Application of Machine Learning in Diesel Engines Fault Identification Denys Pestana-Viana 1 , Ricardo H. R. Guti´ errez 3 Amaro A. de Lima 12 , Fabr´ ıcio L. e Silva 2 , Luiz Vaz 3 , Thiago de M. Prego 2 , and Ulisses A. Monteiro 3 1 Post graduation Program of Instrumentation and Applied Optics (PPGIO), Federal Center of Tech. Edu. Celso Suckow da Fonseca (CEFET-RJ) - Rio de Janeiro, Brazil denys.cefet@gmail.com, 2 Center for Research in Mechatronics (NUPEM), CEFET-RJ - Nova Iguau, Brazil {amaro.lima, fabricio.silva, thiago.prego}@cefet-rj.br 3 Federal University of Rio de Janeiro (UFRJ) - Rio de Janeiro, Brazil {rhramirez, vaz, ulisses}@oceanica.ufrj.br Abstract. The objective of this work is the fault diagnosis in diesel engines to assist the predictive maintenance, through the analysis of the variation of the pressure curves inside the cylinders and the torsional vibration response of the crankshaft. Hence a fault simulation model based on a zero-dimensional thermodynamic model was developed. The adopted feature vectors were chosen from the thermodynamic model and obtained from processing signals as pressure and temperature inside the cylinder, as well as, torsional vibration of the engines flywheel. These vectors are used as input of the machine learning technique in order to discriminate among several machine conditions, such as normal, pressure reduction in the intake manifold, compression ratio and amount of fuel injected reduction into the cylinders. The machine learning techniques for classification adopted in this work were the multilayer perceptron (MLP) and random forest (RF). Keywords: Machine learning, Fault identification, Vibration analysis 1 Introduction In the offshore industry, where the daily operation cost of units rises to exorbitant amounts, unexpected production outages can mean major economic losses. In addition, the unexpected failure of the equipment on board can produce accidents causing damage to the structure, putting at risk the crew, and possibly resulting in environmental impact. Diesel engines can be used in the offshore industry (support vessels and oil production units) in the main propulsion system, in electric power plants and in the mechanical drive of pumps and compressors. Therefore, the proper functioning of the engine components is critical to provide the torque and power for which they were designed. This dependence on diesel engines makes them have a high economic penalty when out of operation, especially when these stops are not programmed. The artificial neural networks as well as other supervised classifiers are ap- propriate for machinery applications as they can be trained offline and tested in