Cooperative flow field estimation via relative and absolute motion-integration errors of multiple AUVs

This paper presents a cooperative flow field estimation algorithm based on the sensor measurements from a group of autonomous underwater vehicles (AUVs). The algorithm uses: (1) the deviation between actual and predicted relative positions between each vehicle and its neighbors (relative motion-integration error), which is available using local measurement, and (2) the deviation between the actual and predicted positions (absolute motion-integration error) of each vehicle, which is available from GPS when the AUVs are on the sea surface, to reconstruct the flow field. Firstly, we establish the relation between the motion-integration errors and the flow field, which is given in the form of a set of nonlinear equations. An iterative algorithm is then proposed to estimate the unknown flow field by solving an inverse problem for these nonlinear equations. Moreover, it is rigorously proven that under this algorithm, the estimation result converges to the real flow field. Finally, simulations are conducted to illustrate the effectiveness of the proposed algorithm.