Information Entropy Model Of Energy Consumption Per GDP And Its Analysis

Energy consumption per GDP has been one of the most important indicatorswhich measure economic development of China and listed in National11th five-yearplan planning target. At a macro-level, this index mainly describes the dependence ofeconomic and social development for energy consumption of one country (or region).Intense research on energy consumption per GDP can enforce to complete variouswork of energy during National12th five-year better, increase the efforts to work ofenergy conservation and emission reduction, enhance energy efficiency of China,build a safe energy system and development strategy, address global climate change,and service for the country’s social and economic construction better.This research started from the concept of “entropy” claimed in1854by RudolfClausius, a German physicist. Then we reviewed the history of information entropyand tracked back to the origin of the maximum information entropy principle and itsdevelopment and application. Through analyzing treatment of agents or indexes,obtain of weights and simulation methods of model, this research has chosen averagedimensionless method for agents’ treatment, neural network method for weights’obtain and self-organizing feature map method for simulation as foundation forstructuring and solving the theoretical model. Basing on the analysis of energyconsumption, gross national product and energy consumption per GDP of China, thisresearch has treated Chinese industry system as an informational network withcharacters of complexity and openness. Through the structure of information entropyand simulation of SOFM nerve net, we have obtained the wave diagram ofcharacteristic value ξ_kwhich could characterize Chinese industrial energyconsumption per GDP, the configuration diagram of ξ_k, the correlation between ξ_kandxi, as well as dynamic trend chart of weights aki. At last, the model has been verifiedand the influencing factors of energy consumption per GDP have been analyzed.Basing on the theory of maximum information entropy principle, this researchhas introduced information entropy into the research of energy consumption per GDPcreatively, which has expanded the application of maximum information entropyprinciple. The case of Chinese industrial system has reflected that, corresponding withtraditional statistical methods, this method can not only reveal interactions amongindustries better, but also reflect the formative progress of characteristic values ξ_kof energy consumption per GDP of Chinese industrial system in each year, relative majorindustries in each period, development pattern of each industry etc. The research hasprovided a new idea to energy research from the perspective of complex open systemby the application.