Computer model of embedded system for real-time estimation of gas emissions for stationary internal combustion engines

摘要

Concern for the deterioration of the atmosphere, and exhaust enission legislation have affected engine installations in the petroleum industry by placing restrictions on the emissions levels of carbon monoxide (CO), nitrogen oxides (NO_x), and hydrocarbons (HC) from spark-ignited stationary internal combustion (SISIC) engines. This paper describes a nvel and universal design of a parametric model, which is capable of calculating eissions produced by natural gas SISIC engines without expensive gas sensors. For the model we used control parameters, monitoring parameters, and claculated parameters. Using data from a tuned engine, 4 multidimensional arrays were created containing emission levels as functions of selected operating parameters. Interpolation methods were used to fill these arrays, thus increasing model resolution. Recorded data from non-increasing model resolution. Recorded data from nontuned engines was used to extend the model. The approach can dynamically estimate harmful emissions from SISIC engines and offers the potential to eliminate the need for site conpliance monitoring. Investigation into the use of Neural Networks to reduce the required computational resources is underway and progress is reorted.