Research On Equivalent Modeling And Parameter Identification Of Small Hydropower Generator Group

The development and utilization of medium and small scale hydropower flourish in China for taking advantage of hydroelectric resource. However, the hydroelectric power plants with small scale are numerous, widely distributed and with small installed capacity, which makes it difficult to build a sophisticated model for power system security and stability analysis. At present, those hydroelectric power plants are built into simplified models with typical parameters or merge into area load as an equivalent power source. However, these approaches cannot reflect dynamic performance of those hydroelectric power plants accurately. Therefore, in this paper,an approache of dynamic equivalent modeling of those hydroelectric power plants is proposed, which consists of equivalent model structure and parameter identification.Firstly, two widely-adopted model structure in medium and small scale hydropower units, named pure load and generator parallelled with a load are introduced briefly. The accuracy of the two model structure are compared and analyzed in 10-machine 39-bus system. The model structure of generator parallelled with a load, which is more accurate than the pure load, is adopted as the equivalent model structure of medium and small scale hydropower units in this paper.Secondly, how to identify the parameters of model structure mentioned above is studied. It shows that the model structure of generator parallelled with a load has more parameters and part of them thinly effects dynamic performance of system. According to the identifiability analysis of model parameters, we divides parameters of identified model into well-conditioned parameters and ill-conditioned parameters. Well-conditioned parameters are sensitive to dynamic performance of system, while ill-conditioned parameters are insensitive. Based on this, a method named multi-cycle parameter identification with adjustment of well-conditioned parameters and ill-conditioned parameters is proposed. This method and traditional identification strategy are both applied to five order generator system. The result shows that multi-cycle identification strategy has higher accuracy than traditional identification strategy.Finally, we choose the Enshi power grid of Hubei Province which is a typical system with numerous small hydroelectric generators to be the test system. Enshi Power grid is modeled as equivalent system under two circumstances of fault in Enshi bus and fault in Sanxia bus. Comparing the results of equivalence with original model, it shows that the proposed equivalent modeling approach of medium and small scale hydroelectric generator group has excellent accuracy and practicability and some practical value in engineering.