A Novel Power System Transient Stability Prediction Method

本文提出了一种新的快速学习的在线预测电力系统暂态失稳的方法，并给出了一个应用于单机和无穷大母线的算例。该算法是从一个经过验证的机器人ball-catchingalgorithm，其中包括快速学习。对于不稳定性预测，球的位置被替换为测量的相对发电机转子角度。使用测量的相对转子角，控制算法预测未来时间的转子角。相对转子角以每秒600次的速率采样。这种新的快速学习算法将预测未来500-1000毫秒的转子角度。预测的发电机相对转子角的增加超过预定阈值是将发生失步的预测。当预测到失步时，保护方案可以启动稳定援助，如发电机跳闸，制动电阻器和/或快速valving.Preliminary调查表明，接球算法可用于暂态稳定预测。本提案要求提供资金证明该方法的可行性.

This proposal suggests a new fast learning, on line method for the prediction of powersystem transient instability and an example of its application to a single machine andinfinite bus. The proposed algorithm is adapted from a proven robotic ball-catchingalgorithm, which includes fast learning. For instability prediction, the ball location isreplaced by the measured relative generator rotor angle. Using the measured relative rotorangle, the control algorithm predicts the rotor angle at a future time. The relative rotorangle is sampled at a rate of 600 times per second. This new fast learning algorithm willpredict the rotor angle 500-1000 milliseconds into the future. The increase of thepredicted generator relative rotor angle beyond a predetermined threshold is a predictionthat loss of synchronism will occur. When loss of synchronism is predicted a protectionscheme can initiate a stability aid such as generator tripping, braking resistor and/or fastvalving.Preliminary investigation shows that the ball-catching algorithm may be applied fortransient stability prediction. This proposal requests funding to prove the feasibility ofthis method.