Search Results

With the implementation of a closed loop fuel control system, operation of lean-burn natural gas engines can be optimized in terms of reducing emissions while maximizing efficiency. Such a system would compensate for variations in fuel composition, but also would correct for variations in volumetric efficiency due to immediate engine history and long-term engine component wear. Present day engine controllers perform well when they are operated with the same gas composition for which they were calibrated, but because fuel composition varies geographically as well as seasonally, some method of compensation is required. A closed loop control system on a medium-duty lean-burn engine will enhance performance by maintaining the desired air-fuel ratio to eliminate any unwanted rich or lean excursions (relative to the desired air-fuel ratio) that produce excess engine-out emissions. Such a system can also guard against internal engine damage due to overheating and/or engine knock.

This paper explores the feasibility of using in-cylinder pressure-based variables to predict gaseous exhaust emissions levels from a Navistar T444 direct injection diesel engine through the use of neural networks. The networks were trained using in-cylinder pressure derived variables generated at steady state conditions over a wide speed and load test matrix. The networks were then validated on previously “unseen” real-time data obtained from the Federal Test Procedure cycle through the use of a high speed digital signal processor data acquisition system. Once fully trained, the DSP-based system developed in this work allows the real-time prediction of NOX and CO2 emissions from this engine on a cycle-by-cycle basis without requiring emissions measurement.