| With the development of power electronics technology,the problem of harmonic pollution in power grid becomes more complex.Accurate determination of each side’s harmonic contribution is a prerequisite for harmonic control.The traditional harmonic contribution determination considers more ideal factors.More complex harmonic sources,such as wind turbine and photovoltaic,aggravate the harmonic fluctuation.The degree of fluctuation of harmonic voltage on the background side will affect the estimation accuracy of harmonic impedance and harmonic contribution.How to carry out harmonic contribution determination under the condition that the background harmonic voltage fluctuates is the main content of this paper.Considering the background harmonic voltage fluctuation,a method of harmonic contribution determination based on shifted blocks for blind source separation(SHIBSS)and data optimization is proposed.Firstly,the initial harmonic impedance solution is obtained by using SHIBSS.Then,the shortcoming of the traditional data identification method is pointed out.Furthermore,the identification method of optimal solution of system harmonic impedance based on data optimization method is proposed.Finally,the harmonic contribution is calculated according to the optimal solution.This method has higher estimation accuracy,and can overcome the problem of high failure of traditional data identification method.So it is more suitable for processing the short-term monitoring data obtained from the power grid with relatively stable operation mode.Considering the background harmonic voltage fluctuation and the system impedance variation,a harmonic contribution determination method based on similarity analysis and ordering points to identify the clustering structure(OPTICS)is proposed.Firstly,data segments with small background harmonic voltage fluctuation are selected by using the principle of correlation analysis.Then,a scatter diagram of harmonic voltage and harmonic current data set at the point of common coupling is established,and the sampled data could be divided into different clusters according to the system impedance by OPTICS algorithm for clustering.Finally,the complex domain robust regression method is used to evaluate the harmonic contribution of different clusters of data.The method can effectively reduce the influence of background harmonic voltage fluctuation and system impedance variation,which can be used in a wider range.This method is more suitable for analyzing a large number of historical data.The above two methods have their own advantages in the evaluation of harmonic contribution,but they all obtain the evaluation result of a certain harmonic.However,harmonic sources often emit harmonics of various frequencies at the same time.How to coordinate the harmonics contribution of each frequency is a problem to be studied.In addition,there are great differences in the evaluation results among different evaluation indices.It may not be reasonable to evaluate from a certain perspective.Therefore,the evaluation index of harmonic comprehensive contribution based on multi-level comprehensive evaluation is proposed in this paper.The index considers harmonic contribution of harmonic voltages,harmonic currents,total distortion of harmonic voltages and total distortion of harmonic currents as a whole.Combined with subjective analytic hierarchy process and national harmonic standards,the weights are determined and the harmonic comprehensive contribution is finally obtained.The comprehensive index overcomes the shortcoming that the single evaluation index is difficult to reflect the actual situation of harmonic pollution and realizes the quantitative evaluation of harmonic influence from the whole point of view. |
PDF Full Text Request