Cement production is an energy intensive process. In South Africa the cost of energy increased since 2006, while cement sales have dropped dramatically. It has become important to focus on methods to optimise energy consumption to achieve cost savings in the cement industry. Various methods of reducing production cost by improving energy efficiency are available, but require extended installation periods and high initial capital expenditure. Other methods such as operational optimisation can reduce production cost, but offer limited savings. The aim of this study is to integrate the optimisation of multiple component operations to improve savings and reduce interruption during implementation. Although integrated optimisation models have been developed, no literature could be found on the application of these models in the cement industry. This thesis reports on the development and implementation of an energy management system at four South African cement plants. The total electricity costs were reduced without installing costly infrastructure upgrades. The results summarise the success of the improved production planning. A conclusion regarding the feasibility of this implementation is compiled by comparing the savings achieved by the implementation of the energy management system to other energy saving methods. Recommendations are also made for further study and the implementation of the energy management system in similar industries. / Thesis (MIng (Mechanical Engineering))--North-West University, Potchefstroom Campus, 2013.

Smart Grid has emerged a very important concept in modern power systems. The integration of different loads such as residential, commercial and industrial into the smart grid and their optimal operation has a significant effect on the system's reliability, stability, peak power demand and energy price. This work presents the mathematical modeling of a Centralized Temperature Control System (CTCS) of a Multi-Unit Residential Building (MURB) and its optimal operation considering electricity prices and weather variations. The model considers comfort levels, preference settings and activity of residents in different units of the building to determine the optimal operation schedules of the CTCS, minimizing its total energy consumption cost. Multi-objective operation of the MURB is also investigated when residents in different units have conflicting interests, and the impact of such conflicting preferences on the operation of CTCS is analyzed. A case-study on optimal energy management of a single unit house considering net-metering is also presented. The proposed CTCS model is a Mixed Integer Non Linear Programing (MINLP) model, where some of the constraints are linearized to reduce the computational complexity arising from the non-linearity, for real-time applications. The model is studied for various customers' preferences using a realistic MURB model. Simulation results show that significant cost savings can be achieved using the proposed mathematical model.

1ST Otokon, a Pretoria based company delivering energy management systems to the industrial and utility market segments, has shown steady growth since its inception 15 years ago. However, after ailing financial results, all aspects of the business were interrogated. It was doubted whether its energy management system (EMS) offering, ecWIN, was fulfilling the needs of the industry. Therefore this study was needed to identify new technology which could improve the effectiveness of ecWIN. After doing a thorough literature study on ecWIN, its competitors and the technologies currently utilised, the product was mapped to various technology management models. Deficiencies in the current software development environment were identified, including a lack of manpower and old software development tools. The empirical study entailed determining the effectiveness of the product through a user survey in the form of a questionnaire. Certain problem areas (e.g. insufficient software testing and a lack of documentation on advanced features) were identified by this survey. The numerous complaints about the slowness of certain reports and lack in user-friendliness were also highlighted. Also forming part of the literature and empirical study respectively, were the identification and evaluation of new technologies not yet utilised in the ecWIN sphere. Taking a holistic view of all this information, recommendations were made on how to improve the effectiveness of ecWIN. This was presented in terms of three dimensions: • Upgrade the current infrastructure over the next 12 months, mainly with new software development tools such as Centura Team Developer 2005, Crystal Reports, etc. • Improve the cost-effectiveness of ecWIN during 2006 and 2007, typically by utilising GPRS, MMS, WiFi and BlueTooth to retrieve information from meters. • Utilise a new business model (mainly centred on the cost-effectiveness dimension together with Linux and MySQL) which could allow the company to directly compete with other firms, even in the commercial and residential markets. A fourth dimension, consisting of technologies that should not be considered for integration with ecWIN at this stage, was also identified (e.g. IPv6, XHTML and Visual Studio .NET). In a world with rapidly changing technology, this study should not be a once-off process, but should be undertaken regularly to ascertain that the effectiveness of 1ST Otokon's EMS offering is constantly improved through the identification of new technology. / Thesis (M.B.A.)--North-West University, Potchefstroom Campus, 2005.

