Electric power system load forecast model based on State Space time-varying parameter theory

Putting forward a new method of power system medium-long term load forecast model establishment based on state space time-varying parameter equation theory, this work brings state variable into model observable for solution, which reflects the varying rules of equilibrium relation of variables and enhances the veracity for model forecast. The State Space time-varying parameter model describes the variational regular of the equilibrium relation between variables and improves the forecast accuracy. Moreover, Due to putting state variables into the observation model, this model based on State Space time-varying parameter theory has more precision in power system load forecasting. Demonstration section uses the model of 1952-2001 Time Series with China Yearly Electricity Consumption and GDP to tests the effectiveness of the new method via forecast of 2002-2005 Yearly Electricity Consumption. The test result proves it to be effective.