Harmonic State Estimation In Distributed Network Under Random Fluctuations Of Source Grid And Load

A large number of harmonics injected into the distribution network,which degrades the power quality,brings economic losses,and affects the safe and stable operation of the distribution network.In order to grasp the harmonic level of all nodes in the distribution network,accurate harmonic state estimation is required.With the dramatic increase in the number of power electronics as well as distributed power supplies,harmonic state estimation is faced with a more complex environment.The randomness and intermittency of the harmonics injected by the grid-connected distributed power supply,the frequently changing situation of network parameters,and the randomly fluctuation of load harmonics bring difficulties to the accurate estimation of the harmonic state.Therefore,the thesis focuses on harmonic state estimation under above three random fluctuations of source network and load are studied.The main contents of the thesis are as follows:1)A method employing both semi-invariant and the least squares method is proposed to estimate the harmonic state under the condition of distributed wind power access to distribution network.Firstly,semi-invariant is used to convert the estimation of the harmonic current at each moment into the estimation of the semi-invariant of the harmonic current at a large number of moments.The least squares method is used to estimate the half-invariant of harmonic currents.The Gram-Charlier series are used to culculate the probability density and probabilistic characteristic quantities of harmonic currents based on the half-invariant.Simulation results on the IEEE 13-bus test system show that the probability harmonic state estimation method can adapt to the randomness and intermittency of the harmonics injected by the grid-connected wind farm,and effectively estimate the probability distribution and probability feature of the harmonic current.2)In order to adapt to the distribution network parameters change,a harmonic state estimation method based on segmented independent component analysis and measurement matrix deviation is studied.A segmented independent component analysis model is established without harmonic impedance matrix,and the harmonic current is estimated by the data segment overlap method.On this basis,the deviation of the measurement matrix is used as the criterion of network parameter change.And then the harmonic current estimation results are modified when the network parameters are judged to change.Simulations are carried out on the 34-bus system,which verify that the proposed method can obtain accurate harmonic state results under the condition of the grid line parameters change.3)Considering the random fluctuation of load harmonics,a harmonic state estimation method based on dynamic feature extraction of load harmonics and adaptive Kalman filter is studied.Firstly,the fast and slow fluctuating components in the historical load harmonics are extracted through the dynamic feature extraction method.Then the fast and slow fluctuating components are used to estimate the Kalman filter noise parameters and state transfer matrix parameters respectively.On this basis,the Kalman filter is applied to solve the harmonic state.To adapt to the complex noise environment,the noise parameters are adaptively modified by the covariance matching criterion and the noise estimator.The simulations on the IEEE 13-bus and IEEE 69-bus system verify that the proposed method has the advantages of high accuracy and adaptability under load fluctuations.