1ST Otokon, a Pretoria based company delivering energy management systems to the industrial and utility market segments, has shown steady growth since its inception 15 years ago. However, after ailing financial results, all aspects of the business were interrogated. It was doubted whether its energy management system (EMS) offering, ecWIN, was fulfilling the needs of the industry. Therefore this study was needed to identify new technology which could improve the effectiveness of ecWIN. After doing a thorough literature study on ecWIN, its competitors and the technologies currently utilised, the product was mapped to various technology management models. Deficiencies in the current software development environment were identified, including a lack of manpower and old software development tools. The empirical study entailed determining the effectiveness of the product through a user survey in the form of a questionnaire. Certain problem areas (e.g. insufficient software testing and a lack of documentation on advanced features) were identified by this survey. The numerous complaints about the slowness of certain reports and lack in user-friendliness were also highlighted. Also forming part of the literature and empirical study respectively, were the identification and evaluation of new technologies not yet utilised in the ecWIN sphere. Taking a holistic view of all this information, recommendations were made on how to improve the effectiveness of ecWIN. This was presented in terms of three dimensions: • Upgrade the current infrastructure over the next 12 months, mainly with new software development tools such as Centura Team Developer 2005, Crystal Reports, etc. • Improve the cost-effectiveness of ecWIN during 2006 and 2007, typically by utilising GPRS, MMS, WiFi and BlueTooth to retrieve information from meters. • Utilise a new business model (mainly centred on the cost-effectiveness dimension together with Linux and MySQL) which could allow the company to directly compete with other firms, even in the commercial and residential markets. A fourth dimension, consisting of technologies that should not be considered for integration with ecWIN at this stage, was also identified (e.g. IPv6, XHTML and Visual Studio .NET). In a world with rapidly changing technology, this study should not be a once-off process, but should be undertaken regularly to ascertain that the effectiveness of 1ST Otokon's EMS offering is constantly improved through the identification of new technology. / Thesis (M.B.A.)--North-West University, Potchefstroom Campus, 2005.

Electric power engineering, as it applies to low-voltage power reticulation in buildings and industrial sites, is ripe for a ‘paradigm shift’ to bring it properly into the Electronic Age. The conventional alternating-current approach, now over a hundred years old, is increasingly unsatisfactory from the point of view of plant and appliance requirements. Alternative approaches can deliver substantial cost savings, higher efficiencies, power quality improvements, and greater safety. Power reticulation systems in the future can be expected to differ from present systems in two key respects. The first is a greatly increased role for direct current; the second is the augmentation of the power system with a wide range of ‘management’ technologies. Combining these two trends, which can already be observed today, leads to consideration of ‘managed DC’ power reticulation systems, operating from AC bulk supply mains via AC-DC converters.

