PMU-Based estimation of voltage-to-power sensitivity for distribution networks considering the sparsity of jacobian matrix

With increasing integration of various distributed energy resources, electric distribution networks are changing to an energy exchange platform. Accurate voltage-to-power sensitivities play a vital role in system operation and control. Relative to the off-line method, measurement-based sensitivity estimation avoids the errors caused by incorrect device parameters and changes in network topology. An online estimation of the voltage-to-power sensitivity based on phasor measurement units is proposed. The sparsity of the Jacobian matrix is fully used by reformulating the original least-squares estimation problem as a sparse-recovery problem via compressive sensing. To accommodate the deficiency of the existing greedy algorithm caused by the correlation of the sensing matrix, a modified sparse-recovery algorithm is proposed based on the mutual coherence of the phase angle and voltage magnitude variation vectors. The proposed method can ensure the accuracy of estimation with fewer measurements and can improve the computational efficiency. Case studies on the IEEE 33-node test feeder verify the correctness and effectiveness of the proposed method.