Solar power generation prediction system using artificial intelligence

Abstract

The present invention relates to a solar power generation prediction system using artificial intelligence, and more specifically, to a solar power generation prediction system using artificial intelligence capable of predicting power generation by collecting various data related to solar power generation and systematically applying various analysis techniques, including multi-complex artificial neural network analysis, based on the collected data. The solar power generation prediction system using artificial intelligence according to the present invention performs performance evaluation, weight determination, and impact assessment of influencing factors whenever an analysis is performed, and accordingly, reflects the reset weights or impact information of the influencing factors in subsequent analyses and can provide them to administrators or users. In addition, the solar power generation prediction system using artificial intelligence according to the present invention can improve the convenience of providing various visual or auditory information, etc., to managers or users through an information provider, regarding expected power generation information. In addition, the solar power generation prediction system using artificial intelligence according to the present invention has the advantage of improving accuracy in parallel execution with multi-complex artificial neural network power generation prediction, by performing prediction analysis processing using other analysis methods such as degradation rate prediction, regression analysis power generation prediction, power generation prediction through statistical models, and deep artificial neural network analysis.

Description

Technology Field

[0001] The present invention relates to a solar power generation prediction system using artificial intelligence, and more specifically, to a solar power generation prediction system using artificial intelligence capable of predicting power generation by collecting various data related to solar power generation and systematically applying various analysis techniques, including multi-complex artificial neural network analysis, based on the collected data. Background Technology

[0002] Global electricity consumption is continuously increasing. However, as the amount of fossil fuel-based energy resources is limited, there is a trend toward increasing the share of electricity production from new and renewable energy sources such as solar and wind power.

[0003] Producing electricity using such new and renewable energy sources is becoming increasingly important as it not only reduces environmental pollutants such as greenhouse gases and fine dust but also enhances energy supply stability from the perspective of sustainable future energy.

[0004] However, since solar power generation depends on various parameters such as solar radiation, temperature, atmospheric humidity, and clouds depending on the specific location and time, generated power is intermittent and very difficult to predict. Furthermore, solar power output fluctuates throughout the day due to external factors such as dust, clouds, and snow, and is also significantly affected by seasonal factors. To mitigate this uncertainty in solar power output, enhance the reliability of the power grid, and improve power quality, it is necessary to accurately predict solar power output. Prior art literature

[0005] Republic of Korea Published Patent 10-2020-0057821 Republic of Korea Published Patent 10-2021-0147366 Republic of Korea Published Patent 10-2021-0043388 The problem to be solved

[0006] The present invention aims to solve the aforementioned conventional problems by collecting weather data module temperature and data provided by the Korea Meteorological Administration, as well as forecast data, visibility data, and precipitation data from current solar power plant area weather data information; verifying the existence of each data for 24 hours a day; constructing a training data set after applying interpolation to empty values ​​to secure time-series characteristics; grouping the data set, which was previously grouped by the Korea Meteorological Administration forecast data in 1-hour units, into 3-hour averages; and providing a solar power generation prediction system using artificial intelligence that can predict the next day's power generation by applying multi-model deep learning composed of RNN and CNN, which are deep learning techniques, to the data for each power plant.

[0007] In addition, the present invention aims to provide a solar power generation prediction system using artificial intelligence capable of predicting power generation by collecting various data related to solar power generation and systematically applying various analysis techniques, including multi-complex artificial neural network analysis, based on the collected data. means of solving the problem

[0008] A solar power generation prediction system using artificial intelligence according to the present invention for achieving the above-mentioned purpose comprises: a data collection unit that collects inverter data of a solar power generation facility and weather data measured by the Korea Meteorological Administration in real time; a signal measurement unit that measures a signal including SV voltage, SA voltage, R current, S current, T current, EFF, and current power generation amount; a customer database in which signal data measured by the signal measurement unit is stored; a degradation rate prediction unit that calculates the annual degradation rate of the solar power generation facility and predicts the degradation rate of the solar power generation facility for the relevant year; a data preprocessing unit that preprocesses the signal data measured by the signal measurement unit and the degradation rate data predicted by the degradation rate prediction unit; a first smart mind database in which data processed by the data preprocessing unit and forecast data from the Korea Meteorological Administration are stored; a prediction module that predicts power generation amount based on data stored in the first smart mind database; and a second smart mind database in which result data predicted by the prediction module is stored. It is characterized by having a monitoring unit that monitors whether there are any abnormalities in the result data stored in the second smart mind database.

[0009] The degradation rate prediction unit of the solar power generation prediction system using artificial intelligence according to the present invention is characterized by learning data from the same month of one year and two years ago to predict the current power generation and annual degradation rate, and learning data within one month to diagnose abnormalities and degradation. Effects of the invention

[0010] The solar power generation prediction system using artificial intelligence according to the present invention performs performance evaluation, weight determination, and impact assessment of influencing factors whenever an analysis is performed, and accordingly, reflects the reset weights or impact information of the influencing factors in subsequent analyses and can provide them to administrators or users.

[0011] In addition, the solar power generation prediction system using artificial intelligence according to the present invention can improve the convenience of providing various visual or auditory information, etc., to managers or users through an information provider, regarding expected power generation information.