Στόχος της παρούσας διπλωματικής εργασίας ήταν η μελέτη και η υλοποίηση στρατηγικής διαχείρισης ενέργειας για τη βελτιωμένη οικονομική λειτουργία υβριδικού οχήματος με χρήση ψηφιακού μικροελεγκτή. Η προτεινόμενη στρατηγική, η οποία βασίζει τη λειτουργία της στην αρχή της ασαφούς λογικής αναπτύχθηκε και αρχικά δοκιμάστηκε, χρησιμοποιώντας το πρόγραμμα Matlab/Simulink. Για την επιβεβαίωση της ορθής λειτουργίας του συνολικού συστήματος αποτελούμενου από το ηλεκτρικό κινητήριο σύστημα, τη μηχανή εσωτερικής καύσης και το σύστημα διαχείρισης της ενέργειας, δημιουργήθηκε ένα μαθηματικό μοντέλο ενός υβριδικού οχήματος παράλληλης διάταξης, επίσης στο περιβάλλον Simulink,στο οποίο η διαχείριση της ενέργειας γίνεται μέσω του ανεπτυχθέντος ασαφούς ελεγκτή. Στόχος ήταν να ελεγχθεί η συμπεριφορά του ελεγκτή αυτού. Τέλος, κατασκευάστηκε το κύκλωμα υλοποίησης του ασαφούς ελεγκτή, χρησιμοποιώντας τον ψηφιακό μικροελεγκτή dsPIC30f4011 της εταιρίας Microchip. Για την υλοποίηση της λειτουργίας του ελεγκτή, αναπτύχθηκε ο αντίστοιχος κώδικας σε γλώσσα C, η λειτουργία του οποίου επιβεβαιώθηκε πειραματικά. / The objective of this master thesis was the study and the implementation of an energy management strategy, aiming for improved economic operation of hybrid vehicle using a digital microcontroller. The proposed strategy, based on the principles of fuzzy logic, was developed and initially tested using the environment Matlab/Simulink. To ascertain the correct operation of the system constituted by the electric motive system, the internal combustion machine and the energy management system, a mathematic model of hybrid vehicle of parallel provision was also created in Simulink. The energy management is implemented by the fuzzy controller. The main objective was to test the behavior of this controller. Finally, the electronic circuit of the fuzzy controller was manufactured. The digital microcontroller dsPIC30f4011 (Microchip company) was used. For the implementation of the operation of the controller, the corresponding code was developed in C language, the operation of which was experimentally confirmed.

Thesis (MScEng)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: Energy efficiency, as the effective use of energy, is recognized as one of the simplest ways to improve the sustainable use of resources and by implication involves the end-user. The 2008 power crisis which South Africa experienced, highlighted supply exigencies and prompted a subsequent emphasis on affordable, rapidly scalable solutions, notably energy efficiency. As the establishment of new supply capacity is both costly and time-consuming, the logical alternative has been to focus intervention on the demand side. Residential electrical end-use has been identified as an area where the potential for change exists and strategies to address residential demand have gained momentum. The vulnerability of energy systems affects energy security on technical, economic and social levels. South African consumers are confronted with rising living costs and a substantial increase in electricity prices according to the Integrated Resource Plan for Electricity (2010-2030). Integral to addressing end-use is the ensuing behaviour of the end-user. End-use analysis aims to grasp and model customer usage by considering the electric demand per customer type, end-use category, appliance type and time of use. This project has focussed on the development of an interactive web application as a tool for residential end-users to improve energy efficiency through modified consumption behaviour and the adoption of energy efficient habits. The objectives have been aimed at educating an end-user through exposure to energy efficient guidelines and consumption analysis. Based on a Time Of Use-framework, a consumer’s understanding of appliance usage profiles can help realize the cost benefits associated with appliance scheduling. In order to achieve the desired functionality and with extendibility and ease of maintenance in mind, the application relies on the provision of dynamic content by means of a relational database structured around end-use categories and appliance types. In an effort to convey only relevant information in the simplest way, current web technology was evaluated. The resulting design has favoured an interactive, minimalistic, graphic presentation of content in the form of a Rich Internet Application. The development process has been divided into two phases. The residential energy