With the increased scope of fires and the widespread destruction of the environment and densely populated urban areas in recent years, researchers have investigated the adoption of rapid and effective firefighting solutions, particularly those based on wireless sensor networks (WSNs). In fact, by evaluating various statistical data and developing a new model of sensors, equipment, and intelligent technologies in a fire sensor network, an effective step toward controlling frequent fires on a wide scale and reducing environmental damage can be taken. In the proposed model, the mobile sensors or firefighting robots based on a fuzzy Q-learning (FQL) algorithm and using two learning strategies in the sensor network, namely partial and perfect, could be used to surround fire in firefighting operations. We also formulate a sensor mode selection strategy as an optimization problem to maximize the lifetime of the energy harvesting-enabled WSN. Furthermore, we determine optimal upper and lower bounds for the number of active sensors in the fire detection system, guaranteeing that the target detection and false alarm probabilities are achieved. Computer simulations show that using such a solution in the optimal selection of moving sensors and determining the moving trajectory in rapid firefighting is effective.
