Search

International audience; A general approach is proposed to determine the common sensors that shall be used to estimate and classify the approximate number of people (within a range) in a room. The range is dynamic and depends on the maximum occupancy met in a training data set for instance. Means to estimate occupancy include motion detection, power consumption, CO 2 concentration sensors, microphone or door/window positions. The proposed approach is inspired by machine learning. It starts by determining the most useful measurements in calculating information gains. Then, estimation algorithms are proposed: they rely on decision tree learning algorithms because these yield decision rules readable by humans, which correspond to nested if-then-else rules, where thresholds can be adjusted depending on the living areas considered. In addition, the decision tree depth is limited in order to simplify the analysis of the tree rules. Finally, an economic analysis is carried out to evaluate the cost and the most relevant sensor sets, with cost and accuracy comparison for the estimation of occupancy. C45 and random forest algorithms have been applied to an office setting, with average estimation error of 0.19-0.18. Over-fitting issues and best sensor sets are discussed.

WOS:000509819200044; International audience; Data quality assesment is a key component for many real applications, since it can drive better modelling. In this work a methodology to asses data quality (Qscore) is proposed and discussed. The validation of Qscore is performed via an interactive learning experiment related to occupancy estimation. Interactive learning has been shown to be crucial to consider and integrate occupant behavior in smart buildings. Indeed, valuable feedback and information can be collected from the occupants by involving them and by improving their consciousness about energy management systems. Users should feel involved to keep developing highly energy-efficient buildings. To reach this goal, occupants should be aware of the building features to feel more in control. This paper proposes a framework to interact with occupants to estimate building occupancy. This framework is based on an enhanced supervised learning approach that involves interaction with occupants, when necessary, to keep collecting training data. The training data consist of the measurements (i.e. features) collected from common sensors, for instance, motion detection, power consumption, and CO2 concentration, and the label (i.e. number of occupants) provided by the occupants during interactions. The considered learning machine in our experiments is the Multi-layer Perceptron regressor (MLP), although other approaches could be easily integrated within the proposed framework. In order to avoid useless interaction with users a new concept is introduced, called spread rate, to measure the quality of the data to decide if an interaction with the user is necessary or not. Extensive simulations have shown the merits of the proposed approach. (C) 2019 Elsevier B.V. All rights reserved.

Power management in homes and offices requires appliance usage prediction when the future user requests are not available. The randomness and uncertainties associated with an appliance usage make the prediction of appliance usage from energy consumption data a non-trivial task. A general model for prediction at the appliance level is still lacking. This work proposes to improve learning algorithms with expert knowledge and proposes a general model using a knowledge driven approach to forecast if a particular appliance will start during a given hour or not. The approach is both a knowledge driven and data driven one. The overall energy management for a house requires that the prediction is done for the next 24 h in the future. The proposed model is tested over the IRISE data and using different machine learning algorithms. The results for predicting the next hour consumption are presented, but the model works also for predicting the next 24 h.

http://www.sciencedirect.com/science/article/pii/S0378778812004987; International audience; Inhabitant's decisions and actions have a strong impact on the energy consumption and are an important factor in reducing energy consumption and in modelling future energy trends. Energy simulations that take into account inhabitants' behaviour are benchmarked at office buildings using controlled activity profiles and predefined scenarios. In this paper we have proposed a co-simulation environment for energy smart homes that takes into account inhabitants' dynamic and social behaviour. Based on this kind of complex behaviour, the setpoints for different controllers are adjusted in the physical simulator. In this platform, human behaviour is modelled using the Brahms environment and the thermal model and controllers for different appliances are modelled as a physical simulator. The thermal model computes the temperature decrease/increase in a room based on the contextual information resulting from the behaviour simulator. This information is then given to the controller to act upon.

This paper proposes a formulation of the global energy management problem of dwellings, which consists in a dynamic predictive control system able to generate optimized controls taking into account the model of the concerned dwellings, i.e. homes or offices. It focuses on the adjustment of the electric energy consumption and production in order to maximize energy usage efficiency, which is seen as a compromise between energy cost and overall comfort. To reach this objective, the concept of service is introduced: basically, a service participates to the comfort and may consume energy. The available flexibilities of the services provided by domestic appliances are used to compute anticipative optimal plans for appliance controls based on a mixed integer linear programming (MILP) algorithm. A reactive mechanism based on a list algorithm is added to face unforeseen events. The paper focuses on the computation of the anticipative plans. Different MILP models of common services are proposed. Application examples are given.