consumption context has been substantiated with a case study of which the main objective and outcome has been to devise a methodology to generate usage profiles for household appliances. Phase one of the development process has been completed, as well as the case study. The conceptualization and framework for phase two has been established and the recommendation is to incorporate the methodology and usage profile results from the case study for implementation of the second phase. The effectiveness of the tool can only be evaluated once phase two of the application is complete. A beta release version of the final product can then be made available to a focus group for feedback. / AFRIKAANSE OPSOMMING: Energie effektiwiteit, gesien as die effektiewe aanwending van energie, word herken as een van die eenvoudigste maniere om die volhoubare gebruik van hulpbronne te bevorder en betrek by implikasie die verbruiker. Die 2008 kragvoorsieningskrisis wat Suid-Afrika beleef het, het dringende tekorte aan die lig gebring en ’n gevolglike klemverskuiwing na bekostigbare, maklik aanpasbare oplossings, vernaamlik energie effektiwiteit. Aangesien die daarstelling van nuwe voorsieningskapasiteit beide duur is en baie tyd in beslag neem, was die voor die hand liggende alternatief om te fokus op vraag-kant toetrede. Huishoudelike elektriese verbruik is geïdentifiseer as ’n area waar die potensiaal vir verandering bestaan en strategieë om residensiële aanvraag aan te spreek het momentum gekry. Die kwesbaarheid van energiestelsels affekteer energie sekuriteit op tegniese, ekonomiese en sosiale vlakke. Suid-Afrikaanse verbruikers word gekonfronteer met stygende lewenskoste en ’n aansienlike toename in elektrisiteitspryse volgens die Geïntegreerde Hulpbron-Plan vir Elektrisiteit (2010-2030). Eie aan die aanspreek van verbruik is die voortvloeiende gedrag van die verbruiker. Verbruiksanalise poog om verbruik te begryp en te modelleer deur die elektriese aanvraag na gelang van verbruikerstipe, verbruikskategorie, toesteltipe en tyd van verbruik in aanmerking te neem. Hierdie projek het gefokus op die ontwikkeling van ’n interaktiewe web-toepassing as ’n instrument vir residensiële verbruikers om energie effektiwiteit te verbeter deur gewysigde verbruiksgedrag en die ingebruikneming van energie effektiewe gewoontes. Die doelwitte is gerig op die opvoeding van ’n verbruiker deur blootstelling aan riglyne vir energie effektiewe verbruik en verbruiksanalise. Gebaseer op ’n Tyd-Van-Verbruik-raamwerk, kan ’n verbruiker se begrip van toestelle se verbruiksprofiele ’n bydrae lewer om die koste-voordele geassosieer met toestel-skedulering te realiseer. Om sodoende die verlangde funksionaliteit te bewerkstellig en met verlengbaarheid en gemak van onderhoud voor oë, steun die toepassing op die verskaffing van dinamiese inhoud deur middel van ’n relasionele databasis wat gestruktureer is rondom verbruikskategorieë en toesteltipes. In ’n poging om slegs toepaslike informasie in die eenvoudigste vorm weer te gee, is teenswoordige web tegnologie geevalueer. Die vooruitspruitende ontwerp is ’n interaktiewe, minimalistiese, grafiese aanbieding van die inhoud in die vorm van ’n sogenaamde "Rich Internet Application". Die ontwikkelingsproses is ingedeel in twee fases. Die huishoudelike energieverbruikskonteks is bevestig deur middel van ’n gevallestudie waarvan die vernaamste doelwit en uitkoms was om ’n metodologie daar te stel om verbruiksprofiele van huishoudelike toestelle te genereer. Fase een van die ontwikkelingsproses is voltooi asook die gevallestudie. Die konsepsuele onwikkeling en raamwerk vir fase twee is reeds gevestig en die aanbeveling is om die metodologie en verbruiksprofielresultate van die gevallestudie te inkorporeer vir implementering van die tweede fase. Die effektiwiteit van die toepassing kan eers geevalueer word sodra fase twee afgehandel is. ’n Beta-weergawe vrystelling van die finale produk kan dan beskikbaar gestel word aan ’n fokusgroep vir terugvoer.

