Harmonic Analysis And State Estimation Of Distribution Network Based On Sparse Signal Processing

After the transformation of the energy structure,the distribution network has become more and more important in the power system.Therefore,it is very important that the distribution network is sensible,appreciable and controllable,and efficient processing of information is the key.At the same time,with the development of power Internet of Things and digital grids,power data information has become the focus of attention at present.Sparse signal processing technology is a hot research topic that has emerged in recent years.By establishing and solving sparse recovery equations(underdetermined equations are solved under certain conditions),the amount of data required can be reduced,data redundancy and transmission pressure can be avoided.The application in signal processing is becoming more and more extensive.In the distribution network,after more and more devices are connected,the system is affected by interference and needs to be effectively monitored.State estimation and harmonic analysis are of great significance to the control and decision-making of the system.This article summarizes the development of sparse signal processing technology,the status quo of harmonic signal analysis and state estimation of distribution network,and the application of sparse signal processing combined with signal analysis and state estimation of distribution network.The original method requires a large amount of data,a large calculation pressure,and high requirements for the system.This paper uses the sparseness of the signal to carry out application research on the compression and sparse reconstruction of harmonic signals,the sparse location of harmonic sources,and the sparse estimation of distribution network power flow Jacobian matrix,which saves calculation time and storage space and improves signal processing.The efficiency of the system reduces the burden of communication and reduces the pressure on the system.The sparse signal processing technology is based on the sparseness of the signal,and a collection of a series of methodologies around this foundation,the core of which is to establish and solve the sparse recovery model.Sparsity is the key to a signal that can break the Nyquist sampling frequency and sample at a lower rate.For a sparse signal,a sparse solution model can be established after a certain matrix compression,and the problem can be solved by a certain sparse recovery algorithm in the case of underdetermined conditions.First of all,in the compression and sparse reconstruction of harmonic signals,in order to reduce the computational pressure of traditional methods,this paper uses Fourier transform bases and non-correlated Gaussian random matrices as its sparse bases and measurement matrices to establish sparse reconstruction equations.Compared with different sparse recovery algorithms,one of the algorithms with better performance is improved,and a multi-(atomic)filtering orthogonal matching pursuit algorithm is proposed,which verifies the adaptability of this algorithm under different conditions.Secondly,for the harmonic sources in the distribution system,in order to reduce the number of measurement devices required under the traditional harmonic positioning method,the sparse distribution of harmonic sources is used to establish the branch current-injected harmonic current estimation equation.Realize sparse positioning under the configuration of fewer measuring devices.The linear system is decomposed by frequency division,the sparse solution model of harmonic sources at different frequencies is established,and the sparse location algorithm of the conversion model is proposed,which is verified by the IEEE33-node power distribution system,and the configuration of the measurement device is optimized.Before this process,through 0-1 planning,the measurement device is configured to make the distribution network considerable,and the SVD-ScOMP algorithm is proposed to obtain the sparse information of the branch current harmonics.The frequency of the harmonics can still be obtained under the influence of noise.Domain information.Finally,in the distribution network state estimation,in order to reduce the number of measurement groups required,in view of the correlation,sparsity and symmetry of the power flow Jacobian matrix,this paper proposes a method based on the measurement data of the synchronous phasor measurement unit.The sparse estimation method of the power flow Jacobian matrix and the voltage/phase angle-power sensitivity matrix effectively estimates the Jacobian matrix and the sensitivity matrix under less measurement than the traditional method.Furthermore,for the bad data that appears in the measurement process,a more robust weighted least square method is introduced to improve the robustness of the algorithm.Finally,the feasibility of the method is verified through the IEEE33 node power distribution system.