with D. aberiae females were able to detect the main male flight periods. D. aberiae caused direct damage to the fruit (deformation and/or reduction in size) by feeding on the ovary of the flower or on small fruits in development. These damages are probably due to their interference with the process of cell division. Finally, in South Africa, native area of the pest, the highest density levels of D. aberiae were found in summer and the highest parasitism rates occurred in autumn. Among the complex of D. aberiae natural enemies, the two most abundant species were Anagyrus sp. nov. 1 (Hymenoptera: Encyrtidae) and Allotropa sp. nov. (Hymenoptera: Platygastridae). Both parasitoids could play an important role in a biological control program against D. aberiae in Spain. For now, Anagyrus sp. nov. 1 seems the best candidate because of its higher rates of parasitism in South Africa. / Delottococcus aberiae (Hemiptera: Pseudococcidae) és un pseudocòccid invasor originari de l'Àfrica subsahariana que va ser detectat causant greus danys en cítrics de l'est d'Espanya l'any 2009. A causa del desconeixement existent sobre aquesta espècie, la gestió de D. aberiae s'ha dut a terme mitjançant l'ús de tractaments químics autoritzats contra aquest tipus d'insectes. No obstant això, l'última Directiva Europea (2009/128/EC) sobre l'ús sostenible de productes fitosanitaris estipula que l'aplicació de plaguicides en l'àmbit agrícola ha de reduir-se, promovent estratègies de maneig més sostenibles com és l'aplicació del control biològic de plagues. A més, quan una espècie invasora arriba per primera vegada a un territori és necessari estudiar la seua biologia, comportament, danys causats i possibilitats de control. En aquesta tesi es presenten per primera vegada els estudis de biologia i comportament de la plaga així com una anàlisi dels danys produïts per D. aberiae. També s'han realitzat estudis amb l'objectiu d'implementar l'aplicació d'un programa de control biològic clàssic sobre aquesta espècie en cítrics a Espanya. Per a analitzar la biologia i comportament de D. aberiae s'han mostrejat durant tres anys diverses parcel·les de cítrics amb poblacions de D. aberiae a la Comunitat Valenciana. En aquestes parcel·les s'han recollit mostres i comptabilitzat tots els pseudocòccids presents, el seu estadi de desenvolupament i l'estrat i òrgan de l'arbre on es trobaven. El període de danys al fruit va ser estudiat en semicamp i camp mitjançant la infestació artificial amb D. aberiae de fruits de diferents mides de diàmetre. Finalment es va estudiar el comportament i les possibilitats de control biològic de D. aberiae mostrejant diverses parcel·les de cítrics en la seua zona d'origen (Sud-àfrica). Els resultats van mostrar que l'abundància de les poblacions de D. aberiae en cítrics és elevada a la primavera i estiu, reduint-se a nivells molt més baixos a la tardor i hivern. A més l'insecte completa diverses generacions al llarg de l'any, estant dos d'elles molt clarament definides i sent les que donen lloc a elevades poblacions de la plaga. Quant a la seua distribució, D. aberiae es va instal·lar principalment en la copa de l'arbre i es van observar migracions entre els diferents òrgans, mostrant una clara preferència pel fruit en desenvolupament. Entre febrer i setembre part de les poblacions de D. aberiae es van trobar en tronc i sòl, existint moviments de pujada i de baixada a la copa en funció de la fenología de la planta i les condicions climàtiques. La comparativa entre tècniques de mostreig va revelar que les trampes de cartró corrugat proporcionen una mesura quantitativa de l'abundància de D. aberiae en les parcel·les. Per la seua banda, les trampes apegaloses proveïdes de femelles de D. aberiae van aconseguir detectar els principals vols de mascles. D'altra banda, D. aberiae va causar danys directes al fruit (deformació i/o reducció de mida) en l'alimentar-se de l'ovari de la flor o dels primers estats de desenvolupament d'aquest. Aquest danys són possiblement deguts a la seua interferència amb el procés de divisió cel·lular. Finalment, a Sud-àfrica, lloc d'origen de la plaga, els majors nivells poblacionals de D. aberiae es van trobar a l'estiu i la taxa de parasitisme va ser màxima a la tardor. Entre el complex d'enemics naturals de D. aberiae trobats van destacar dues espècies, Anagyrus sp. nov. 1 (Hymenoptera: Encyrtidae) i Allotropa sp. nov. (Hymenoptera: Platygastridae). Tots dos parasitoides podrien tindre un paper important en un programa de control biològic contra D. aberiae a Espanya. Per ara, Anagyrus sp. nov. 1 sembla el millor candidat per les seues majors taxes de parasitisme a Sud-àfrica. / Martínez Blay, V. (2018). Biology and management, by application of classical biological control, of the invasive mealybug Delotoccoccus aberiae (Hemiptera:Pseudococcidae) in citrus orchards in Spain [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/106363 / Compendio

