User:Mbr5002/sandbox

GPS/INS refers to the use of GPS satellite signals to correct or calibrate a solution from an Inertial Navigation System (INS). Inertial navigation systems usually can only provide an accurate solution for a short period of time. The INS accelerometers will produce an unknown bias signal that appears as a genuine specific force. This is integrated twice and produces an error in position. Additionally, the INS software must use an estimate of the angular position of the accelerometers when conducting this integration. Typically, the angular position is tracked through an integration of the angular rate from the gyro sensors. These also produce unknown biases that affect the integration to get the position of the unit. The GPS gives an absolute drift-free position value that can be used to reset the INS solution or may be blended with it by use of a mathematical algorithm such as a Kalman Filter. The angular orientation of the unit may be inferred from the series of position updates from the GPS. The change in the error in position relative to the GPS may be used to estimate the unknown angle error.

The benefits of using GPS with an INS are that the INS may be calibrated by the GPS signals and that the INS can provide position and angle updates at a quicker rate than GPS. For high dynamic vehicles such as missiles and aircraft, INS fills in the gaps between GPS positions. Additionally, GPS may lose its signal and the INS can continue to compute the position and angle during the period of lost GPS signal. The two systems are complementary and are often employed together.

Applications
GPS/INS is commonly used on aircraft for navigation purposes. Using GPS/INS allows for smoother position and velocity estimates that can be provided at a sampling rate faster than the GPS receiver. This also allows for accurate estimation of the aircraft attitude (roll, pitch, and yaw) angles. In general, GPS/INS sensor fusion is a nonlinear filtering problem, which is commonly approached using the Extended Kalman Filter (EKF) or the Unscented Kalman Filter (UKF). The use of these two filters for GPS/INS has been compared in various sources     , including a detailed sensitivity analysis. The EKF uses an analytical linearization approach using Jacobian matrices to linearize the system, while the UKF uses a statistical linearization approach called the Unscented Transformation which uses a set of deterministically selected points to handle the nonlinearity. The UKF requires the calculation of a matrix square root of the state error covariance matrix, which is used to determine the spread of the sigma points for the Unscented Transformation. There are various ways to calculate the matrix square root, which have been presented and compared within GPS/INS application. From this work it is recommended to use the Cholesky decomposition method.

In addition to aircraft applications, GPS/INS has also been studied for automobile applications such as autonomous navigation,  vehicle dynamics control , or sideslip, roll, and tire cornering stiffness estimation.