Research On Identification Of Grid Equivalent Model Parameters Based On Transient Voltage And Current Fault Components

With the continuous expansion of power grid interconnection, grid structure isbecoming more and more complex, grid online analysis and calculation is becomingmore difficult. The existing online analysis and calculation of complex grid areprimarily based on simplifying the complex grid and the basis of this work is gridequivalent. The accuracy of the grid equivalent depends on the validity of parameteridentification method. Existing methods have some disadvantages, such as complexidentification process, large amount of information needed, low identification efficiencyand so on, which will severely affect the parameter identification accuracy of theequivalent model and the real-time tracking capability. Transient voltage and current hasa characteristic of changing with equivalent model parameters. Alone with transientvoltage and current fault components of one running condition, equivalent parametercan be identified, which can realize real-time grid parity. Therefore, it’s worthwhile tostudy the parameter identification method of the equivalent grid, which based ontransient voltage and current fault components. The main contents are as follows:①The grid equivalent model parameter identification strategy based on voltageand current fault components was proposed. Transient components informations wereused to calculate equivalent model parameters by building the high-order differentialequations matched with network port transient characteristics. Analyze the existingproblems and causes of the strategy to identify the static, transient and dynamicequivalent model. Based on the above parameter identification strategy, this paperproposes a new ideas for parameter identification for the grid thevenin equivalent,electromagnetic transient equivalent and dynamic induction motor.②For the mismatch problem between transient electric parameters of theequivalent network and the thevenin equivalent model. The thevenin equivalent modelparameter identification methods based on voltage and current fault components wasproposed. Firstly, high-order differential equations matched with network port transientcharacteristics were established. Then, equivalent network parameters were calculatedwith the relations between linear coefficients of differential equations and grid theveninequivalent parameters. The method needs less informations,and the parameteridentification process is simple. Finally, a simple single-phase system and IEEE9system cases verified the correctness of this method. ③For unidentifiable problems of some parameters when directly identify theactive linear network using transient electric parameters of the equivalent network, astep identification strategy based on trajectory sensitivity is proposed. Active linearnetwork differential equations are established. High sensitivity parameters were firstlyestimated according to transient information, and others were estimated through globalconstraints on network parameters using Thevenin equivalent impedance. The methodcan accurately identify circuit parameters under one operating condition, theidentification process is simple and of high recognition efficiency. Finally, thesimulation of two order active linear network has verified the correctness andeffectiveness of this method.④For the problem that it can’t directly identify the practical parameters ofinduction motor based on transient electric parameters. The dynamic induction motorparameter identification methods based on voltage and current fault components wasproposed.The differential equations model in which all the variables can be measuredwas deduced in detail based on the IM Parker model, and IM electrical parameters wereestimated step by step with trajectory sensitivity strategy. Then practical parameterswere identified by establishing the parameters relationship between IM practicalparameter model and its Parker model. Besides, the differential equations model inwhich all the variables can be measured was deduced, and mechanical parameters wereobtained by direct integral method. This method avoids solving complex differentialequations, and simplifies the parameter identification method,and improves theefficiency of identification. Finally, one induction motor case verified the correctnessand effectiveness of this method.⑤The IM non-linear equivalent model parameter identification methods wasproposed. Induction motor rotor resistance parameters are a nonlinear function of theslip s. After the non-linear parameters of induction motor on different points wereidentified, the changing relationship between rotor resistance parameters and the slip swas formed with the ladder network. Besides, the changing relationship between rotorresistance parameters and the slip s was formed using the relationship between IMpractical parameter model and its Parker model,and non-linear equivalent modelparameter identification were achieved.