In order to reduce the fuel consumption of a vehicle while keeping the performance the same, there
are several possible options, from which drivetrain hybridization is one of them. Hybrid vehicles make
use of 2 or more energy sources to propel the vehicle. For optimal collaboration of all components in
terms of fuel consumption or emissions, a power management strategy can be introduced. In previous research, a model of the vehicle is made and tested in Matlab/Simulink model to compare dierent control strategies. This model is based on a given driving cycle and is processed o-line with previously obtained experimental data. The vehicle in question, might be deployed oshore and there are no driving cycles available which represents such an environment. This report describe an architecture to create a methodology which can include fuzzy modelling of the complex submodels of the vehicle. This approach must allow the adaption of the fuzzy model to unknown environments and capture dynamic behaviour as well. This is due to a variable time-window approach thus resulting on a model which is regularly updated.