On The Uncertainty Theory Based Analysis Of Power System Small Signal Stability And Online Prediction And Service Restoration Strategy

With the interconnection of power grids, power systems become more and more complicated. The uncertainties, such as output of power generations and loads will affect small signal stability of power systems. Therefore, online prediction of small signal stability and small signal stability analysis considering uncertainties are highly expected. On the other hand, distributed generation acts as supplement to large power grid, which can restore loads with high priority after blackout. Therefore, the service restoration strategy for power systems with distributed generations is also worth deep study. The main contents of research are as follows:Firstly, uncertainties such as random fluctuation of power generations and loads are taken into consideration for effective analysis of small signal stability of power system. A TPE (Two Point Estimate) based method is presented to fulfill probabilistical analysis of small signal stability with uncertain factors. Comparing to the MCS (Monte Carlo simulation) method, the TPE method has much lower computational burden while keeping equivalent performance. However, the assumption of distribution function of uncertain factors has important influence on the result of small signal stability analysis.Secondly, a credibility theory (CT) based method is proposed, which takes the uncertainties into consideration. The CT method is able to not only give the expectations and variances of eigenvalues of DAEs (Differential-Algebraic Equations), but also calculate the small signal stability indices based on credibility theory for analysis and decision of planning. The CT method avoids hypothesis of the uncertain parameters distributed function’ type and the function parameters’ statistics, and does not require independence of uncertain parameters. The membership function of uncertain parameters is required, but it has no influence on the result. Compared with certain analysis method, two point method (TPE), and Monte Carlo simulation (MCS) method, the feasibility and applicability of the CT method are verified.Thirdly, an approach for online prediction of small signal stability based on the Case-based Reasoning (CBR) theory is proposed. Based on EMS and WAMS records of previously occurred low frequency power oscillations, their characteristics are off-line investigated and then stored into a database. Once current operation status is available via EMS or WAMS, its small signal stability can be predicted by carrying on line CBR in the database. Further, control strategy can also be obtained to improve small signal stability in case it is no satisfied.The last, a service restoration strategy is proposed using distributed generations. A multi-objective optimization model is formed, in which the two objective functions, namely the minimum load shedding and minimum breaker action, are optimized in two steps separately, so as to reduce computational burden. Further, power quality index, in particular, fundamental frequency is introduced into constrains of the optimization model, and a flexible power flow algorithm is used to obtain fundamental frequency of each solution of optimization procedure. Consequently, final result of service restoration accords well with practical situation.Performance of the proposed methods are verified using IEEE benchmarks and practical data recorded by the WAMS in a regional power grid. The results proved the validity,effectiveness and advantages of the proposed methods.