Based On Chaos Theory Urban Natural Gas Daily Load Forecast

With word's rapid economic development, people's living quality gradual increasing energy consumption's continual grouth,developing gas industry is the most best method to improve the environment and maintain sustainable economic development. City has become the main natural gas consumption,to accurately predict short-term natural gas consumption in which in real time is the important part and means of ensuring the effective development and operation of natural gas supply system,and benefits optimization, engineering analysis, management of natural gas pipeline network and economic, security and reliable operation of supply system.For some city's natural gas consumption,the article has given the research on phase space reconstruction, chaos discriminationbased and consumption prediction of which,based on the domestic and international research and application of chaos theory with the actual engineering.1. Selecting depolarization multiple autocorrelation function combining autocorrelation function and G-P algorithm as the best method on the detailed description of the means to choose the best reconstruction parameters,and to improve the classic G-P algorithm through introducing time window, balance offset factor and amendment of critical distance.2. There is detailed description of the qualitative and quantitative means for chaotic identification,among which, saturated correlation dimension and maximum Lyapunov index are the most important indicators,and the more authoritative small amount of data calculation method is used to calculate maximum Lyapunov index.3. Maximum Lyapunov index prediction model, improved weighted first order local prediction model, arithmetic mean combination prediction model and combination prediction model based on information entropy are selected to predict some city's natural gas consumption after describing reconstruction parameters, chaotic identification and leading time window and contribution rate of cumulative distance into weighted first order local prediction model,then some indexes for evaluating predictive results are list.4. There is research and discussion about some city's daily natural gas consumption based on the detailed introduction and analysis about chaotic theory, in the end,the combination prediction model based on information entropy is considered to be the best model.