| Large-scale events can represent a competition of overall national strength,technological strength,cultural charm,and economic strength,play an important role in country’s political influence,economic development,cultural exchanges,and the improvement of national quality.With the improvement of our international prestige and comprehensive strength,large-scale events,such as the Olympic Games and the World Cups,will be held more frequently in China.During the whole period of an event,safe,reliable,and stable electricity supply is a must.What is more important is that,in the face of potential electricity accident,emergency plans shall be strategized in advance and ensure the event is delivered in an orderly manner.Scientific and orderly power emergency supplies will dispatch electricity emergency supplies accurately in a timely and efficiently manner.It also avoids the accidents caused by electricity power equipment problems or power blackout.Stable power supplies will prevent the negative impacts on unstable broadcasting connections,on the potential security risks of venues staffs and audience,and on the event sponsors.Take the Beijing 2022 Olympic Winter Games as an example.With the aim of predicting the electricity emergency supplies reserve,supplies reserve location layout,and electricity emergency supplies distribution.On the basis of a comprehensive literature review,the philosophy of "Prediction-Siting-Distribution" power emergency supplies dispatch was brought up.By using emergency supplies dispatch,XGBoost regression prediction theory,and linear programming theory.It builds XGBoost power emergency supplies reserve prediction model,multiple coverage and multi-level quality supplies reserve model and power emergency supplies accurate distribution model This addresses the problem of electricity power emergency supplies and improve the efficiency.The main research contents are as follows:(1)Power emergency supplies reserve prediction.This paper introduces the knowledge graph spectrum clustering algorithm,using the methodology of offline training prediction plus online real-time crawler data.It builds the XGBoost power emergency supplies reserve prediction model.This model,by using the large event data provided by the Olympic games,conducts data training,combined with the real scenarios of the Olympic winter games.This provides theoretical reference for the dispatch of electricity emergency supplies by predicting the power emergency supplies reserves.(2)Siting and the dispatch of the power emergency supplies reserve bank.Based on the power emergency supplies reserves forecast,for the multiple power emergency supplies reserve warehouse and the supply point demand,multiple coverage and multi-level quality supplies reserve Siting model is built.Aiming at the shortest response time,the maximum satisfaction of reserve warehouse coverage demanded,and the minimum cost,power emergency supplies reserve warehouse is combined to meet the demand of venues.Taking the Beijing 2022 Olympic winter games as a case study and the possibility of secondary allocation,introducing the solution method of multi-objective decision,through Lingo operating the scene description and conducting the model solution,and analyzing the results,the effectiveness of the model construction is verified.(3)Accurate dispatch of electricity power emergency supplies.Through quantitative analysis of large event power emergency supplies distribution model,the impacted venues and the demand of electricity power emergency supplies are predicted.The power emergency supplies reserve warehouse is selected according to the power emergency supplies for supply and demand distribution.Based on the supply and demand distribution,the electricity power emergency supplies are allocated for the first time.Based on the first allocation,and the latest supply and demand distribution model,the second distribution of power emergency supplies is completed.The effectiveness of the allocation model is proved by combining the scheme made by the allocation model with the actual situation of the Winter Olympics.The innovation of this paper is reflected in three aspects: First,the comprehensive dispatching mode of electric emergency materials for large-scale events is optimized.With the starting point of "prediction-site-distribution",the triple optimization objectives of maximizing the overall time benefit,maximizing the satisfaction degree and maximizing the utility of emergency materials are ensured.To a certain extent,the theory of electric emergency rescue for large-scale events is expanded and the connotation of comprehensive dispatching of electric emergency materials for large-scale events is enriched.Secondly,a reserve prediction method based on characteristic clustering of power emergency materials was proposed,and a XGBoost regression model based on knowledge graph clustering of power emergency materials was established,which solved the problems of data sparsity and cold start,effectively reduced the prediction deviation of traditional forecasting methods,and improved the accuracy of reserve prediction.It provides reference for emergency decision-making departments to formulate emergency resource allocation plan.Thirdly,multiple coverage and multi-level quality supplies reserve Siting model is constructed,which realizes the multilevel coverage of demand points by the storage warehouse,divide the level of supplies rescue according to the emergency situation,optimizes the traditional coverage mode,meet the needs of different emergency situations,which improves rescue efficiency and supplies usage rate,and provides theoretical support for the practical rescue of power emergency in large-scale events.In this paper,the integrated dispatching method of "prediction-site-allocation" for electric emergency supplies can provide reference for the power emergency repair and rescue work of large-scale events,and has certain theoretical and practical significance. |