One of the gray areas in the implementation of regulations limiting the generation of pollutants from mobile sources is the actual effectiveness of the exhaust gas emissions control strategy in vehicles that have been in use for some time. While it is possible today to conduct limited diagnostics with the on-board engine computer by performing periodic checks to verify the validity of the signals measured by the on-board sensors, and to measure tailpipe emissions during routine inspection and maintenance, the task of correlating these measurements with each other to provide an on-line, accurate diagnosis of critical malfunctions has thus far proven to be a very challenging task, especially in the case of misfire.
A typical automotive emission control system relies on the measurements provided by various sensors, including for example air charge temperature, intake manifold pressure, exhaust gas oxygen concentration, and engine speed, to adjust the air and fuel mixture as closely as possible to the stoichiometric air-to-fuel ratio (AFR). Maintaining the AFR close to stoichiometry allows the catalytic converter to operate in a high conversion efficiency region, and to significantly reduce the emission of the three major pollutants (for gasoline engines), that is, CO, HC, and NOx. Improper operation of any of the sensors, as well as actual component malfunctions (e.g., a fouled spark plug or fuel injector, or a vacuum or compression leak), may sufficiently upset the delicate balance sought by the electronic controller to cause the mixture to significantly deviate from the desired set point. Such a condition results not only in increased emissions, but possibly in further damage to components (e.g., the catalytic converter), with significant cost to the consumer or to the manufacturer.
It is clearly important to diagnose malfunctions in the emission control system at the incipient stage. Since in practice the first visible “symptom” of a problem with the emission control system may be the measurement of excessive tailpipe emissions during an inspection and maintenance test, it is very important that the exhaust gas measurements be somehow correlated with the health of the emission control system. This is a particularly challenging problem, because the chemistry of exhaust gas production is very complex, and not easily modeled by conventional dynamic system modeling methods. One approach to understanding the correlation between emission control system malfunctions and measured tailpipe emissions is to generate an extensive data base, and to analyze the data to uncover characteristic patterns that may assist in the diagnosis of specific faults. The ultimate aim of the project is to provide a methodology for the development of diagnostic tools that can integrate the information supplied by conventional tailpipe inspection programs with on-board diagnostics to provide fast and reliable diagnosis of misfire.