Optimization of the distance-constrained multi-based multi-UAV routing problem with simulated annealing and local search-based matheuristic to detect forest fires: The case of Turkey

Abstract

Forests cover nearly a third of the Earth’s land area, and they are a key factor for all life on Earth, but unfortunately, forest fires are the greatest danger to their presence. The wildfires jeopardize general wellbeing, security, and require high levels of government resources. They also lead to noteworthy debasement of nature, property loss, and high rates of human death and injury. This paper proposes an algorithm to use and route unmanned aerial vehicles (UAVs) to mitigate forest fire risks. The developed matheuristic algorithm hybridizes simulated annealing and local search metaheuristics with an integer linear programming model. The mathematical model was developed to solve the distance-constrained multi-based multi-UAV routing problem, and because of the complexity of the problem, the generated metaheuristics helps the model to find better solutions. The effectiveness of the proposed matheuristic is tested with a real-life case study for Turkey and is also compared with a genetic algorithm. The Turkish State Meteorological Service generates forest fire-risk maps countrywide every day to predict fire risks 3 days later by using meteorological data. These maps are used to generate the risky regions to be visited by the UAVs, and the existing airports are considered for the UAVs to take off and land. The algorithm is coded using MATLAB and ILOG. The metaheuristics are designed with problem-based operators, and their parameters are tuned by experiments. Computational results demonstrate the effectiveness of the algorithm and the hybridization procedures. Results demonstrate that the CPU times for the methods are acceptable.