Research On Traction Load Forecasting Method For High-speed Railway Traction Substation Based On Train Timetable

The High-Speed Railway(HSR)is one of the most significant loads of the utility power system.Typically,it has various characteristics,including high power,strong fluctuation,and massive energy consumption.The State Grid Corporation requires the railway company to predict the power consumption of each traction substation accurately for safe and stable operation and economic dispatch of the power system.Generally,there are two challenges to predict the traction load of the HSR accurately.In one aspect,there are many complicated conditions when the EMU is Running on the rails Among the power supply Interval of the Traction Substation(EMURAITS),and its power fluctuates frequently.Moreover,modern EMU is regenerative.All the features make it hard to simulate the dynamic power consumption of EMU.On the other hand,the China Railway Company adjusted the scheduled traintimetable frequently to meet the requirements of economic development.It makes the quantity and positions of the EMUs variable and complicated.Therefore,this paper established a load model library,which contains the dynamic power consumption of typical EMURAITS,using the measured data at the point of common coupling(PCC)of the traction station.Ultimately,it predicts the dynamic traction loads of the traction station based on the information in the scheduled train timetable,including the train trips,and marshaling.Generally,the main contents include the following four aspects:(1)Primarily,it summarized the typical EMURAITSs considering the factors as running direction,entering the station or not,and the EMU marshaling.Then,it introduced the field measurement scheme to obtain the field data.Ultimately,it analyzed the dynamic power consumption features of all the typical EMURAITSs with the field data.(2)It introduced a load detection method with sliding window technology,which detected the starting and ending moment of the EMURAITS using the active power fluctuation index.On this basis,it extracted the dynamic power consumption of each EMURAITS from the field data of the PCC.Consequently,it established a load library.Moreover,it calculated some typical indices,including the statistics,the regenerative electricity of each EMURAITS,as well as the similarity.Ultimately,it classified the EMURAITSs with the indices.(3)It found that there are some biases of the different models in one EMURAITS.It studied a unified modeling method using linear-regression technology and probability statistics technology.Moreover,it established the model of no-traction-load using the probability statistics technology.Furthermore,it verified the accuracy of the proposed method with case studies.(4)It studied a traction load forecasting method with the model of typical EMURAITSs and the no-traction-load condition and the information in the scheduled train timetable,including the train trips,departing time,stop station,and EMU’s marshaling.Moreover,it predicted the energy consumption demand of the high-speed railway traction substation in case studies,the results of the total power consumption and the regenerative electricity are within the error range,which verifies the accuracy of the traction load prediction method proposed in this paper.Ultmatly,it evaluated the power supply capacity of the traction power supply system under different scheduled train timetables.Finally,it summarized all the works of this paper,and the future research direction has prospected.