Model-in-the-loop development for fuel cell vehicles -

Model-in-the-loop development for fuel cell vehicles

Increasing environmental concerns and awareness of the fact that fossil fuel will be depleted in the not long distant future have led to the rapid development of hybrid, fuel cell and battery electric vehicles in the past decade. Fuel cell technology has been considered the most promising alternative to substitute internal combustion engine for its zero emission characteristics and efficient use of renewable fuels.

In the area of fuel cell vehicle (FCV) development, Ford Motor Company has been a technology leader.Ford's Focus vehicle based technology demonstration FCV fleet has been operated by various partners around the globe for over five years and it has accumulated more than one million miles in total. The fuel cell vehicle considered in this paper is a type of hybrid electric vehicle, based on Ford Explorer.

In this paper, it will be referred to as new generation FCV (NGFCV). The NGFCV has a primary power source, the fuel cell (which is controlled by the fuel cell control unit (FCU)), a secondary power source, the battery (which is controlled by the battery control module (BCM)), and a high voltage energy converter (HVEC) which converts DC/DC current between the battery bus and vehicle bus. Propulsion is achieved through the IPT (Integrated Power Train, or the drive motor) connected to the wheels (front, rear or all wheel drive).

In this paper, the work on developing and validating a model-in-the-loop (MIL) simulation environment for a group of prototype fuel cell vehicles is presented. The MIL model consists of a vehicle plant model and an integrated vehicle system controller model.

First, the vehicle simulation plant model is functionally validated with a simple vehicle system controller (VSC) model and then improved to satisfy the input output interface and fidelity requirements. The developed MIL system is then verified for basic functionality against the simple VSC controller model and shows uniform correlation results. It is further validated against vehicle dynamometer test data and demonstrates satisfactory consistency.

A rapid model building approach which is suitable for model-based controller design process was also discussed. This approach enabled the developers to use model-to-code algorithms unlike many comparable simulation models.

To read this external content in full, download the complete paper from the author archives online.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.