The Model Of Innovation Diffusion Based On Cellual Automata And Using It In Natual Gas Consumption Predication

The estimation on natural gas consumption is very important for the management of production,layout,design,optimization and control of net pipeline's transportation.Thus,according to natural gas consumption trends,establishing an appropriate model to forecasting is very significance. Currently, forecast natural gas consumption always use the methods of regression analysis,ARMA model,gray prediction model and so on.These methods have many defects. For example: the prediction accuracy of regression analysis will be affected by the selection of factors, if introduction of other factors, it will not only increase the calculation, reducing the stability of the model, also easy to affect the prediction accuracy. gray prediction method's defect is it only fit to the original data which change not quickly, and the turning point occurred when the system changes, using the GM (1,1) model to predict will have low prediction accuracy. ARMA forecasting method's defect is that it will automatically to clear the random fluctuations of original data sequence, making the forecasting methods is not fit to predict the consumption have certain random fluctuations .By analyzing the trend of natural gas consumption, we can get that at first,its consumption increased slowly, then it will accelerat after being adopted, with the new energy innovation and adoption, at some point in the future natural gas consumption will reach saturation, the growth of consumption will become slowly, until it stops growing. This shows that the basic line with natural gas consumption fits the process of innovation diffusion, and foreign scholars have used innovation diffusion model to forecast natural gas consumption, such as GUSEO has used the Bass expansion model to forecast the United States, Norway and the world's consumption of natural gas and crude oil , and get a better prediction. Therefore, in order to predict more accurately the level of consumption of natural gas, natural gas consumption as this is a process of diffusion of innovation, based on cellular automata model of innovation diffusion to fit the natural gas consumption trends and projections. The model of Innovation Diffusion Based on Cellual Automata was expanded on the cellular automata model, and through the link between the extended model and a separate non-autonomous Riccati equation , can get a perturbation expansion of cellular automaton model. The model of Innovation Diffusion Based on Cellual Automata to predict the gas consumption has some advantages: 1) the perturbation functions and variables of the model have consider the uncertainties and the phenomenon of repeat purchase of natural gas consumption,it can forecast an uncertainty process of innovation diffusion ;2) this model considers the heterogeneity of consumers, the adoption of individual behavior and interaction mechanisms, it can reveal microscopic diffusion of innovation market mechanisms to overcome the defects of Bass model; 3) this model not only has closed form solutions, and can use macro-statistical data to do empirical research ,it overcomes the defects of complex systems which can only Computer simulation process and the results of technology innovation diffusion.This paper analyzes the status of China's natural gas consumption, the nesscery of forecasting natural gas consumption forecast and several commonly used approaches'defects; then introduced the innovation diffusion model and cellular automata model, and introduced the model of innovation diffusion based on cellual automata; Finally, using natural gas consumption to do empirical study, in empirical study firstly using regression analysis and gray prediction on natural gas consumption forecast, and then using the perturbation cellular automaton extended model to forecast it, and compared the results with the classic Bass model and GBM model. Finally, comparing the results of traditional forecasting methods with innovation diffusion model that forecast natural gas consumption , indicating that the model of innovation diffusion based on cellual automata is more suitable for the forecast of China's natural gas consumption.