Prediction method for power generating system output power of grid-connected type photovoltaic power station

Abstract

The invention discloses a prediction method for power generating system output power of a grid-connected type photovoltaic power station in the technical field of photovoltaic power generation. The prediction method comprises the following steps of: firstly constructing a sample set: denoising the sample set to obtain the sample set after denoising; then establishing a kernel matching pursuit training model by utilizing the sample set after denoising; afterwards optimizing a coefficient of the kernel matching pursuit training model by utilizing a harmony search algorithm; and finally, predicting the generating power of the photovoltaic power station by utilizing the kernel matching pursuit training model with the optimized coefficient. The prediction method avoids the local optimum phenomenon to enable a prediction model to be accurate, and has very good nonlinear processing capability and high-dimensional data processing capability.

Description

technical field [0001] The invention belongs to the technical field of photovoltaic power generation, and in particular relates to a method for predicting output power of a grid-connected photovoltaic power station power generation system. Background technique [0002] Solar photovoltaic power generation is an energy source with low energy density, poor stability, and poor regulation ability. The power generation is greatly affected by weather and region. In some cases, photovoltaic grid-connected power generation will have a series of impacts on the existing power generation mode and the stable, economical, safe operation and power supply quality of the power grid, such as the impact of load peaks and valleys on the power grid, day and night changes, time differences between east and west, and seasonal changes The impact on the power grid, the impact of changes in weather conditions on the power grid, the impact of long-distance photovoltaic power transmission, etc. [000...

AI Technical Summary

Problems solved by technology

The time series method has a small amount of calculation and a fast speed, and the prediction effect is good when the weather changes little, but when the factors affecting the power generation change greatly, such as temperature, light intensity, etc., it is difficult to reflect the relationship between the force and these variables. Due to the dynamic and non-linear relationship, the prediction effect on complex and changeable photovoltaic power generation is poor.

Expert systems can avoid complex numerical calculations, but they have poor versatility and lack of learning ability

The neural network algorithm is only a model established based on the principle of empirical risk minimization, the convergence speed is slow, and it may converge to a local minimum point, knowledge expression is difficult, it is difficult to make full use of the experience knowledge of the dispatcher, and it requires a long training time

Support vector machines also have some obvious shortcomings, mainly because support vector machines need to solve a quadratic programming problem in the training process, which makes it have high time complexity and space complexity for large-scale sample collections. In order to reduce To solve the time and space complexity of quadratic programming, the optimization problem needs to be decomposed into several sub-problems

Algorithms for solving these subproblems usually obtain suboptimal solutions, thus reducing the generalization ability of SVMs

Images