Estimation of circular error probability of strapped down inertial navigation system by propagation of error covariance matrix

Covariance propagation methodology has been tested successfully for estimating errors of a strapdown inertial navigation system. The error models for both the inertial frame and local vertical frame have been developed and used using the relative quaternion parameterization for attitude. The circular error probability (CEP) predicted from covariance of system errors is shown to be lower than from the conventional root sum square (RSS) technique used in the Monte Carlo approach. It has been found that the covariance method of predicting CEP provides a large saving in computational time compared to the Monte Carlo approach