An arbitrary segmentation method for loss allocation in power grids with distributed generation

The issue of network loss allocation is crucial for the economic operation of power grids, as fair distribution ensures that participants allocate the cost of losses in proportion to their actual contributions. This, in turn, optimizes resource allocation and supports the sustainable operation of the power grid. In response to this challenge, this paper proposes a novel network loss allocation method based on the arbitrary segmentation method and graph theory, applicable to complex transmission and distribution networks with distributed generation integration. Initially, a theoretical approach for calculating network loss improvement in power grids with distributed generations is proposed, utilizing bus voltage equations and the PQ-PV subnetwork disassembly method to mitigate the impact of bidirectional power flow problem. Subsequently, a new method for identifying abnormal network loss measurement data is developed, leveraging the “rank sum” characteristic to identify discrepancies. On this basis, an anomaly correction process is then applied using a segmentation region estimation model. Finally, the power grid is divided into simpler regions using the arbitrary segmentation method and graph theory, and network losses are allocated to users and distributed generations via a power flow tracing method, coupled with enhancements to the average network loss coefficient method. The proposed framework is validated on two distinct network structures, the IEEE 14-bus transmission system and an NC-10 kV distribution network, demonstrating its effectiveness and applicability. Numerical results indicate that this method can be successfully applied to any type of power grid with distributed generation integration.