Matching error of the iterative closest contour point algorithm for terrain-aided navigation

The algorithm used for terrain-aided navigation (TAN) is crucial for high performance of such a totally autonomous and long-duration underwater vehicle navigation technique. This paper presents results of an investigation into the matching errors of the revised iterative closest contour point (ICCP) algorithm for underwater TAN. In particular the quantitative relationship of the matching errors with terrain features is studied in this paper. Among 10 terrain factors, 6 of them have been shown to have the most influence on the accuracy of the revised ICCP algorithm. Three statistical methods, including multiple regression, logistic regression, and discriminant analysis, are applied to mathematically derive the different relationships between the terrain factors and the matching errors. Each formula uses no more than three terrain factors to fit the matching errors. This paper also studies the effect of other factors, including the initial error of the inertial navigation system (INS), the INS accuracy, the map resolution, the vehicle speed, the matching path length, and the sonar accuracy, on the matching errors.