Research On Voltage Stability Evaluation Index And Its Algorithm Of Large-scale Power Grid

Voltage stability has become one of the most important and urgent problems in modern bulk power systems due to the continuous advancement of technology, the rapid development of economy, the widely use of dc transmission and novel power electronic devices. Voltage stability evaluation index is an effective method to assess voltage stability. Moreover, it is also the foundation of voltage stability control.With the widely use of computer and network communication, as well as the increasing attention paid to power quality and grid security, the real-time performance of estimating operating condition is emphasized in modern power systems. This paper conducted a comprehensive comparative analysis of existing voltage stability evaluation index, especially on the way to obtain voltage collapse point of the margin index with continuation power flow method.The widely used continuation power flow is a quite reliable approach while the calculation speed is not that satisfied. Aiming at this problem, an improved continuation power flow algorithm is presented in this paper. This approach combines Lagrange quadratic interpolation with secant method. Moreover, a step control strategy, which is suitable for quadratic curve, is raised in order to accelerate the speed as much as possible at the same accuracy level. The effectiveness of the proposed approach is tested on New England-39 system and IEEE-118 system.Data transmission and storage is the basis of voltage stability analysis. Regarding to SCADA system cannot transfer dynamic information, especially in fault situation, a data compression and reconstruction method for SCADA system based on orthogonal least square fitting is proposed. Data is compressed in substation side and reconstruction in SCADA side, which reduced the requirements of communication network greatly. What's more, the control centre scheduling is able to observe the dynamic information of power grid. The effectiveness of this method has been proved by EPRI 36-node system. Compared to the least square fitting method, it is more precise, especially in fault situation.