Research Of Harmonic Assessment Based On Improved Binary Linear Regression Method

With the rapid development of scientific, now the load structure and the characteristics of the power grid have happened a qualitative change, putting forward the higher requirements to the power quality. At the same time, the harmonic currents produced by a large number of nonlinear loads float in the grid, which produce the harmonic voltage in the bus. In the harmonic assessment and management, It’s significant to clear the harmonic responsibility. In the paper, the harmonic characteristics of nonlinear loads and the generation principle of harmonic are analyzed above all, the meaning and properties of power system harmonic are described, the relevant measurement indexs of harmonic distortion are described in theory.And then, the "invasive" and "non-invasive" assement methods which are based on the harmonic emission level are described detaily, and the advantages and disadvantages of various methods are analyzed. Now at home and abroad these methods are mainly used in the calculating of the voltage drop because of the harmonic currents caused by the system and users respectively, use the voltage drops to assess the respective responsibility for the harmonic pollution. But in reality, voltage distortion in each of the bus is not caused by a single nonlinear load, but It’s the effect of all the nonlinear loads. So finally, the paper puts forward the estimation method of binary linear regression to calculate the background harmonic voltage and the harmonic impedance for the more bus and more harmonic sources network, the observation datas are based on the harmonic voltages and currents in the busbar which we are care about, according to the equivalent network circuit model deduces the regression equations, calculating the transfer harmonic impedance and the background harmonic voltage. Then divise the responsibilities, and through the PSCAD simulation to validate the method is effective and feasible.