Method for optimizing parameters of load prediction algorithm based on Gaussian liquid level method

Abstract

The invention discloses a method for optimizing parameters of load prediction algorithm based on a Gaussian liquid level method in the field of load prediction of a power system. The technical scheme comprises the following steps of firstly, initializing a Gaussian liquid level, i.e. parameters, such as queue length, number of dimensions and range; secondly, generating a random parameter through the Gaussian liquid level, and performing load prediction calculation; thirdly, evaluating a comparison result corresponding to the random parameter and an output result corresponding to an optimal parameter, and determining whether to update the optimal parameter or not; and finally, judging whether the Gaussian liquid level is converged or not, updating or initializing the Gaussian liquid level, and performing load prediction of the next time. The Gaussian liquid level which is continuously updated in a load prediction process serves as a probability density function for generating the random parameter, the method is clear in concept, clear in thought and low in calculated amount, and parameter optimization and prediction precision improvement of a super-short period load prediction algorithm of the power system can be realized.

Description

technical field [0001] The invention belongs to the field of electric power system load forecasting, in particular to a parameter optimization method of a load forecasting algorithm based on the Gaussian liquid level method. Background technique [0002] Ultra-short-term load forecasting is an important part of the energy management system. Online real-time load forecasting with a step size of 5 minutes to 1 hour is required. The forecast results are mainly used for: system economic dispatch, preventive control, and advanced control of automatic power generation control etc. At the same time, accurate ultra-short-term load forecasting helps to improve the operating efficiency and reliability of the microgrid. [0003] How to optimize the parameters of the load forecasting algorithm and improve the forecasting accuracy is the key technology for ultra-short-term load forecasting. At present, a large number of parameter optimization algorithms have been used for load forecast...

AI Technical Summary

Problems solved by technology

[0004] In view of the current power system mentioned in the background technology, the ultra-short-term load forecasting algorithm takes a long time to optimize parameters and has poor adaptability to rapidly changing load characteristics. The present invention proposes a load forecasting based on the Gaussian liquid level method Algorithm parameter optimization method

Images