Practical considerations in precise calibration of a low-cost MEMS IMU for road-mapping applications

This paper addresses the theoretical and experimental development of a calibration scheme to overcome the intrinsic limitations of a low-cost Micro-Electrical-Mechanical System (MEMS) based Inertial Measurement Unit (IMU). The two-stage calibration algorithm was developed and tested successfully on a six-degree of freedom prototype MEMS IMU to determine the deterministic and stochastic errors of the sensor. This paper makes use of artificial observations known as pseudo-velocity measurements resulting from a specific scheme of rotation to calibrate the IMU in the laboratory environment. The proposed structure is then modified and utilised as a basis for the IMU's error estimation in outdoor navigation applications. For this purpose, the designed calibration method is applied to an integrated GPS/MEMS IMU system, showing improved navigational and road sign positioning performance in a test vehicle.