High blood pressure (HBP) is at the root of many cardiovascular diseases, causing millions of deaths every year worldwide. Continuous monitoring of blood pressure (BP) is necessary, not only to prevent cardiovascular disease, but also to help those already affected. While many methods of blood pressure measurement exist at present, they present two major inconve-niences, these methods do not allow continuous measurement and they are frequently very invasive. In this paper, we present an alternative method to measure systolic and diastolic blood (SBP and DBP) pressure values with photoplethysmographic (PPG) signals using a deep learning architecture. PPG signals come from MIMIC II, a noisy and unprocessed database. Using a ResNet neural network coupled with LSTM layers, we succeeded in predicting SBP and DBP values, with an RMSE of 8.96 mmHg.

A Passive Fault Tolerant Control (PFTC) based on deep Reinforcement Learning (RL) approach has been proposed in this work. The RL agent is responsible for controlling a heating system in order to regulate the indoor temperature of an area. During operation, faults may occur in the heating system, particularly at the level of the internal resistance of the boiler and the pump responsible for the flow of water injected into the pipes to heat the area. The RL agent has been trained without the presence of a fault either in a holy environment and the goal is to see his behavior when the fault appears. The agent manages even without prior training to adapt and propose a PFTC to guarantee an interior temperature faithful to the set temperature while ensuring minimizing the energy consumed and good use of the heating system to guarantee a long equipment life.

Energy consumption in buildings has become a real concern for scientists and seeking to reduce this consumption is essential. Heating, ventilation, and air conditioning (HVAC) systems account for more than 50% of this consumption. One of the solutions to reduce this excessive consumption is to detect and diagnose faults that can appear instantaneously and quickly with fault diagnostic detection systems (FDD) based on artificial intelligence. The paper presents a strategy based on a data-driven approach for the detection and diagnosis of sensor faults that may appear in the Air Handling Unit (AHU) systems. A Decision Tree, Random Forest and SVM algorithm were used to detect and diagnose temperature sensor faults occurring in the AHU. The comparison between these methods shows that the Random Forest gives the best result with 96% accuracy.