Stacking-based estimation of maximal transient voltage drop with unified fault location representation

With the development of high voltage direct current (HVDC) technology, the transient voltage drop problem of HVDC inverter bus severely threatens the secure operation of receiving-end power grid. Based on the stacking algorithm of ensemble learning, this paper proposes a maximal transient voltage drop estimation method adapting to different fault locations. Firstly, based on electrical distance, a novel expression method of fault location features is proposed to represent different fault locations quantitatively. Then, in order to improve the prediction precision of single learner, a stacking algorithm is presented. Finally, based on the proposed fault location representation method and stacking algorithm, a unified estimation model of maximal transient voltage drop is built. The simulation results of a simplified provincial power system of China demonstrate that the proposed method can accurately estimate the maximal transient voltage drop under different fault locations.