Modelling and filtering of wheel speed signals to determine driveline behavior in an all-wheel drive vehicle

Abstract: In vehicles there are many issues involving the safety and performance. The safety issues are often regulated by governments, whereas the performance by the customer. One solution to both these issues is to take use of an all-wheel drive system, where a coupling engages and disengages driving axles when needed. Such a solution will, through added traction and stability, increase performance and safety when needed and optimize fuel efficiency when not needed, due to driving on only two wheels. An essential factor to optimize the performance is to know the temperature of the coupling at any given time. The simplest way to solve this problem is to use a temperature sensor, but due to the ever optimization of the vehicle, such features ought to be removed in an effort to drive down cost and weight. Instead the temperature can be determined using a model of the vehicle’s driveline, which estimates the speed difference over the coupling. This can in turn decide how much friction is exerted on the coupling and from there the temperature can be decide to a high degree of exactness, given that the speed difference estimation is exact. The aim of this thesis has been to investigate a novel method of modelling the driveline, in Matlab/Simulink, to get a more exact estimated speed difference over the coupling in certain driving scenarios where the existing method tends to fail. Using sensor measurement of the speed difference on a real world vehicle, the produced model could be tuned and validated to mimic the actual speed difference behavior in the driving scenarios. The final temperature estimation from the speed difference was not in scope of the project. The work has been performed at BorgWarner Sweden AB as a Master Thesis Project at Lund University.