BP neural network photovoltaic power station generating capacity prediction method based on genetic algorithm

Abstract

The invention discloses a BP neural network photovoltaic power station generating capacity prediction method based on a genetic algorithm. The method comprises the steps of firstly performing data sample acquisition and preprocessing, then constructing a BP neural network structure, optimizing the network by means of the genetic algorithm, and introducing an optimal individual after optimization into the BP neural network for prediction. According to the BP neural network photovoltaic power station generating capacity prediction method, the genetic algorithm is introduced for optimizing the BP neural network; the initial weight threshold of the network is represented by the individual; prediction error of the BP neural network with initialized individual value is used as fitness value of the individual; the optimal individual is searched through selection, interaction and variation operation methods in the genetic algorithm, so that the network reaches an optimal initial weight threshold, thereby improving final prediction precision.

Description

technical field [0001] The invention relates to a genetic algorithm-based BP neural network photovoltaic power station generating capacity forecasting method, which belongs to the technical field of photovoltaic power plant generating capacity forecasting Background technique [0002] In the 21st century, science and technology are changing with each passing day, and the photovoltaic industry has also developed rapidly. Today, photovoltaic systems are moving toward automation, networking, and intelligence. The prediction of the power generation of photovoltaic systems greatly fits this theme. As we all know, photovoltaic power generation is easily affected by external factors, and there are many influencing factors. The safety is reduced. When the proportion of photovoltaic power generation in the entire power grid is relatively small, the above-mentioned impact may not be obvious. Once the proportion of photovoltaic power generation in the entire power grid exceeds a certai...

AI Technical Summary

Problems solved by technology

As we all know, photovoltaic power generation is easily affected by external factors, and there are many influencing factors. The safety is reduced. When the proportion of photovoltaic power generation in the entire power grid is relatively small, the above-mentioned impact may not be obvious. Once the proportion of photovoltaic power generation in the entire power grid exceeds a certain range, grid-connected photovoltaic power plants will have a negative impact on the power system. It will have a great impact on the safe and stable operation of the power grid, and the technology of predicting the power generation of photovoltaic power plants has emerged as the times require.

However, the domestic photovoltaic power plant power generation forecasting technology is not mature at present, and the forecasting error is relatively large, some of which are as high as 30%.

[0003] BP neural network is a multi-layer feed-forward neural network, which is one of the most popular algorithms at present. The essence of BP algorithm is to solve the problem of the minimum value of the error function. The direction correction weight threshold has the characteristics of simplicity, ease of operation, small amount of calculation, and strong parallelism. The power generation forecasting system of photovoltaic power plants is a nonlinear system with changing factors. Very applicable, however, the initial weight threshold of the BP algorithm is generally a random value in the (-1,1) interval assigned by the system. Whether the BP algorithm reaches the local minimum during the learning process, the length of the training time, whether the training is convergent, and the final prediction accuracy are related to The adjustment of the weight threshold has a lot to do with it. The insufficient fitness of the initial weight threshold will lead to low prediction accuracy.

Images