[0012] In addition, the solar power generation prediction system using artificial intelligence according to the present invention has the advantage of improving accuracy in parallel execution with multi-complex artificial neural network power generation prediction, performing prediction analysis processing using other analysis methods such as equipment degradation rate prediction, regression analysis power generation prediction, power generation prediction through statistical models, and deep artificial neural network analysis. Brief explanation of the drawing

[0013] FIG. 1 is a block diagram showing a solar power generation prediction system using artificial intelligence according to the present invention. Specific details for implementing the invention

[0014] Hereinafter, a solar power generation prediction system using artificial intelligence according to a preferred embodiment of the present invention will be described in detail with reference to the attached drawings.

[0015] FIG. 1 illustrates a solar power generation prediction system using artificial intelligence according to the present invention. Referring to FIG. 1, the solar power generation prediction system using artificial intelligence according to the present invention comprises: a data collection unit (10) that collects inverter data of a solar power generation facility and weather data measured by a meteorological agency in real time; a signal measurement unit (20) that measures a signal including SV voltage, SA voltage, R current, S current, T current, EFF, and current power generation amount; a customer database (30) in which signal data measured by the signal measurement unit (20) is stored; a degradation rate prediction unit (40) that calculates the current power generation amount and annual degradation rate through learning data of the same month from one month, one year, and two years ago of the solar power generation facility, and predicts the degradation rate for the current year of the solar power generation facility; and a data preprocessing unit (50) that preprocesses the signal data measured by the signal measurement unit (20) and the degradation rate data predicted by the degradation rate prediction unit (40). The system comprises: a first smart mind database (70) in which data processed by the data preprocessing unit (50) and forecast data (60) from the meteorological agency are stored; a prediction module (80) that predicts power generation based on the data stored in the first smart mind database (70); a second smart mind database (90) that stores result data predicted by the prediction module (80); and a monitoring unit (100) that monitors whether there is an anomaly in the result data stored in the second smart mind database (90).

[0016] In addition, the degradation rate prediction unit (40) of the solar power generation prediction system using artificial intelligence according to the present invention learns data from the same month of one year ago and two years ago to predict the current power generation and annual degradation rate, and learns data within one month to diagnose abnormalities and degradation of the equipment.

[0017] The solar power generation prediction system using artificial intelligence according to the present invention collects weather data module temperature and data provided by the Korea Meteorological Administration, as well as forecast data, visibility data, and precipitation data among current solar power plant area weather data information, confirms the existence of each data for 24 hours a day, and constructs a training data set after applying interpolation to empty values ​​to secure time series characteristics.

[0018] In addition, the data set, which was grouped into 1-hour units of weather forecast data, is grouped into 3-hour averages, and the next day's power generation can be predicted by applying multi-model deep learning, consisting of RNN and CNN, which are deep learning techniques, to the data for each power plant.

[0019] In addition, the present invention aims to provide a solar power generation prediction system using artificial intelligence capable of predicting power generation by collecting various data related to solar power generation and systematically applying various analysis techniques, including multi-complex artificial neural network analysis, based on the collected data.

[0020] The solar power generation prediction system using artificial intelligence according to the present invention operates as a system composed of a plurality of devices, each of which is independent or interconnected, constituting the solar power generation prediction system using artificial intelligence according to the present invention.

[0021] In addition, the solar power generation prediction system using artificial intelligence according to the present invention may further include an analysis processing unit through regression analysis to perform machine learning through regression analysis.

[0022] The solar power generation prediction system using artificial intelligence according to the present invention, as described above, performs performance evaluation, weight determination, and influence assessment of influencing factors whenever an analysis is performed, and accordingly, the reset weights or influence information of the influencing factors can be reflected in subsequent analyses and provided to administrators or users.

[0023] In addition, the solar power generation prediction system using artificial intelligence according to the present invention can improve the convenience of providing various visual or auditory information, etc., to managers or users through an information provider, regarding expected power generation information.

[0024] In addition, the solar power generation prediction system using artificial intelligence according to the present invention has the advantage of being able to improve the accuracy of prediction analysis processing using other analysis methods such as annual degradation rate prediction of facilities, regression analysis power generation prediction, power generation prediction through statistical models, and deep artificial neural network analysis, as well as parallel execution of multi-complex artificial neural network power generation prediction.

[0025] The solar power generation prediction system using artificial intelligence according to the present invention described above has been explained with reference to the attached drawings, but this is merely illustrative, and those skilled in the art will understand that various modifications and equivalent alternative embodiments are possible therefrom.

[0026] Therefore, the true scope of technical protection of the present invention should be determined solely by the technical concept of the appended claims.

Claims

Claim 1 A data collection unit that collects inverter data of a photovoltaic power generation facility and weather data measured by the Korea Meteorological Administration in real time; a signal measurement unit that measures signals including SV voltage, SA voltage, R current, S current, T current, EFF, and current power generation amount; a customer database in which signal data measured by the signal measurement unit is stored; a degradation rate prediction unit that calculates the annual degradation rate of the photovoltaic power generation facility and predicts the degradation rate for the relevant year of the photovoltaic power generation facility; a data preprocessing unit that preprocesses signal data measured by the signal measurement unit and degradation rate data predicted by the degradation rate prediction unit; a first smart mind database in which data processed by the data preprocessing unit and forecast data from the Korea Meteorological Administration are stored; a prediction module that predicts power generation amount based on data stored in the first smart mind database; a second smart mind database that stores result data predicted by the prediction module; and a monitoring unit that monitors whether there are any anomalies in the result data stored in the second smart mind database; wherein the degradation rate prediction unit is one year ago and two years ago A solar power generation prediction system using artificial intelligence characterized by learning data from the same month to predict current power generation and annual degradation rate, and learning data within one month to diagnose abnormalities and degradation. Claim 2 delete

Images