Method for predicting short-period microgrid load power interval probability

Abstract

The invention discloses a method for predicting short-period microgrid load power interval probability. The method comprises the following steps: acquiring a plurality of history load power of a microgrid as a sample set; combining prediction interval coverage rate and prediction interval average bandwidth to construct an optimization criterion; establishing a short-period microgrid load interval probability prediction model based on an artificial bee colony recurrent neural network, and carrying out neural network weight threshold value searching and updating on the optimization criterion through an artificial bee colony algorithm; introducing trend operation in a bacteria foraging optimization algorithm into a bee following local search strategy, and introducing an optimal bee source guiding mechanism to carry out optimization on the artificial bee colony algorithm so as to improve the algorithm performance and quicken the rate of convergence. According to the method, through the improvement of the artificial bee colony algorithm, the disadvantage that a traditional artificial bee colony algorithm is slow in rate of convergence and low in precision is preferably overcome, and the load prediction level of the microgrid is effectively improved.

Description

technical field [0001] The invention belongs to the technical field of power systems, and in particular relates to a short-term micro-grid load power interval probability prediction method. Background technique [0002] A microgrid is a small decentralized system that connects distributed power sources, energy storage devices, energy conversion devices, and monitoring and protection devices to supply power to users. Accurate prediction of microgrid load is an important basis for microgrid operation and energy management, and will directly affect microgrid operation strategy. In the field of load forecasting, the current methods for microgrid load forecasting are mainly divided into traditional forecasting methods and modern intelligent forecasting methods. Traditional forecasting methods mainly include curve extrapolation method, gray forecasting method, regression analysis method, time series method and load derivation method, etc. This kind of method uses probability the...

AI Technical Summary

Problems solved by technology

[0003] At present, the load power prediction methods of microgrid are single-point prediction, only a certain value is given, and the possible fluctuation range of the prediction result cannot be determined.

Images