基于自回归模型的汽车姿态角在线预测系统

针对汽车姿态角在线预测问题，利用微惯性测量单元以捷联方式测量汽车6自由度运动参数，通过四元数法解算汽车的姿态角，应用卡尔曼滤波器对姿态角进行最优估计处理，在此基础上建立自回归模型对未来1 s~3 s的姿态角进行实时预测，使用Visual Basic 2005、NI Measurement Studio开发汽车姿态角在线预测软件，并对其进行实车道路试验，结果证明该软件具有较高的预测准确性。

On-line Prediction System of Vehicle Attitude Angle Based on Auto-regressive Model

Aiming at the on-line prediction problem of vehicle attitude angles,this paper uses 6 degree of freedom vehicle motion parameters to measure directly by Micro Inertial Measurement Unit（MIMU） in strap-down way,then quaternion is adopted for attitude algorithm and Kalman filter is used to get the optimal estimation of attitude angles.Auto-regressive（AR） model is set up for future 1 s3 s prediction of attitude angle.In addition,on-line prediction software for vehicle attitude angles is developed based on Visual Basic 2005 and NI measurement studio,and the road test is carried out.The result shows the validity and feasibility of on-line prediction software for vehicle attitude angles