Pazychev

The traditional method of gyroscope biases determination is very difficult to be used during a calibration procedure. To calculate the errors the method must use a lot of navigation data saved during the calibration procedure. The equations of a strapdown inertial navigation system development were analyzed to show how the errors can be detected using the special control signals added to a tradition inertial navigation system algorithm. The control signals basing on the velocity errors are summarized to the navigation system angular velocity which is used by the direction cosines matrix calculation procedure. The method was tested during the calibration of SINS-TEK navigation system based on fibre optic gyroscopes(FOG) and quartz accelerometers. The method was only used for 60 seconds and the gyroscope biases were detected. After the biases being detected the output velocity errors calculating during 2 hours of work were improved 10 times.     

The inertial sensor signals caused by the temperature changing are very difficult to determinate. To observe the errors, this paper presents an error model of MEMS sensors. The error model consists of the arrays of the coefficients defined for each navigation system sensor. Processing sensor signals, I show how to calculate the parameters of the error model during the temperature changing. The method uses the least square method adapted by an increment of the temperature. This adaptation provides the calculation of the coefficients without influencing of the sensor temperature not changing during the calibration. Navigation system CompaNav2 based on ADXRS300 and ADXL210 Analog Devices sensors was calibrated using the method. The calibration results demonstrate that the method provides a practical and accurate technique for calibrating a strapdown inertial navigation systems based on MEMS sensors