Verbesserung eines Frischluftmassenmodells mit Hilfe von Umgebungsgrößen an einem modernen Dieselmotor

Modern cars with diesel engines are usually equipped with a sensor to measure the actual air flow. This Air Flow Sensor is a very complex device and comes along with some drawbacks. It is difficult to manufacture and thus expensive. Additionally, the measuring tolerance of the sensor can increase over its lifetime. This master thesis shows a possible approach to generate an air flow signal base on opther sensor inputs. The system developed here is based on a model for the mass flow into the engine. In addition, the mass flow of the returned exhaust gas is estimated. Setting up the mass balance for the intake manifold leads to the desired fresh air flow before mixing with the exhaust gas. To cope with static errors, a variable lookup table is included in the system. At sationary engine operatig points, a reference air flow value is calculated from the Lambda-signal and the difference between modelled air flow and Lambda-air flow is adapted and stored in the lookup table. To calculate the amount of returned exhaust gas it is necessary to estimate the actual temperature and pressure in the exhaust manifold. For the temperature, this is done by a lookup table. For the pressure, a model of the exhaust manifold and the turbocharger is used. This system is based on physical models and offers an easy opportunity to apply the structure to different types of diesel engins.