How to solve the accumulated errors of inertial navigation products over time?

The error of inertial navigation mainly comes from the inherent defects of inertial sensors, which increase in a quadratic (position) or linear (attitude) manner over time.

The main reasons for the accumulation of errors over time are: gyro drift, accelerometer bias; the integration of accelerometer bias error can lead to velocity error, and the secondary integration can cause position drift (for example, a 1mg bias for 1 hour can result in a position error of about 60 meters); Temperature sensitivity; Installation error; The sensor coordinate system is not strictly aligned with the carrier coordinate system.

Common methods for addressing error accumulation include:

• Multi-sensor integrated navigation: (1) GNSS/INS combination: periodically correcting the position and velocity errors of inertial navigation through GPS/Beidou and other satellite signals, suppressing long-term drift; (2) Milometer/Doppler radar assistance: vehicle odometer or radar speed information can constrain the speed drift of inertial navigation.

• Advanced filtering algorithm (Kalman filtering algorithm) real-time estimates and compensates for sensor errors (such as zero bias, scale factor).

• Suppress temperature drift through built-in temperature sensors and compensation algorithms.

• Reduce initial attitude error through static calibration during startup; And methods such as forcing the speed to zero and resetting the speed integration error when the carrier is stationary (such as when the vehicle is parked).