Regional electricity market price forecasting based on an adaptive spatial–temporal convolutional network

The accurate prediction of electricity prices has great significance for the power system and market, regional are difficult to predict due congestion issues in transmission lines. A price framework is proposed based on an adaptive spatial–temporal convolutional network. expected better explore prices’ dynamic characteristics spot market improve predictive accuracy prices. First, different areas regarded as nodes. Then, each area’s historical data used corresponding node’s characteristic information constructed into a graph. Finally, graph containing input price. Operational from Australian adopted, results compared with those existing methods. numerical example show that than baseline similar In twelve-step forecast this paper, considering spatial dependence can by at least 10.3% up 19.8%.