惯性导航系统的误差估计

惯性导航系统（INS）以其自主的工作能力广泛应用于军事武备的导航、制导与控制系统和国民经济的诸多领域。它的主要缺点是定位误差随其工作时间的增长而增大。对惯导系统的误差进行估计和补偿是在保证性能价格比的前提下，提高惯性导航系统精度的有效途径。目前，对惯导系统的误差修正均采用外信息（如GPS的输出信息）校正，即在INS工作的全部时间内，定期地利用GPS输出的速度和位置信息与INS输出的相应信息的差值作为观测量，对INS误差进行估计和补偿。Kalman滤波的方法广泛地应用于惯导系统的误差修正初始对准。本研究了当地水平惯导系统的误差估计和补偿问题。分析结果表明，采用Kalman滤波的方法，可以精确地估计惯导系统的误差（包括陀螺漂移和加速度计零偏），误差估计的精度高，并且估计的方差阵收敛快。

Estimation of the INS's Errors

Inertial navigation system (INS) is used widely in many military and civil scopes because of its ability of independent operation. The primary disadvantage of INS is that the position error increases with the time. Error estimation and compensation is an effective approach to improve the accuracy of INS under the same ratio of performance to price. At present, the outside information (for instance, the outputs of GPS ) is used to correct the errors of INS, i.e. the errors of INS are estimated and compensated regularly, based on taking the difference between the outputs of GPS and INS as the observation during the whole time of the INS operating. The Kalman filter is applied widely in error correction and the alignment of INS. In this paper, the research of error estimation and compensation has been made for INS of the local-level frame. The analysis results show that the errors of INS (the gyro drifts and accelerometer biases are included) can be estimated accurately by Kalman filter. The referenced velocity can be derived from either GPS or Doppler log, and the referenced heading from compass. The estimation precision is high and the relevant variances can rapidly convergence.