A shape-based cutting and clustering algorithm for multiple change-point detection (Q2293636)

The reduced complexity algorithm using data-cutting is designed for change-point detection problem. The suspected change-points are selected by the cutting step and the later steps are only implemented on the set of the suspected change-points. As result, the time cost of the proposed method is less than \(O(n)\). The effiency of the proposed algorithm is demonstrated on the scenario reduction for the planning of the grid company which needs analysis of huge amount of data including voltage, current and load for each transformers in a city electric distribution network. The proposed shape-based cutting and clustering algorithm enables auto-segmentation of the objective planning periods, which determine the boundary conditions of the electric distribution network operation.