A method and system for intelligent location selection of distributed photovoltaic power plants

Abstract

The application provides a method and system for intelligent location selection of distributed photovoltaic power plants, which belong to the technical field of prediction or optimization. The scheme includes: fine-grained grid division of pre-selected areas; Text semantic information and numerical data information, and based on the pre-trained deep learning model, respectively perform feature extraction; based on the extracted features and pre-trained multi-core learning model, obtain the evaluation scores of all grid positions under fine grid division in the region value, based on the evaluation score to provide decision-making basis for intelligent site selection; the scheme is based on the relevant data of existing power stations in the area, using time-space correlation to predict power generation at its adjacent locations, and further combining other texts that affect photovoltaic power generation Semantic information and numerical information are used to comprehensively evaluate all grid positions under grid division, ensuring the accuracy and rationality of intelligent site selection for photovoltaic power plants.

Description

technical field [0001] The application belongs to the technical field of prediction or optimization, and in particular relates to a method and system for intelligent location selection of distributed photovoltaic power plants. Background technique [0002] Photovoltaic power generation, as a kind of green energy that is more convenient to develop, is attracting more and more attention. However, in order to optimize the site selection and layout of photovoltaic power stations, ensure the reasonable system configuration of distributed photovoltaic power stations, improve the efficiency of photoelectric conversion, and reduce the cost of photovoltaic power generation, photovoltaic power generation The evaluation and site selection research of the power station is particularly important. [0003] However, the current site selection of photovoltaic power plants still has the following problems: [0004] (1) The current site selection of photovoltaic power plants is generally des...

AI Technical Summary

Problems solved by technology

[0004] (1) The current site selection of photovoltaic power plants is generally described as an optimization problem, such as the minimum annual operating cost, the minimum voltage deviation, etc. All of the above are NP-hard problems, and the solution process is easy to fall into local optimum;

[0005] (2) The power generation prediction of photovoltaic power plants is the key to site selection, but the current research on power generation forecasting generally predicts power generation based on the historical power generation time series data of a certain power station combined with relevant factors that affect power generation; the main problems It is manifested in two aspects: one is that it can only be predicted for one or more specific existing power stations; the other is that for candidate sites, if there is no historical power generation data, effective evaluation cannot be carried out;

[0006] (3) In addition to the historical time series data and forecast time series data of power generation, solar radiation, planning and control conditions, transportation conditions, grid access methods, natural resources and environmental conditions, social economy, operation and maintenance costs, operation and maintenance process, etc. Information is also an important factor in site selection, and the existing methods have not considered it effectively. At the same time, the existing evaluation methods include analytic hierarchy process, gray relational analysis, fuzzy evaluation, etc. The main shortcomings of the above evaluation methods are manifested in two aspects First, the weight setting of each influencing factor is relatively subjective and lacks scientific and objective quantitative basis; Semantic understanding of text among the important factors of URL

[0007] (4) Most of the existing photovoltaic site selection methods are considered from the above-mentioned single view (such as any one of power generation time series information, text semantic information, and numerical data information). The simple combination of all viewport features makes it impossible to make full use of multi-viewport features

[0008] (5) In the existing related research, most of the predictions or evaluations are made on the existing power stations, but in a certain area, the power stations that have already generated power are relatively sparse compared to the entire area. Overall distributed power station planning, the current research lacks a comprehensive and comparative evaluation of all grid locations under fine-grained grid division

Images