Research On Short-term Load Forecasting Model Of Power System That Takes Into Account Generalized Demand-side Resources

Due to the unique superiority in demand-side resource scheduling and allocation,as well as the substantial contributions to energy conservation and environmental protection,demand-side management,an important means to balance the supply and demand of energy in the power market,has become an extensivelly recommended alternative to power-generation side in the power system.Under the circumstances of an increasingly open demand-side market,some users with the ability to adjust and respond will gradually get to change the rigid identity in traditional load and participate in the power market activities in the form of demand response through the integrated operation of aggregation mechanism.The widespread participation of demand side users,to some extent,effectively alleviates the peak load stress of the power grid,while it also causes the changed daily load curve at the same time.This study focuses on the impact of scheduling behavior of generalized demand side resources on daily load curve and short-term load forecasting to work out a new method with better forecasting performance and stronger adaptability for short-term load forecasting of modern power systems.To start with,we introduced and explained the new connotation of short-term load forecasting technology in smart grid.Based on the background of the intelligent development of power grid,the inadaptability of the current short-term load forecasting model was analyzed in detail,and some reasonable suggestions were given for the improvement.Then,this paper discussed the impact of scheduling behavior of energy storage resource on short-term load forecasting.According to the operation control strategies of energy storage scheduling based on electricity price and contract,the corresponding charging and discharging models were constructed respectively,and the charging and discharging behavior of energy storage was described in the form of load curve.Electricity price and contract factors related to energy storage scheduling behavior were integrated in the construction of the improved short-term load forecasting model,and Elman-NN was adopted for the short-term load forecasting.The two kinds of short-term load forecasting methods before and after improved were compared through the example simulation,and the results showed that the improved model taking account of energy storage scheduling factors better behaved than the conventional short-term load forecasting model in prediction performance,which provided a reliable basis for research projects aiming at the impact of considering generalized demand side resource scheduling behavior on short-term load forecasting.Finally,to further consider the influence of diverse demand response factors on short-term load forecasting,this article,based on the new model for demand side resources taking part in spinning reserve market under the actual power market mechanism,proposed a new generalized demand side resource scheduling model which could be realized through electricity price contract on the basis of single resource scheduling model of energy storage.This model took the maximum economic benefit of all kinds of demand-side resources at the load aggregator as the goal of scheduling optimization,and could submit the scheduling results to the system dispatcher in the form of demand response signals to make use of.A short-term load forecasting model of power system considering generalized demand side resources was constructed in this paper,with demand response factors incorporated into the input of the improved model.Through the example simulation to compare the prediction outcomes of the two short-term load forecasting models considering the generalized demand side resources or not,it was turned out that the improved forecasting model had more significant advantages,which provided theoretical support for the improvement of short-term load forecasting model under the smart grid in the future.