UKF-based adaptive variable structure observer for vehicle sideslip with dynamic correction

摘要

Knowing vehicle sideslip angle is important and useful in various vehicle active safety applications. However, equipment which could directly measure sideslip angle is usually too expensive for industrial applications, so a reliable, affordable sideslip angle observer is necessary. Model-based observer is affected by the model uncertainty, non-linearity and the road friction inaccuracy. Sensor-based observer might be easily affected by the sensor noise and the bias induced by the road micro-uneven. Thus, an unscented Kalman filter (UKF)-based adaptive variable structural observer with dynamic correction (AUKF) for vehicle body sideslip angle is brought forward in this study. First, an UKF observer with adaptive parameters is used to compensate the model uncertainty. Then, a sensor-based observer is applied to realise a variable structure observer to compensate the estimation error in the highly non-linear region. A pseudointegral method and a zero point reset method are adopted to dynamically correct the bias drift problem in the integral. Both simulations and real vehicle tests validated the proposed approach. The accurate vehicle sideslip angle is measured by a differential global positioning system. Results show that the proposed approach has better performance compared with traditional UKF method, and makes a good foundation for stability control.