Research On Load Forecasting And Dispatching Strategies For Electricity Replacement Heating Load

Under the background of double carbon,the adoption of electric energy to replace traditional heating methods such as coal and gas is the future development trend,which is conducive to the realization of energy saving and emission reduction and sustainable development.Howerver,the electric energy replacement heating load has the characteristics of being distributed,random and volatile in the process of heating,which is not easy to regulate and control.Therefore,this paper studies the electric heating user load power prediction and optimal scheduling technology,and integrates the scheduling layer and user layer to complete the scheduling strategy The implementation of the scheduling strategy is completed by considering the scheduling and user layers.The main research contents of this thesis are as follows:(1)The model of electric heating heating equipment was constructed,load characteristics were analysed and indicators were designed,and a load cluster clustering algorithm was developed.Firstly,a heat balance model of the heating building and a thermodynamic model of the electric heating equipment were designed to analyse the influence of heating power and heat dissipation coefficient on indoor temperature;Then the characteristics of electric heating user loads are analysed and indicators for evaluating load changes are designed considering load time series,extreme value characteristics,trend characteristics,fluctuation characteristics and abrupt change characteristics;Finally,a clustering algorithm for clustering heating loads based on district heating and K-Means algorithm is proposed to classify electric heating user loads in clusters.(2)A load power prediction model incorporating a data-physics-knowledge inference model is established.Firstly,based on the real-time tariff response mechanism,the real-time tariff and the load heating elasticity coefficient are used to study the heating demand and establish a physical model for predicting the load demand.Then,based on the GRNN algorithm and the load index set,a data model for predicting the short-time power of the load is constructed.Therefore,a knowledge base of the load power trend changes is established and the LSTM algorithm is used to construct a knowledge inference model.On this basis,a power prediction fusion model that fuses the above three models is designed,and the fusion weights of each model are calculated using the entropy weight method.Finally,evaluation indexes are designed,including mean absolute error,mean square error,root mean square error,median absolute deviation,and coefficient of determination,etc.(3)An optimal dispatching model for electric heating user loads and the process of implementing the dispatching strategy are designed.The scheduling evaluation index system,including economy,comfort,satisfaction and fairness of scheduling,is firstly designed by considering the heating and heating demand at the scheduling and user levels;then the objective function and constraints of the scheduling model are designed;then the inertia weights,learning factor dynamic update algorithm and simulated annealing mechanism are studied,and then the traditional particle swarm algorithm is improved;finally the electric heating user load optimisation scheduling model is established.In the dispatching process,the smart energy platform and the dispatching platform complete the load power prediction and power distribution respectively,and the load control terminal deploys the dispatching model,uses the particle swarm algorithm to complete the iterative optimisation of the dispatching strategy according to the predicted power,constraints,and objective function,and implements power distribution to the subordinate loads.Through experimental verification,the load power prediction algorithm designed has a prediction variance of 0.375,which meets the power prediction requirements of the scheduling model.Meanwhile,after the implementation of the scheduling strategy,the scheduling model designed in the paper has a certain contribution to guaranteeing the revenue of the scheduling layer and satisfying the comfort and satisfaction of customers.This strategy can provide some reference for the design of electric heating user load dispatching strategy under the transformation of heating mode,and help energy saving and emission reduction.