Research On Forecasting Method For Regional Grid Saturated Power Consumption

With the evovlement of modernization and urbanization in China, energy demand is rising quickly. Resource restriction and power demand, development of power grid and economy face with great challenge. Saturated power consumption is a pivotal indicator for power network planning in regional power grid. Through the analysis and forecast for saturated power consumption, important saturated power information is provided for regional power grid, which could provide guideness for the power grid planning and construction.Owing to the rare definitions of saturated power consumption, in this thesis, electricity, society and economy characteristics in developed countries and regions are firstly concluded based on energy development history. The definition of saturated power consumption is then clearly formulated. Judgment index system for saturation occurrence is presented to quantify the judgment indicators through generalization of foreign developed regions.Due to the success of Solow model in economic problems, it is also introduced in this thesis to yield some useful conclusion. GDP and population are used to represent the economies of scale; Industrial structure and power consumption structure are used to represent the structural effects, and the progress of technology is also applied to Solow long-run equilibrium model. From the experiences of Solow extended model, panel data model among area to forecast and foreign developed regions could be built to have an overall consideration of all countries with different period.End of this thesis, Shanghai is selected as the subject for future saturated power consumption analysis. Time of saturated occurrence is studied according to judgment index system by cyclical fluctuating Gompertz model and isomorphism analysis. While saturated power consumption is calculated and verified from the Solow long-run equilibrium model and panel data model, which gives out influential factors and other forecasting index. Comparing results could reveal it a beneficial exploration of saturated load forecasting work in new situation.