Overview

The accuracy of an inertial navigation system is often expressed as a position uncertainty after a given period of navigation, or on reaching a given destination. Alternatively, it is expressed in terms of the rate at which the navigation error builds up with time, in nautical miles per hour for instance. The actual form of expression used for this overall accuracy figure depends upon the application. For example, for an inter-planetary missions, the accuracy refers to the desired point of closest approach to the 'destination' planet. For navigation in the vicinity of the Earth, it is usually the errors in two dimensions which are of most interest, the along-track and cross-track position errors over the surface of the Earth. These errors are often combined to yield a single number which expresses navigation accuracy after a given navigation time, the circular error probable (CEP) or circular probable error (CPE) as it is sometimes called. Essentially, this defines a circular area within which the navigation system estimates its true position to be, with a certain probability. The 50 per cent CEP is a frequently quoted figure. When the probability value is not stated, it usually means a 50 per cent value should be assumed.

Consider now the composition of this navigation performance figure. In practice, navigation errors propagate owing to a large number of error sources which include alignment errors, a variety of inertial sensor errors and errors attributable to computational inaccuracy. In general, each errors may be regarded as comprising...