Actuellement, les stratégies de gestion de l’énergie dans les réseaux intelligents sont pour la plupart limitées à l’intérêt d’un sous-système. En règle générale, chaque acteur est géré de façon autonome sans tenir compte du fait qu’il est intégré dans un réseau électrique à proximité. Par exemple, un système de gestion de l’énergie des bâtiments vise à fournir le niveau de service souhaité aux occupants et ne se soucie pas de son impact sur le réseau, sauf s’il doit en respecter certaines contraintes.Cette manière de gérer peut conduire bien entendu à un équilibre donné mais la résultante ne sera qu’un ensemble de sous-systèmes optimisés qui amèneront rarement à un optimum global dans la poche à laquelle ils appartiennent.Compte tenu de ce qui est dit ci-dessus, et au vu d’une architecture de réseaux de distribution en évolution rapide; la restructuration physique et algorithmique en sous réseaux physiques ou virtuels permettra de répondre efficacement aux problématiques liées à :— La sûreté de la fourniture— L’intégration massive de renouvelable— La qualité de l’énergie— L’apparition de nouvelles charges non conventionnelles— Aux services systèmesDans la littérature, les aspects du contrôle et de la gestion de l’énergie de microréseaux sont traités séparément, et l’interaction de réseau intelligent est simplement proposée.Pour relever ces défis, le concept de réseaux intelligents est apparu au cours de la dernière décennie. Il s’appuie sur les capacités des systèmes de communication modernes qui permettent le flux continu de données entre les acteurs d’un réseau intelligent et sur les capacités de calcul évolutives permettant de mettre en œuvre des stratégies avancées de gestion de l’énergie à grande échelle.Cette thèse se propose de mener une étude systémique du contrôle de microréseaux lequel contrôle vise une gestion optimisée de l’énergie en lien avec une structure de ce qui est communément appelé « réseau intelligent » et ce, tout en optimisant la puissance locale sous un modèle prédictif de contrôle (MPC).Le MPC se distingue parmi les stratégies avancées de contrôle de réseau pour plusieurs raisons. D’abord, il permet de traiter facilement des systèmes multi variables qui sont soumis à de multiples contraintes. En second lieu, il est capable d’anticiper les événements futurs en tenant compte des prévisions (par exemple, prévisions météorologiques, prévisions de charges, ...). Pour ces raisons, une partie de cette thèse est dédiée aux algorithmes MPC qui visent à coordonner de manière optimale un grand nombre d’acteurs dans un microréseau (PV, Batteries, Éolienne, charges, ...). L’idée est d’avoir un contrôleur MPC local pour chaque microréseau et au-dessus, un coordinateur de contrôleur de gestion MPC qui influence le contrôleur local de telle manière que l’optimalité globale du réseau intelligent soit respectée. L’objectif de maximiser la consommation locale d’énergie produite localement est considéré. Cet objectif est une étape vers l’indépendance énergétique des microréseaux locaux vis à vis du réseau principal lequel toutefois peut intervenir pour acheter l’excès de puissance de l’ensemble des microréseaux de la coopérative.Cette thèse a été préparée en co-tutelle entre le Gipsa-Lab de l’Université Grenoble-Alpes (UGA) et le PREEA de l’université de technologie et de sciences appliquées libano-française dans l’application du projet PARADISE.Ce dernier projet vise par ses contributions à optimiser des réseaux de distribution ilôtables en présence d’un fort taux de production intermittente à base de renouvelable ; et ce, par des architectures physiques et algorithmiques incrémentales. / Currently, energy management strategies in smart grids are mostly limited to the interest of a subsystem. As a general rule, each actor is autonomously managed regardless of whether it is integrated into a nearby power grid. For example, a building energy management system aims to provide the desired level of service to occupants and does not care about its impact on the system unless it has to meet certain constraints.This way of managing can of course lead to a given equilibrium but the resultant will be only a set of optimized subsystems that will rarely lead to an overall optimum in the pocket to which they belong.In view of what has been said above, and in view of a rapidly evolving distribution system architecture; The physical and algorithmic restructuring in physical or virtual sub networks will allow to answer efficiently the problems related to:- Security of supply- Massive integration of renewable energy- The quality of energy- The appearance of new unconventional loads- System servicesIn the literature, aspects of microgrid energy control and management are treated separately, and intelligent network interaction is simply proposed.To meet these challenges, the concept of smart grids has emerged over the last decade. It builds on the capabilities of modern communication systems that enable the continuous flow of data between the players in an intelligent network and the scalable computing capabilities to implement advanced large-scale energy management strategies ladder.This thesis proposes to carry out a systemic study of the control of microgrid which control aims at an optimized management of the energy in connection with a structure of what is commonly called "intelligent network", while optimizing the local power under a model Predictive control (MPC).The MPC stands out among advanced network control strategies for several reasons. Firstly, it makes it possible to easily handle multi-variable systems which are subjected to multiple constraints. Secondly, it is able to anticipate future events by taking into account forecasts (for example, weather forecasts, forecast loads, etc.). For these reasons, part of this thesis is dedicated to MPC algorithms which aim to coordinate optimally a large number of actors in a microgrid (PV, Batteries, Wind, loads ...). The idea is to have a local MPC controller for each microgrid and above it, an MPC management controller coordinator that influences the local controller in such a way that the overall optimality of the intelligent network is respected. The objective of maximizing local consumption of locally produced energy is considered. This objective is a step towards the energy independence of the local microgrids with respect to the main network, which however can intervene to buy the excess power of all microgrids of the cooperative.This thesis was prepared in co-supervision between the Gipsa-Lab of the Grenoble-Alpes University (UGA) and the PREEA of the Lebanese-French University of Technology and Applied Sciences in the application of the PARADISE project.This project aims, through its contributions, to optimize distribution networks that are portable in the presence of a high rate of intermittent production based on renewable energy; And this, by physical architectures and incremental algorithm.

