MACHINE LEARNING FOR MISFIRE DETECTION IN A DYNAMIC FIRING LEVEL MODULATION CONTROLLED ENGINE OF A VEHICLE

Using machine learning for misfire detection in a Dynamic firing level modulation controlled internal combustion engine is described. A neural network is used to calculate expected crank acceleration from various inputs, including the dynamically defined cylinder skip fire sequence. The output of the neural network is then compared to a signal indicative of the measured crank acceleration. Based the comparison, a prediction is made if a misfire has occurred or not. In alternative embodiment, the neural network is expanded to include the measured crank acceleration as an additional input. With the latter embodiment, the neural network is arranged to directly predict misfire events.