Predictive Maintenance presents an important and challenging task in Industry 4.0. It aims to prevent premature failures and reduce costs by avoiding unnecessary maintenance tasks. This involves estimating the Remaining Useful Life (RUL), which provides critical information for decision makers and planners of future maintenance activities. However, RUL prediction is not simple due to the imperfections in monitoring data, making effective Predictive Maintenance challenging. To address this issue, this article proposes an Evidential Deep Learning (EDL) based method to predict the RUL and to quantify both data uncertainties and prediction model uncertainties. An experimental analysis conducted on the C-MAPSS dataset of aero-engine degradation affirms that EDL based method outperforms alternative machine learning approaches. Moreover, the accompanying uncertainty quantification analysis demonstrates sound methodology and reliable results.

Modeling and control of the heating feature of living spaces remain challenging tasks because of the intrinsic nonlinear nature of the involved processes as well as the strong nonlinearity of the entailed dynamic parameters in those processes. Although nowadays, adaptive heating controllers represent a crucial need for smart building energy management systems (SBEMS) as well as an appealing perspective for their effectiveness in optimizing energy efficiency, unfortunately, the leakage of models competent in handling the complexity of real living spaces’ heating processes means the control strategies implemented in most SBEMSs are still conventional. Within this context and by considering that the living space’s occupation rate (i.e., by users or residents) may affect the model and the issued heating control strategy of the concerned living space, we have investigated the design and implementation of a data-driven machine learning-based identification of the building’s living space dynamic heating conduct, taking into account the occupancy (by the residents) of the heated space. In fact, the proposed modeling strategy takes advantage, on the one hand, of the forecasting capacity of the time-series of the nonlinear autoregressive exogenous (NARX) model, and on the other hand, from the multi-layer perceptron’s (MLP) learning and generalization skills. The proposed approach has been implemented and applied for modeling the dynamic heating conduct of a real five-floor building’s living spaces located at Senart Campus of University Paris-Est Créteil (UPEC), taking into account their occupancy (by users of this public building). The obtained results assessing the accuracy and addictiveness of the investigated hybrid machine learning-based approach are reported and discussed.

Most already advanced developed heating control systems remain either in a prototype state (because of their relatively complex implementation requirements) or require very specific technologies not implementable in most existing buildings. On the other hand, the above-mentioned analysis has also pointed out that most smart building energy management systems deploy quite very basic heating control strategies limited to quite simplistic predesigned use-case scenarios. In the present paper, we propose a heating control strategy taking advantage of the overall identification of the living space by taking advantage of the consideration of the living space users’ presence as additional thermal sources. To handle this, an adaptive controller for the operation of heating transmitters on the basis of soft computing techniques by taking into account the diverse range of occupants in the heating chain is introduced. The strategy of the controller is constructed on a basis of the modeling heating dynamics of living spaces by considering occupants as an additional heating source. The proposed approach for modeling the heating dynamics of living spaces is on the basis of time series prediction by a multilayer perceptron neural network, and the controlling strategy regarding the heating controller takes advantage of a Fuzzy Inference System with the Takagi-Sugeno model. The proposed approach has been implemented for facing the dynamic heating conduct of a real five-floor building’s living spaces located at Senart Campus of University Paris-Est Créteil, taking into account the occupants of spaces in the control chain. The obtained results assessing the efficiency and adaptive functionality of the investigated fuzzy controller designed model-based approach are reported and discussed.

The forecasting of building energy consumption remains a challenging task because of the intricate management of the relevant parameters that can influence the performance of models. Due to the powerful capability of artificial intelligence (AI) in forecasting problems, it is deemed to be highly effective in this domain. However, achieving accurate predictions requires the extraction of meaningful historical knowledge from various features. Given that the exogenous data may affect the energy consumption forecasting model’s accuracy, we propose an approach to study the importance of data and selecting optimum time lags to obtain a high-performance machine learning-based model, while reducing its complexity. Regarding energy consumption forecasting, multilayer perceptron-based nonlinear autoregressive with exogenous inputs (NARX), long short-term memory (LSTM), gated recurrent unit (GRU), decision tree, and XGboost models are utilized. The best model performance is achieved by LSTM and GRU with a root mean square error of 0.23. An analysis by the Diebold–Mariano method is also presented, to compare the prediction accuracy of the models. In order to measure the association of feature data on modeling, the “model reliance” method is implemented. The proposed approach shows promising results to obtain a well-performing model. The obtained results are qualitatively reported and discussed.

The HVAC system accounted for a significant portion of the building's energy consumption, resulting in enormous CO 2 emissions. Among the numerous HVAC control methods, reinforcement learning (RL) gives the ability to control complex systems without requiring an explicit model of the building's thermal dynamics. This study conducted a concise review of previous research on the application of RL to HVAC systems in buildings, it offered a thorough explanation of the theoretical foundations of RL and a summary of several recent RL studies that employ a particular variant of each main component of the RL: environment, state-space, action-space, rewards function, number of time steps and training episodes. Most studies construct the training environment as a stationary MDP due to the use of a predefined single sequence of transitions for non-action-controllable state vector components (e.g., outdoor temperature and occupancy schedule). This type of MDP is solved using tabular RL and DRL algorithms. Future research should focus on using the Meta-RL approach for HVAC systems, which solves the problem of non-stationarity in the environment (non-stationarity-MDP) where the sequence of transitions for (outdoor temperature and occupancy schedule) are changeable, as is the case during the real-world implementation of the RL controller in buildings.