Improved Research For The Online Identification Of The Dominent Dynamic Parameters

The digital simulation is the important basis for the design and plan of power system as well as for the decision making. The precision of parameters in power transient simulation would directly affects the difference between simulation results and actual measuring results. Correct and fast identification of parameters would be helpful to improve the online application of power system transient simulation. Over the years, the dynamic parameter identification technology based on field trials and tests has made great progress. However, there is still quite a long way to go to meet the needs of actual application.The dominant dynamic parameter identification method on the basis of PMU measurement makes use of Volterra series transition and pattern classificaiton technology. With the help of dominant dynamic parameter identification based on WAMS, this method employs little measure information and intends to realize the online fast identification of some parameters which have great influences on the dynamic process of the power system. This paper mainly applies the composite load model which is composed in parallel by equivalent dynamaic load of induction motor and equivalent static load of impedance according to the percentage, and conducts a systematic research on the identification of the dynamic component percentage which is the reflection to the composition of the load model. Firstly, this paper analyzes the dominance of the parameters in composite load model system and further ensures the importance and feasibility of identification of dynamic component percentage. Secondly, this paper works on the major factors which influence the precision and accuracy of the results in the identification of dynamice component percentage with the help of dominant dynamic parameter identification method on the basis of PMU measurement, and shows that there is relativity among the typical value of parameters, sample classification and identification result accuracy. It also puts forward an identification plan to build sample classification storeoom for typical load. Moreover, this paper designs an offline simulation method to measure the rational range of parameter typical value on the premise of the correct identification of dynamic component percentage. It makes use of PSASP simulation results of one provicial power system and realizes the parameter identification of dynamic component percentage, verifying the feasibility of the identification method. At the same time, this paper carries a simulation test of supposed fault in the composite stable simulation programme and proves that the identification accuracy of dynamic component percentage in composite load model with the help of dominant dynamic parameter identification method on the basis of PMU measurement has nothing to do with fault spots and types, which holds some generality. Finally, this paper investigates the relationship between accuracy of identification result and partition distance.