Algorithm Function
Mathematical algorithms in inertial navigation systems resolve dynamic errors generated by the simultaneous presence of oscillatory linear acceleration and angular velocity. The process of sculling integration is used to calculate the net velocity change of a platform under high-frequency vibrations. This calculation prevents the accumulation of false velocity increments that occur when digital sampling fails to track the phase relationship between the two motions.
It is a critical component of the strapdown inertial navigation computations.
Dynamic Error
Oscillatory movement creates the error when the sensor platform undergoes angular oscillations about one axis and linear acceleration oscillations along an orthogonal axis. If these oscillations occur at the same frequency with a non-zero phase difference, a rectifying effect creates a false linear velocity. Software routines must run at high update rates to accurately capture these inputs and apply the necessary corrections.
Failure to perform sculling integration leads to a rapid, continuous drift in the estimated position. This drift is especially severe in high-vibration environments like aircraft or missile systems.
Numerical Processing
Modern inertial measurement units process raw sensor outputs using multi-state integration algorithms to perform the compensation. The algorithms utilize high-speed accelerometer and gyroscope data to calculate the cross-product terms that represent the sculling effect. These corrections are applied to the velocity state vector before it is used in the main navigation loop.
This architecture ensures that high-frequency vibration effects do not corrupt the low-frequency navigation solution.
Sensor Limit
The effectiveness of the correction is limited by the sampling frequency of the sensor processor. If the vibration frequency exceeds half the sampling rate, aliasing prevents the algorithm from resolving the phase relationship. The system must utilize high-frequency sampling to prevent aliasing.
