Data-Driven Multiple-Task Dynamic Security Assessment and Its Application on Optimization of Emergency Control Strategy for Frequency

Maintaining the dynamic security and stability is a primary task of the power system operation. Dynamic security assessment (DSA) of multiple security risk attributes is the basis of formulating security control strategy. For the coupling problem of multiple dynamic security risks of the power system after emergency control strategy, a data-driven multiple-task DSA method based on hybrid convolutional neural network (CNN) is proposed. Multiple dynamic security risks are evaluated in the meantime, and the computation resources are made use of efficiently. The constructed multiple-task DSA model is applied on the optimization of emergency control strategy for frequency of receiving-end power grid, which significantly improves the online optimization efficiency of control strategy. Taking a receiving-end power grid with multi-infeed DC as an example, the validity of the proposed method is verified.