IMU Signal Software Simulator

摘要

This paper describes the implementation and experimental results of a software based inertial measurement unit (IMU) signal simulator. The simulated signal generation of inertial sensors is an inverse process of the IMU mechanization (navigation) equations. Compared with using hardware, the IMU simulator (IMUS) saves significant time and money for research work and is very flexible when developing new integration algorithms as it does not impose experimental limitations. Furthermore, the IMUS can be used efficiently when choosing or designing the required hardware characteristics for a given application. It is often difficult to accurately translate IMU errors into state errors (position, velocity, and attitude) which the simulator does with ease. rnThe IMU simulator was developed by the MMSS group, the University of Calgary. It can simulate a variety of sensor errors such as bias instability, random walk, scale factor errors, sensor errors due to thermal drift, gsensitivity, non-orthogonalities, misalignments, and their combinations. The user can also define the output data rate, bandwidth, low pass filter cutoff and so on, based on a variety of vehicle dynamics such as straight line, accelerations, turns, U-turns, bumpy roads, constant velocities, static periods as well as varying attitude, and their combinations for different applications. Meanwhile, the simulator provides GPS position and velocity simulated measurements as an optional function. This tool gives users a fast and effective way for evaluating new inertial sensors using datasheet characteristics provided by the manufacturers or obtained through lab testing. Or, if the IMU characteristics are not known, they can be inferred by generating state errors for given IMU errors, in this way developing the required IMU parameters. rnThe correctness and effectiveness of the IMUS has been verified not only in theory but also in practice by comparing the results from the IMUS to the results from a real hardware IMU using field test data. A real case example for both a simulated and a real hardware MEMS ADI IMU on the IMU/GPS integration level is presented in the paper. It shows that the average position drift during GPS outages using simulator data corresponds well to field test results using the ADI hardware IMU and GPS from a NovAtel OEM4 receiver.