9.4.2 Initialization and Alignment

9.4.2.1 Navigation Initialization INS initialization is the process of determining initial values for system position, velocity, and attitude in navigation coordinates. INS position initialization ordinarily relies on external sources such as GNSS or manual entry by crew members. INS velocity initialization can be accomplished by starting when it is zero (i.e., the host vehicle is not moving) or (for vehicles carried in or on other vehicles) by reference to the carrier velocity. (See alignment method 3 below.) INS attitude initialization is called alignment.

9.4.2.2 Sensor Alignment INS alignment is the process of aligning the stable platform axes parallel to navigation coordinates (for gimbaled systems) or that of determining the initial values of the coordinate transformation from sensor coordinates to navigation coordinates (for strapdown systems).

Alignment Methods Four basic methods for INS alignment are as follows:

1. Optical alignment, using either of the following:
   (a) Optical line-of-sight reference to a ground-based direction (e.g., using a ground-based theodolite and a mirror on the platform). Some space boosters have used this type of optical alignment, which is much faster and more accurate than gyrocompass alignment. Because it requires a stable platform for mounting the mirror, it is applicable only to gimbaled systems.
   (b) An onboard star tracker, used primarily for alignment of gimbaled or strapdown systems in space or near-space (e.g., above all clouds).
2. Gyrocompass alignment of stationary vehicles, using the sensed direction of acceleration to determine the local vertical and the sensed direction of rotation to determine north. Latitude can be determined by the angle between the earth rotation vector and the horizontal, but longitude must be determined by other means and entered manually or electronically. This method is inexpensive but the most time-consuming (several minutes, typically).
3. Transfer alignment in a moving host vehicle, using velocity matching with an aligned and operating INS. This method is typically several times faster than gyrocompass alignment, but it requires another INS on the host vehicle and it may require special maneuvering of the host vehicle to attain observability of the alignment variables. It is commonly used for in-air INS alignment for missiles launched from aircraft and for on-deck INS alignment for aircraft launched from carriers. Alignment of carrier-launched aircraft may also use the direction of the velocity impulse imparted by the steam catapult.
4. GNSS-aided alignment using position matching with GNSS to estimate the alignment variables. It is an integral part of integrated GNSS/INS implementations. It does not require the host vehicle to remain stationary during alignment, but there will be some period of time after turnon (a few minutes, typically) before system navigation errors settle to acceptable levels.

Gyrocompass alignment is the only one of these that requires no external aiding. Gyrocompass alignment is not necessary for integrated GNSS/INS, although many INSs may already be configured for it.

INS Gyrocompass Alignment Accuracy A rough rule of thumb for gyrocompass alignment accuracy is

where σ_gyrocompass is the minimum achievable RMS alignment error in radians, σ_acc is the RMS accelerometer accuracy in g values, σ_gyro is the RMS gyroscope accuracy in degrees per hour, 15° per hour is the rotation rate of the earth, and φ_geodetic is the latitude at which gyrocompassing is performed.

Alignment accuracy is also a function of the time allotted for it, and the time required to achieve a specified accuracy is generally a function of sensor error magnitudes (including noise) and the degree to which the vehicle remains stationary.

Gimbaled INS Gyrocompass Alignment Gyrocompass alignment for gimbaled systems is a process for aligning the inertial platform axes with the navigation coordinates using only the sensor outputs while the host vehicle is essentially stationary. For systems using ENU navigation coordinates, for example, the platform can be tilted until two of its accelerometer inputs are zero, at which time both input axes will be horizontal. In this locally leveled orientation, the sensed rotation axis will be in the north-up plane, and the platform can be slewed about the vertical axis to null the input of one of its horizontal gyroscopes, at which time that gyroscope input axis will point east-west. That is the basic concept used for gyrocompass alignment, but practical implementation requires filtering⁸ to reduce the effects of sensor noise and unpredictable zero-mean vehicle disturbances due to loading activities and/or wind gusts.

Strapdown INS Gyrocompass Alignment Gyrocompass alignment for strapdown systems (see Fig. 9.16) is a process for "virtual alignment" by determining the sensor cluster attitude with respect to navigation coordinates using only the sensor outputs while the system is essentially stationary.

If the sensor cluster could be firmly affixed to the earth and there were no sensor errors, then the sensed acceleration vector a_output in sensor coordinates would be in the direction of the local vertical, the sensed rotation vector ω_output would be in the direction of the earth rotation axis, and the unit column vectors

⁸The vehicle dynamic model used for gyrocompass alignment filtering can be "tuned" to include the major resonance modes of the vehicle suspension.

would define the initial value of the coordinate transformation matrix from sensorfixed coordinates to ENU coordinates:

In practice, the sensor cluster is usually mounted in a vehicle that is not moving over the surface of the earth but may be buffeted by wind gusts and/or disturbed by fueling and payload handling. Gyrocompassing then requires some amount of filtering to reduce the effects of vehicle buffeting and sensor noise. The gyrocompass filtering period is typically on the order of several minutes for a medium-accuracy INS but may continue for hours or days for high-accuracy systems.