Research On Demand Response Mechanism And Short-term Load Forecasting Model Under Smart Grid

With the continuous development and improvement of the power market,the interests of the main body gradually diversified,and electricity pricing mechanism has been adjusted and updated.In order to optimize the resource allocation of the whole power system and alleviate the shortage of short-term load capacity,the demand response becomes a hot research topic in the field of electric power.In the technical support of the smart grid,demand response can promote real-time interaction between grid and users by making time-of-use price,which is conducive to the rational allocation of power resources,and ultimately benefit both sides cooperated with short-term load forecasting.Therefore,it is of great significance to analyze the demand response mechanism of the users under TOU price,and to study the new method of short-term load forecasting considering demand response.According to the principle of consumer psychology,a demand response mechanism model based on piecewise function is established and used to achieve quantitative analysis of TOU price.Because the model does not fully consider the influence of the non price factors,the fuzzy attribute is given to demand response.The load time series clustering method on the basis of data mining is adopted to cluster the annual load sequences of a power grid,which takes the intersection constraint formed by the Euclidean distance between load sequences and the standard deviation of difference sequences as criteria.According to the clustering results,the main characteristics of the load are analyzed from such two aspects as the numerical value of the load and the morphological changes so as to make the dynamic peak-valley electricity price more reasonable.According to the practical example,the demand response load curve of the day-forecasted is fitted,and the results show that the effect of dynamic peak-valley electricity price is significant and participating users can gain income.In addition,this paper describes the demand response mechanism in the form of curve,and establishes the user's actual demand response model based on the time-varying function.It studies basic characteristics of the load which takes demand response into consideration through frequency spectrum analysis.On this basis,the input component of the prediction model is determined,and the RNN,Elman-NN,RBF-NN prediction models considering demand response are established.Through the actual example,the prediction performance is compared before and after considerinig demand response factors in the three prediction models.The results show that the RBF-NN model has the best prediction performance,and the prediction results can be improved greatly by introducing the demand response quantification results into the prediction model.Combined with dynamic time-of-use price,the paper describes the customers' demand response mechanism based on Logistics Function in accordance with the theory of customer psychology.This paper constructs a short-term load forecasting model based on RBF-NN,which takes into consideration the comprehensive factors affecting demand response by quantizing such external factors as price,the degree of customers' response,temperature and so on.The demand response mechanism based on Logistic function takes into account the psychological response of power users to different electricity prices,and the response curve can be continuously guided at the piecewise points of different prices.It is more in line with the objective fact than the subsection function describing the demand response mechanism.This paper demonstrates the importance of taking into overall account such factors as price and the degree of customers' response in RBF-NN,which provides theoretical basis for further studies on short-term load forecasting considering demand response.