TOLERANCE CALCULATION SCHEME Overview

A tolerance calculation scheme for AMC has been set up. It is fed by a Monte Carlo algorithm (4) which provides a large variety of operation point parameter sets. Monte-Carlo simulation describes a stochastic procedure that is based on the „Law of Large Numbers". In the context of the tolerance calculation scheme described here it is used to generate a large number of random operation conditions. Four sub algorithms are involved in the calculation as shown in Fig. 4: Controller core with
pre-control, fuel injection demand evaluation and a Design-of-Experiments (DoE) based pump model. The basic idea is to generate a large set of learning information

I. e. AMC values by using an emulation of the ECU software and a mathematical model which represents the physical behaviour of the high pressure pump. Tolerances of the participating components are considered in the ECU algorithm as well as for the pump model. Tool operation requires additional input such as calibration data like a characteristic pump curve or system demand values and tolerance data e. g. of the rail pressure sensor, pressure control valve, high pressure pump and injection quantity accuracy functions. Functional tolerances can be made visible by drawing learning value ranges over characteristic input parameters.

TOLERANCE CALCULATION SCHEME Overview

Pre control

Model

Quantity

Demand

TOLERANCE CALCULATION SCHEME Overview

PID governor

подпись: pid governor

DoE based pump model

подпись: doe based pump model^ O = Engine operation mode

Fig. 4: Calculation scheme for AMC learning values

Добавить комментарий

Ваш e-mail не будет опубликован. Обязательные поля помечены *