Application of Fuzzy Logic to develop a wildland fire-predicting UAS

Abstract

While development of wildland firefighting tools has focused on better on-spot attack equipment, fire-managers still suffer large number of loss of lives of personnel. Many incidents have recorded the deaths of several firemen during air-drops or on-field action. Studies have shown that this is caused to a large extent by the unavailability of accurate information about the growth of fire. Enhanced situational awareness tools can help improve proper planning of such activities and decrease response-time. The key is to develop a real-time and dynamic fire-prediction model that can clearly depict the perimeter of affected area during the expected response time. Mathematical models involve the tedious manual classification of the vegetation cover into select models for different behavior with fire ignition. Following is an attempt to imply tools of Fuzzy logic to make this process less consuming. Land cover data has been taken from national agencies and then categorized into prescribed models using a fuzzy-logic based decision making tool.