摘要

As the active vehicle warning technology develops with built-in vehicle-mounted platforms as a carrier, this paper studied and designed a vehicle rollover warning system through information fusion, to cut traffic accident that result from vehicle rollover. The improved Sage-Husa adaptive Kalman filtering algorithm was employed to fuse the automobile real-time angular velocity and lateral acceleration to get real-time roll angle, of which the range was monitored in real time by the vehicle-mounted platform for deciding to deliver a warning to the driver or not. When the roll angle exceeded the threshold value, the warning would be triggered. MATLAB/Simulink was used for simulating proof of the algorithm. The result showed that the algorithm was of high stability and tracking performance, and could effectively enhance the detection accuracy of the automobile rollover warning system.