Real-time accurate odometer velocity estimation aided by accelerometers

In strap-down gyro-compass in-motion alignment, the alignment accuracy depends not only on the quality of the gyroscopes and accelerometers, but also on the accuracy of the velocity provided by the aiding sensors such as odometers. To improve the accuracy of the in-motion alignment, real-time accurate odometer velocity estimation is required. In this paper, the effect of the noise of the odometer velocity on strap-down gyro-compass in-motion alignment accuracy is presented, based on the strap-down gyro-compass algorithm. A velocity tracking model is designed as the state model, to describe the relationship between and among the vehicle's velocity, acceleration and jerk in the vehicle frame. Based on the velocity equation applied to strap-down navigation system mechanizations, the vehicle's acceleration in the vehicle frame can be obtained from the specific forces measured by accelerometers. With the observations including the vehicle's acceleration in the vehicle frame and the vehicle's velocity in the vehicle frame obtained from the odometer using the first order difference algorithm, real-time velocity estimates are produced by a Kalman filter. The field test results show that the proposed method can successfully improve the accuracy of the odometer velocity. The comparison with the traditional method highlights the superior performance of the proposed method. (C) 2016 Elsevier Ltd. All rights reserved.