GPS and Low Cost INS Integration for Positioning in the Urban Environment

摘要

GPS and INS integrated systems are expected to become widely available as a result of low cost Micro-ElectroMechanical Sensor (MEMS) technology. However, the current performance achieved by low cost sensors is still relatively poor due to the large inertial sensor errors. This is particularly prevalent in the urban environment where there are significant periods of restricted sky view. Currently GPS and low cost INS are commonly integrated using a loosely coupled architecture which is suitable in most applications where there is good satellite availability. However, significant performance improvements can be made in the urban environment by integrating GPS at the measurement level using the tightly coupled approach. Furthermore, when it is not critical to obtain the position solution in real-time, a Kalman filter smoothing algorithm can be applied to post process the data. Kalman filter smoothing is rarely considered in the case of low cost INS, however it can result in significant performance improvements, particularly for low cost INS integration. This paper investigates the performance of a low cost Crossbow IMU integrated with GPS measurements in the urban environment. The tightly coupled integration filter is combined with a Kalman filter smoothing algorithm to obtain a high quality integrated solution. The results obtained in this paper demonstrate that a GPS and low cost INS system is capable of meeting the performance requirements for a number of applications where GPS availability is restricted. Furthermore, by integrating GPS and INS data using a Kalman filter smoother, it is demonstrated that in post processing, high performance positioning can be achieved for a range of applications where low cost and high quality, continuous positioning is required.