A Stochastic Model of the Fuel Injection of the SI Engine

The paper is devoted to the issues of the non-stationary, stochastic analysis of the fuel injection. The aim of the experiment was to obtain data about SI engine operating parameters during operation. The object randomness in sequentially injected combustion engines are investigated. A numerical procedure is designed for the intake manifold mixture preparation to compensate the mixture variations.

There are some direct consequences of inherent non-linearities in the basic dynamics of mixture preparation. Compensating for the mixture dynamics (fuel path, air path consisting of intake manifold and wall-wetting dynamics and the oxygen sensor dynamics) performs well only if the statistical properties of the stochastic processes are sufficiently matched. Since this a priori knowledge is hardly accessible in the present case, the authors implemented an experimental procedure. The results of the tests confirm the great potential in the future application of adaptive techniques in the field of mixture ratio control.

The issues of the non-stationary, stochastic analysis of the engine operation have been formulated. The paper presents the results of engine research in city drive and road run tests. The experiment was performed using the Engine Monitoring System developed by the authors. The test vehicle was equipped with a 2,2 L Multi-Point Fuel Injection engine.