Actuellement, l'inconfort intérieur dans les bâtiments est l'une des questions cruciales, ainsi que la consommation énergétique du bâtiment. En effet, les gens passent 60 à 90% de leur vie dans les bâtiments. Le confort intérieur est indispensable en termes de bienfaits sur la santé, la productivité et le bien-être des occupants. C'est pourquoi diverses stratégies d'optimisation fondées sur des règles, anticipatives ou prédictives ont été proposées pour atteindre le confort perçu en tenant compte de la consommation d'énergie. Dans la pratique, il existe un écart entre l'anticipation et la réalité. Habituellement, les plans anticipatifs sont synchronisés avec une période d'anticipation d'une heure et ne tiennent pas compte des différentes sources de conflits ainsi que des configurations d'enveloppes des bâtiments possibles. Il peut en résulter des conséquences négatives sur le coût et le confort. Pour résoudre ce problème, le système de gestion du bâtiment doit être conçu aussi comme réactif, de sorte à ce qu'il puisse répondre à toutes sortes de divergences par rapport au plan anticipatif, de manière réactive. Dans ce but, un système multi-échelle d'analyse de diagnostic réactif anticipatif (ARD- BMS) est proposé dans cette thèse. ARD-BMS est une gestion interne qui effectue trois actions importantes, c'est-à-dire la détection de défaut, l'isolement des causes et, enfin, les actions correctives. ARD-BMS permet ainsi une gestion réactive à court terme, à savoir 10 minutes pour analyser les tendances des défauts et la mise à jour de la dynamique du bâtiment et prendre ainsi les mesures correctives nécessaires pour maintenir le niveau de confort désiré. Cette thèse propose donc un modèle réactif à dynamique rapide simplifiée qui peut être utilisé pour estimer l'état actuel du bâtiment. Les bâtiments modernes sont un système très sophistiqué avec un grand nombre de capteurs, de contrôleurs et de CVC. La plupart des installations de construction utilisent des services prévus de maintenance préventive provenant des opérations périodiques des bâtiments. Ces problèmes imprévus puce causer des répercussions inexplicables sur le confort de l'occupant pendant le cycle de fonctionnement de 24 heures. Ces problèmes ne sont pas inadmissibles tels que les situations imprévues, les pannes de prévisions météorologiques. Le diagnostic des causes d'inconfort à court terme est encore un problème difficile au niveau de l'opération de construction intégrale. En outre, pour analyser cette situation, proposez une méthodologie diagnostique pour la détection et l'isolement des causes (fautes) dans les bâtiments. La méthodologie proposée comprend une HAZOP fondée sur les règles (analyse des risques et de l'optimisation) et une approche basée sur un modèle. / Currently, indoor discomfort in dwellings is one of the crucial issues along with the building energy consumption. Indeed, people spend 60-90% of their lives in buildings. Indoor comfort plays a vital role in occupants health, productivity, and well-being. However, various optimization and rule-based anticipative or predictive building strategies have been proposed to achieve the perceived comfort taking into account the energy consumption. However, in practice, anticipation or plans are far from the reality. Usually, anticipative plans are synchronized with one-hour anticipation period and do not consider the various sources of discrepancies as well as current envelope configurations. Unbeknownst to many, discrepancies from different sources could cause big penalty over cost and comfort. To tackle this issue, building management system needs to be designed as reactive or almost with no planning so that it can respond to all discrepancies re-actively. To address this problem, a multi-scale Anticipative Reactive Diagnosing- Building Management System (ARD-BMS) is proposed in this dissertation. ARD-BMS is an internal management and performs three important actions i.e., Discrepancydetection, Cause isolation, and finally Corrective actions. ARD-BMS follow the short-time resolution i.e., 10-minute to analyze the fault trends and current the building dynamics and take necessary corrective actions to maintain the desired level of comfort. This thesis proposes a fast dynamics simplified reactive model that can be used to estimate the current status of the building. Modern buildings are very sophisticated system with a large number of sensors, controllers, and HVACs. Most of the building