Engine Misfire Detection For Green Combustion Automobiles: Robust Algorithms For Engine Misfire Detection

Current regulations in North America and Europe require engine manufacturers to develop an on-board real time misfire detection system to monitor emissions related systems in automobiles. These regulations create a challenge for the engine manufacturers to develop a low cost on-board real time system. In this book different techniques are presented to design such a system. The overall algorithm consists of two main steps. First, it employs a frequency domain pattern classifier to detect and identify the type of misfire pattern, and then it utilizes two different techniques to locate the misfire instances. The first technique for misfire instances detection is based on a weighted Wiener adaptive filtering technique, while the second technique is based on time-frequency. The overall algorithm is intended to detect the engine misfire pattern type, identify engine misfire instances, the misfiring cylinder(s), and detect the total number of misfire events resulting from each cylinder under various speed and load conditions and misfire patterns. The performance of the presented algorithms is superior to other algorithms found in the literature.