The output from the Inertial Measurement Unit (IMU) consists of specific force and rotational velocity. The rotational velocity is integrated to acquire the angular displacement, which is used to calculate the rotational matrix, C^t . The Inertial Navigation System projects the specific force from the IMU’s body axes onto the global axes, using the rotational matrix.
Detector hypothesis is that the platform has zero velocity, and thus the error is zero minus estimated velocity.
Gravitational acceleration is subtracted from the specific force, which is then integrated twice to get an estimate of the displacement, and for each integration a correction factor, δv^t,δs^t , from the Kalman filter is added to the output.
The rotational velocity and the force are also passed to the Kalman filter, whereupon they are used to update the current estimate of the state vector of the system, s^t, and its covariances.
Meanwhile the specific force is also passed to the zero velocity detector (ZPD), which detects when the IMU is at zero velocity, at which point it triggers the Kalman filter to reset the process errors.