Study On Dynamic Equivalence Methods For AC/DC Hybrid Large Scale Power Systems

With the advantages of HVDC on large capacity and long distance power transmission, a great deal of HVDC projects have put into service in China, thus promoting the form of AC/DC hybrid large scale power systems such as China Southern Power Grid and East China Power Grid.When study on the stability of AC/DC hybrid large scale power systems, the electromagnetic transient model must be used to reflect the rapid transient characteristics of HVDC accurately. For the scale of electromagnetic transient simulation is limited,the AC power system should be simplified to meet the requirements of electromagnetic transient simulation. This paper focuses on the simplification of large scale AC power systems,study on the dynamic equivalence methods based on short circuit current calculation by using BPA.This paper include the following four points.(1)A comparison and analysis is made between common dynamic equivalence methods and engineering used dynamic equivalence methods, the dynamic equivalence method based on short circuit current calculation is proposed.Equivalence principles and steps of the method are given with formula in details.(2) An equivalent is made of East China Power Grid in the year of2012with the dynamic equivalence method based on short circuit current calculation. Dynamic simulation analyses of the reduction system show that the method is appropriate, effective and suitable.(3) Influence of different load models to the dynamic characteristics of systems is contrasted through examples, the results show that dynamic load must be considered when research the stability of AC/DC hybrid large scale power systems.(4) Common load modeling methods are compared, and common parameter identification methods are introduced.This paper focuses on the application of weighted least-squares with constraints on parameter identification of synthesis load model. Parameter identification is made of Shanghai Yanggao with this method; the results show that the algorithm is effective and accurate.