FOREST FIRE PREDICTION MODELING APPROACH BASED ON DETERMINATION OF FIRE SPOT DISTANCE THROUGH TEMPERATURE RATE USING MACHINE LEARNING AND MULTILATERATIONI

• Main Author: Amaliyah, Shafira
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/73153
• Institution: Institut Teknologi Bandung

According to Global Forest Watch, forest fires are becoming an increasingly complex problem related to climate change. Therefore, an effective solution is needed to minimize the risk of forest fires by predicting fire events early and accurately. Fire spot detection using wireless sensors and multilateration methods can be a cost-effective and efficient. This approach combines the concept of machine learning based on the rate of change of temperature over distance to estimate the distance between the sensor and the fire spot, and uses a multilateration method to determine the coordinates of the location of the fire spot. The results showed that the machine learning approach using a linear regression algorithm can predict the distance of the fire point to the sensor well in an area of 120 x 120 cm. In addition, the multilateration method is able to estimate the position of fire spot accurately using the least squares approach. This approach provides fairly good position estimation accuracy with an average offset distance of 2.09 cm indicating the potential of this approach as an early warning system for forest fires. For further research, a wider research area and more complex variations with a non-linear fitting approach or other machine learning algorithms can be used to improve the accuracy of forest fire detection.