Study Of Forecasting Technology For City Natural Gas Load In Medium And Long Term

Chinese natural gas market is one of the world’s fastest-growing markets.Chinese natural gas market has some characteristics: proven reserves, output andconsumption are growing rapidly, natural gas supply pattern is more diversely,natural gas infrastructure construction is developing fast, the consuming field isexpanding, and market demand exceeds supply, gas consumption structures becomemore reasonable etc.With establishment and imple ment of the12th Five-Year natural gas industrydevelopment goals, our country’s natural gas market faces new challenge. Theproduction capacity, transmission capacity, distribution capacity, peaking capacity,storage capacity of the gas industry have been asked to be more flexible, safer, morereliable and more effective. To achieve these we must know gas load first.Researching the technology of Medium and Long term Load Forecasting for UrbanNatural Gas is proposed to meet new challenges. The main research contentincludes research of the forecasting methods and models, establishment of database,development and design of forecasting software.First of all, this paper processed and analyzed the example A city’s formeryears’ historical data, and established the database of medium and long term loadforecasting software for urban natural gas, then analyzed all kinds of users’consumption law of natural gas and influence factors, used the least square supportvector machine, correlation analysis, load density method, non-uniform coeffic ientmethod, grey forecasting technology to established the medium and long t erm loadforecasting model, optimized selection and edited prediction module. Then, authoranalyzed the requirement of the software to give the system function module,interface and interface’s function. At last, author commissioned the software, andimproved the problems of the data calls，software interface. What’s more, authorchecked the calibration of existing data and analyzed the errors. From the results ofthe error analysis, the city gas medium term load forecasting model is in goodprediction accuracy, and the city gas long term load forecasting model is in highprecision accuracy, which can be applied in practical projects. Not only does thesoftware provide the basis for the development of the city gas storage plan, but alsotechnical support for the scheduling, management and ope ration of the enterprise.