Estimation of the dynamics of two-dimensional clusters
Daniele Casagrande, Mario Sassano, Alessandro Astolfi · 2009
A method to estimate the dynamics of clusters is presented. An existing static clustering method, which makes use of a Hamiltonian function, is exploited in the dynamic case, when the points to cluster do not remain in fixed positions but move with unknown dynamics. The time is discretised and the static algorithm is applied to each time-instant. While identifying each cluster, the algorithm also performs the computation of the moments of the cluster. By estimating the dynamics of the moments an approximate estimation of the dynamics of the cluster is obtained. An illustrative example shows the performance of the method.