It is aimed to estimate solar energy estimates with high accuracy by developing a new method and device. In this context, many methods have been presented and developed in the literature for solar energy production forecasts. However, none of these methods are generally accepted methods in the scientific community.
The superiority of each method over the other, that the methods address different geographic areas of application, is an indication that it is not the perfect solution for every environment condition. As it is known, some parts of the cloud directly in the region where the solar panels are shading between the panels and the sun, and production is expected to fall or fall, although the clouds do not block the sun as a result of the production to continue at the same level as a result of the calculation results in serious losses and cause serious losses. In addition to this material loss, it causes energy fluctuations, deterioration of system stability, wasting of excess energy, frequency mismatches, and occasional power outages as a result of voltage fluctuations. Although they are located in the same city, the production values of the solar panels located in different regions vary. This makes the accuracy of estimates very difficult for each plant.
In addition, energy production is not entirely dependent on atmospheric conditions. Many variables such as the life, structure and deformation coefficients of heating and cooling systems directly affect the energy to be produced. Therefore, the estimation of solar energy production, which depends on many factors and variables, is a big and challenging issue. According to the literature reviews, even the most successful predictions in real environment applications are not satisfactory. (The accuracy rate for Turkey is 60-65%). As a result of the meetings with TEİAŞ and Tegneta company, it is calculated that the accuracy rate in the estimations will directly affect the productivity, and even with a 5% improvement, it is calculated that the profit will be only 3 billion TL for our country. Considering the legislation and national standardization studies, production estimates are of great importance for distribution networks.
Considering all these difficulties and lost energy, a device that can monitor the sky regionally with thermal cameras and analyze the parameters such as temperature, humidity, pressure, wind, altitude, rain, DNI with the help of sensors from that point will be developed and the correct decision capacity of machine learning algorithms will be increased. Ensuring that the predictions are made by these algorithms will be the solution to all these problems. Moreover, instead of only one of these machine learning algorithms to make decisions, a method in which the most popular algorithms known are combined will be used to make common decisions. This will be a first in the GES systems. Because, contrary to all known methods, for the first time, a device will be developed that can monitor the sky with thermal imagers and associate it with energy production according to all state variables. The project will bring innovation both in terms of hardware and software and is also an internationally patented product.
As a result, it is aimed to be a project with high impact, integrated with industry and high added value to the country and it is foreseen that the outputs of the project will be guiding to other stakeholders, especially distribution system operators.