Application of Neural Networks in the Estimation of Tire/Road Friction Using the Tire as Sensor

The importance of friction between tire and road for the dynamic behavior of road vehicles has been emphasized in many publications. Continuously updated knowledge of the friction potential and the friction demand can help to improve maneuverability and thereby safety of vehicles under slippery road conditions. An on line estimation method, based on combination of side force and self aligning torque, generated by the tire, is theoretically founded on a simple brush type tire model.

The system is implemented in the front wheel suspension of a passenger car. To cope with the highly non-linear behavior of the wheel suspension and the actual tire, various static neural networks have been applied in the estimation procedure. Experiments have been carried out both in simulation using a full vehicle multi-body model and with an actual vehicle. Conclusions are drawn regarding the estimation principle, the application of neural networks and the implementation in a test vehicle.