Aerial firefighting, agent water-dropping model, fixed-wing firefighting aircraft, genetic algorithms, water-dropping scheme, neural networks, Electrical engineering. Electronics. Nuclear engineering, TK1-9971

The water-dropping scheme plays an important role in the aerial firefighting mission. When a wildfire occurs, the main problem faced by the aircrew is how to choose the appropriate water-dropping scheme (speed, height and trajectory) to achieve the optimal firefighting effect. However, so far, the crew members can only rely on the existing experience to generate the scheme, which lacks the corresponding theory method support. To solve above problem, this paper proposes a fast optimization method of water-dropping scheme for fixed-wing firefighting aircraft. Through the feature extraction of grid distribution, the agent model between aircraft water-dropping scheme and water distribution polygon is built taking advantage of neural network. Meanwhile, the corresponding extinguishing efficiency evaluation model is established with discrete points of fire area. Based on these researches, the combined genetic algorithm is used for searching the optimal water-dropping scheme. Finally, the feasibility and rapidity of the method are demonstrated through the application of examples. It is also shown that the method can optimize the water-dropping scheme quickly instead of aircrew experience and play a role of auxiliary decision support for the firefighting mission of fixed-wing firefighting aircraft.