Full vehicle and tyre identification using unscented and extended identifying Kalman filters

This paper considers identification of all significant vehicle handling and driveline dynamics of a test vehicle, including identification of a combined-slip tyre model, using only those sensors currently available on most vehicle CAN buses. The method extends previous work using augmented Kalman Filter state estimators to concentrate wholly on parameter identification, and it compares Extended and Unscented Kalman filter algorithms. Using an appropriately simple but efficient model structure, all of the independent parameters are found from test vehicle data, with the resulting model accuracy demonstrated on independent validation data. The method is suited to applications of system identification, but also in on-line model predictive controllers or estimators. It can also operate in real-time, so the model could be continuously identified to maintain accuracy with each new journey.