Estimating Angular Velocity from Output of Rate-integrating Gyro

The Kalman filtering technique can be used to improve estimates of angular velocity (in a strapped-down navigation system) using angular increment (rate integral) data. A simple, second-order, constant coefficient digital filter can yield significant improvement for a very high or a very low noise-to-signal ratio·r = 4 σθ /σ α T2, where σ θ is the rms noise in measuring angular increments, σ α is the rms angular acceleration, and T is the sampling time. For r ≈ 1, the filter affords negligible performance improvement.