Abstract
This paper describes an online (real-time) optimization for control target values of NOx emission in diesel engine. Diesel engines are required to realize high efficiency and low fuel consumption while achieving exhaust emissions regulations which will become even more severe in the future. For that purpose, it is necessary to manage together the NOx reduction rate on the engine side and the aftertreatment system side. The target value manager performs online optimization based on the actual operating state of the engine and aftertreatment system. Online optimization uses a gradient method. By this method, the control target value that minimizes the operation cost (fuel and urea water) is calculated while suppressing the NOx emission to the regulated value or less. In addition, multi-objective particle swarm optimization is used to calibrate Lagrange multiplier parameters that affect NOx emissions and total operation cost, and Pareto optimal solutions for the two evaluation functions are obtained. This optimization improves the performance of the target value manager and visualizes tradeoffs between NOx emissions and total operating costs. The effectiveness of this approach is evaluated by simulation using data measured with a diesel engine test bench.
