The objective of this work is to improve the combustion management in W18V50DF dual fuel engine by determining for a desired power, the optimal values of the parameters of pressures and subsequently to map them in real time based on a power set point. The interest was mainly focused in pressure parameters, other being considered as constant. Two methods have been used, namely mathematical modeling and learning by neural networks. The results show that, in the beginning mathematical modeling result helps to monitor the ongoing process and with longer learning period the result with neural network become better and significant due to the adaptation to the reality. Furthermore, the neural network method improves significantly in the long term the rationalization of fuel consumption in such a system in order to significantly reduce the carbon dioxide emission rate. Finally, work has proved that for an immediate result mathematical model can be used but without robustness on the control process, this is obtained by a neural network. But this approach requires a good data base and long learning time.

Throughout the world, thousands of passengers travel by air, their quality depends on that of the equipment used. Predictive maintenance is increasingly used to estimate. The remaining useful life of system components and in particular turbofan engines as an essential component. It is used to predict failure before it occurs, optimize component design, extend equipment life, and reduce maintenance costs. However, the algorithms proposed in the literature to date to determine the remaining useful life lack precision with a quadratic error around 20 while the physical models have errors of the order of 0.02. The problem here is how to increase the accuracy of predicting the remaining useful life of a turbofan engine. The objective of this study is to develop a more realistic and accurate algorithm for calculating the remaining useful life of a turbofan engine. To do this, we considered the degradation of the high pressure compressor and the fan as essential organs of the turbojet engine and we used deep learning, known for its high precision linked to a great capacity for extracting information. More specifically, it involved acquiring data on a turbojet engine in operation, pre-processing this data, developing the prediction model, training the model and finally validating the approach in comparison with other diagnostic methods. and to model these defects. We compared two deep learning architectures per application against the CMAPSS dataset to assess their performance. The LSTM architecture we developed prevailed with an RMSE of 13.76, well positioned compared to the literature architecture.

This paper presents a functional and dysfunctional behavioral study of a telecommunication system, with the aim to evaluate the performance of its constituent units. It is question of taking advantage offered by artificial intelligence in order to evaluate by modeling and simulation in system reliability. The methodological approach consists in combining ANFIS neuro-fuzzy networks with hybrid stochastic automata. The Neuro-Fuzzy ANFIS networks provide a prediction for the passage from nominal mode to degraded mode, by controlling the occurrence of malfunctions at transient levels. This allows to anticipate the occurrence of events degrading system performance, such as failures and disturbances. The objective is to maintain the system in nominal operating mode and prevent its tipping in degraded mode. The results are implanted around a demonstrator based on Scilab, and implemented on Matlab / Simulink.