Les cultures de Brassicacées sont très courantes en agriculture urbaine à Montréal. Elles comportent cependant leur lot de problèmes liés aux insectes ravageurs, tels que la piéride du chou, Pieris rapae. L’utilisation de méthodes de lutte classique contre les insectes ravageurs, tels les insecticides, engendrent de sérieux dommages environnementaux, incluant la contamination de l’eau et du sol ainsi que la toxicité pour les plantes environnantes, les insectes auxiliaires et les microorganismes du sol. Le compagnonnage végétal, une pratique agroécologique alternative, est connue en tant que stratégie d’IPM (lutte intégrée contre les insectes ravageurs). L'objectif général de l'étude était d'évaluer in situ l'importance de diverses méthodes de compagnonnage sur l'infestation de P. rapae au sein du chou chinois, Brassica rapa. Les plantes utilisées dans les systèmes de compagnonnage étaient des cultures-pièges (Eruca sativa et Brassica carinata), des plantes compagnes principales qui étaient des plantes insectaires et répulsives pour les insectes ravageurs (Tagetes erecta, Amaranthus cruentus et Ocimum grattissimum) ainsi que des plantes compagnes secondaires (Solanum aethiopicum, Ocimum basilicum et Hibiscus sabdariffa). Les paramètres de physiologie végétale ainsi que les structures des communautés microbiennes et d’insectes ont été suivis de près tout au long de la saison croissance où l’expérience a eu lieu. La taille des larves était significativement plus importante au sein de la monoculture (contrôle) et les taux de concentration de glucosinolates dans les feuilles de Brassica rapa étaient deux fois plus élevés dans les contrôles comparativement aux systèmes de cultures plus diversifiés. Une PERMANOVA a confirmé une différence significative entre les méthodes de cultures concernant la composition des communautés d’insectes bénéfiques. Les communautés bactériennes du sol ont été améliorées par l'agriculture durable par rapport au sol d'origine et ont été bonifiées dans les systèmes de compagnonnage végétal (plus diversifiés). L'étude a mis en évidence les avantages de l'agroécologie, y compris le compagnonnage végétal, en termes de lutte intégrée contre les insectes ravageurs et d'autres composantes de l'agroécosystème. / Brassica crops are very common culture for urban farmers in Montreal where insect pests like the cabbage whitefly, Pieris rapae, are a real concern. The use of conventional insect pest control methods, such as insecticides, causes serious damages to a highly anthropized and therefore already constrained environment. These include air, water and soil contamination, as well as toxicity to surrounding beneficial insects, soil microorganisms, plants and the entire food chain. Companion planting, an agroecological practice, is an alternative strategy for insect pest management. The general objective of the study was to evaluate in situ the relevance of various traditional methods of companion planting on infestation of Chinese cabbage, Brassica rapa, by P. rapae. The plants used in the companion systems were two trap crop species: Eruca sativa and Brassica carinata; Tagetes erecta, Amaranthus cruentus and Ocimum grattissimum were used like companion plants because of their properties as insectary plants and repellent species against the targeted pest; Solanum aethiopicum, Ocimum basilicum and Hibiscus sabdariffa were also considered (secondary) companion plant species. Plant physiological parameters as well as microbial and insect community structure were carefully monitored over the growing season where this experiment took place. Larval size was significantly greater in the monoculture and glucosinolate concentrations in leaves of Brassica rapa was two-times higher in control than in more diversified cultivation systems. With larger yield and the absence of pest in the field, ‘trap cropping system’ was overall the most efficient albeit the 'mix of companion planting’ system had the lowest leaf area damage. A PERMANOVA confirmed a significant difference between the cultivation methods regarding beneficial insect communities’ composition. Belowground, soil bacterial communities were readily modified by sustainable agriculture practice, even more so in biodiversified systems. The study highlighted benefits of agroecology, including companion planting, in terms of integrated pest management and other components of the agroecosystem.

Regression is a central problem in statistics and machine learning with applications everywhere in science and technology. In probabilistic regression the relationship between a set of features and a real-valued target variable is modelled as a conditional probability distribution. There are cases where this distribution is very complex and not properly captured by simple approximations, such as assuming a normal distribution. This thesis investigates how conditional Generative Adversarial Networks (GANs) can be used to properly capture more complex conditional distributions. GANs have seen great success in generating complex high-dimensional data, but less work has been done on their use for regression problems. This thesis presents experiments to better understand how conditional GANs can be used in probabilistic regression. Different versions of GANs are extended to the conditional case and evaluated on synthetic and real datasets. It is shown that conditional GANs can learn to estimate a wide range of different distributions and be competitive with existing probabilistic regression models.