DECISION SUPPORT SYSTEM DEVELOPMENT FOR MONITORING AND PREVENTION OF FOREST FIRES IN UKRAINE

Abstract

The article presents the results of research and development of a decision support system for forest fire monitoring and prevention in Ukraine. The issue of forest fires on the territory of Ukraine was considered. Traditional methods of monitoring and recording forest fires were studied. Their imperfection was revealed. An overview of the literature dealing with this problem is given. The paper notes that traditional methods of forest fire monitoring and prevention, such as ground patrols and aerial surveillance, are often ineffective and expensive. The use of modern technologies, such as remote sensing of the Earth, artificial intelligence and decision support systems (DSS), can significantly improve the effectiveness of forest fire monitoring and prevention. That is why the problem of developing software that would ensure the detection of forest fires is extremely important. The study was carried out using images from the open sources of the NASA Earth Observatory. Python libraries: Keras, TensorFlow, PyTorch were used to process and analyze satellite images. With the help of modeling methods, the system architecture is designed and usage options are shown. The article describes the concept of DSS. During the development of the CSDP for forest fire monitoring and prevention in Ukraine, the needs of users were identified and a diagram of use cases was constructed. The main components of the created system were described and the architecture was developed. The working principles of the DSS algorithm for image analysis are considered. First, the image data is read and converted to a single format. Next, with the help of neural network training methods, you need to remove noise, segment the image by highlighting areas of fire, transform color, analyze boundaries and key points. Based on this, an image assessment is formed. The algorithm is repeated for several cycles to obtain more accurate prediction results. After primary image processing and analysis, a neural network learning algorithm was developed. As a result, a convolutional neural network model was built, the accuracy of which reaches 92%. The developed decision support system shows good results in fire detection in the early stages, but there are still opportunities for further improvement of the algorithm.