facilities are using a scheduled preventive maintenance services derived from periodic operations of the buildings. These preventive actions do not take into account the other inadmissible issues such as unplanned situations, weather prediction failures etc. These unplanned issues could cause unaccountable impacts over occupant’s comfort during the 24-hour operation cycle. Diagnosability of short-term discomfort causes is still a challenging job at whole building operation level. Furthermore, to analyze this situation the thesis proposes a diagnostic methodology for detection and isolation of cause (faults) in buildings. The proposed methodology includes a rule-based HAZOP (Hazard and Operability analysis) and model-based approach. Further, in order to oversee unplanned discomforts, a short-term reactive optimization has been proposed.

Il a été développé 3 approches pour l'estimation d'occupation (nombre d'occupants et potentiellement leur activité) à base de techniques de machine learning :- une technique d'apprentissage supervisé exploitant différentes typologies de capteurs,- une technique à base de connaissance sans capteurs- et une technique interactive sollicitant lorsque c'est opportun les occupants via des questions. / Building energy management and monitoring systems (EMMS) should not only consider building physics and HVAC systems but also human behavior. These systems may provide information and advice to occupants about the relevance of their behavior regarding the current state of a dwelling and its connected grids. Therefore, advanced EMMS need to estimate the relevance of occupant activities. Additionally, innovative end-user services such as replay past situations, anticipate the future or mirror the current state are under development and require models together with building state estimations including the human part of the state. However, to define the state of a zone, non-measured values should be known in both physical (i.e. heat flows) and human part (i.e. occupancy and activities).The problem is to identify and calculate data processed from sensors, calendars, etc… that could be used in a classification model to estimate the number of occupants and various activities happening in offices/homes. The sensor data must provide a rich context for a classifier to have a broad separation plane and represent the office situation closely. Since the use of video cameras is a problem in many areas, the solution must respect privacy issues and relies largely on non-intrusive sensors.The thesis identifies the most relevant calculation from the sensor data in order to classify the number of people in a zone and their activities in offices/homes at a given time period. The proposed approach is inspired from machine learning and interactive learning to avoid using the camera and build a general estimation method.Three approaches are proposed for occupancy and activities estimation:- supervised learning approach. It starts to determine the common sensors that shall be used to estimate and classify the approximate number of people (within a range) in a room and their activities. Means to estimate occupancy include motion detection, power consumption, CO2 concentration sensors, microphone or door/window positions. It starts by determining the most useful measurements in calculating the information gains. Then, estimation algorithms are proposed: they rely on decision tree learning algorithms because it yields decision rules readable by humans, which correspond to nested if-then-else rules, where thresholds can be adjusted depending on the considered living areas. An office has been used for testing.- knowledge base approach using sensor data and knowledge coming respectively from observation and questionnaire. It relies on hidden Markov model and Bayesian network algorithms to model a human behavior with probabilistic cause-effect relations and states based on knowledge and questionnaire. Different applications have been studied for validation: an office, an apartment and a house.- an interactive learning approach is proposed. It estimates the number of occupants in a room by questioning occupants when relevant, meaning limiting the number of interactions and maximizing the information gains, about the actual occupancy. Occupancy and activities estimation algorithms use information collected from occupants together with common sensors. A real-time application has been done in an office case study.