Distributed state estimation of hybrid AC/HVDC grids by network decomposition

Power system state estimation is traditionally formulated and executed as one integrated problem while considering all the network components and their interactions in the grid simultaneously. Such simultaneous approach presents challenges when the concerned network consists of different types of systems operated by different entities (ISO, TSO or utility). A leading example is provided by hybrid AC/HVDC grids. We propose a method that complements the simultaneous approach, where the system network is decomposed into several subsystems on which the state estimation is executed separately in an iterative way. Hybrid AC/HVDC grids naturally decompose into AC and HVDC subsystems. The method uses LaGrangian relaxation based approach and blockwise Gauss-Seidel solution technique to arrive at a solution. The solution approach is illustrated using hybrid AC/HVDC grids formed by IEEE AC grid networks along with monopole and bipole VSC HVDC grids. The presented technique is capable of handling monopole/bipole converter configurations, parallel converters, converter-less DC buses, wind farm connections and grid contingencies. Although hybrid AC/HVDC grids is considered as a primary application, the method employs a general formulation applicable to address any power grid that can be decomposed into a number of parts.