Search Results

Beginning in 2010, implementation of on-board diagnostics (OBD) is mandatory for all the heavy-duty engine applications in the United States. The task of developing OBD strategies and calibrating them is a challenging one. The process involves a strong interdependency on base engine emissions, controls and regulations. On top of that the strategies developed as a result of the regulatory requirements need to go through a stringent and time-intensive process of software implementation and integration. The recent increasing demands to minimize the development process have been pushing the envelope on the methodologies used in developing the strategies and the calibration for robust monitoring. The goal of this paper is to provide a concise overview of a process utilized to help the development, testing and calibration of the OBD strategies on a 2010 model year heavy-duty diesel engine.

With the advent of advanced diesel after-treatment technologies, sophisticated sensors are becoming a critical cost challenge to OEMs. This paper describes an approach for replacing the engine out NOx sensor with an artificial neural network (ANN) based virtual sensor. The technique centers around inferring NOx concentration from readily available engine operating parameters, eliminating the need for physical sensing and the cost associated with it. A multi-layer perceptron network was trained to estimate NOx concentration from engine speed, load, exhaust gas recirculation, and air-fuel ratio information. This supervised learning was conducted with measured engine data. The network was validated against measured data that was excluded from the training data set. The paper details application of this technique to both a heavy duty and light duty diesel engine. Results show good agreement between predictions and measured data under the steady state conditions studied.