Transient voltage stability assessment of grid-connected wind power systems based on the grcforest model

Abstract

In view of the shortcomings of machine learning in the fast and accurate evaluation of the transient voltage stability of wind power grid-connected systems at the present stage, a grcForest model-based transient voltage stability assessment method for wind power grid-connected systems. Firstly, the number of input features increases with the number of cascading forest layers The gradient growth or gradient reduction that may occur with the increase of the number of layers in the cascade forest is optimised by using a residual network to ensure that the model can still maintain its initial learning capability after the number of layers increases. Secondly, the key factors influencing the transient voltage of the grid-connected wind power system are analysed and input features are constructed by combining transient faults; then the model is evaluated by The model is then trained offline to complete the parameter setting and performance optimization; finally, the completed input features are applied to the grcForest wind power grid-connected system transient voltage The simulation analysis of the 39-node system of the IEEE10 machine validates the rapidity and accuracy of the method. The simulation analysis of the IEEE 10-machine 39-node system verifies the rapidity and accuracy of the method.