A novel dynamic spatio-temporal graph convolutional network for windspeed interval prediction

It is crucial to predict the wind speed for the utilization of renewable wind energy and the operation oftransmission lines with increased capacity. The intermittency and stochastic fluctuations of wind speed posea significant challenge for the high-quality wind speed prediction, and a novel wind speed interval prediction(WSIP) model is constructed in this study by employing the residual estimation (RE)-oriented dynamic spatio-temporal graph convolutional network (DSTGCN) approach. Firstly, a dynamic adjacency matrix is designed toobtain time-varying global spatial weight allocations among each wind speed node. Then, the spatio-temporalfeatures are extracted by using gated recurrent units (GRUs) and GCNs to construct the wind speed graphnetworks. Moreover, the RE-oriented strategy incorporating the pinball loss is designed to provide a guidancethe parameter training of the constructed model, thus eliminating the quantile crossings problem. As a result,the deterministic point prediction of the wind speed is expanded to the quantile-based probabilistic intervalprediction. Finally, the experimental results are presented to demonstrate the validity and superiority ofproposed scheme in both qualitative and quantitative performance