| Under the background of "carbon peak,carbon neutral",electric power substitution is developing rapidly in the energy industry.With the change of energy structure and the increase of electricity demand,electric heating will become one of the main heating modes.Large-scale integration of electric heating load into distribution network will have a great impact on various indicators of distribution network.Considering the large-scale integration of electric heating,accurate prediction of electric heating load is carried out.Comprehensive evaluation of distribution network is the main content of this paper.Firstly,according to the weather classification method matching weather factors and power fluctuation process,the forecast model of electric heating is divided into 5 weather types.The forecasting process of electric heating power is divided into fluctuating process and smoothing process.Prediction of Fluctuation Process by variational mode decomposition and CNN-LSTM network,The results of electric heating load forecasting are more accurate by adding the fluctuation process and smoothing process.Secondly,in the case of large-scale connection of electric heating load to distribution network,the comprehensive evaluation of distribution network is studied.Based on the traditional evaluation system of distribution network,the uniformity of power flow distribution is defined and used as reliability index,and the economic index of investment capacity ratio is proposed.On this basis,a comprehensive evaluation index system including security,reliability,quality and economy is established.Objective weight and subjective weight are obtained by using intuitionistic fuzzy AHP and entropy weight method,and the optimal combination weight is obtained by relative entropy weight method.So that the weight reflects the subjective attitude and preferences of users,while maintaining an objective state.Finally,a comprehensive evaluation model of distribution network based on trapezoidal cloud matter-element model is proposed.Using multidimensional matter-element to qualitatively and quantitatively express the evaluation index characteristics of distribution network,and fitting the fuzzy uncertainty in the evaluation process with the randomness of the related degree calculation of index cloud.Trapezoidal cloud model is used to balance the membership degree of evaluation index to evaluation grade,and the concept of cloud correlation function is used to describe the degree of belonging to certain evaluation grade.This paper compares and analyzes the evaluation results of four typical users in the northwest area under the heating load,and the evaluation results are more accurate and humane.Practice shows that the model and method proposed in this paper are feasible,which provides a reference for the research of distribution network comprehensive evaluation when the heating load is connected to the grid. |