Abstract: The flight control principle of four-axis vehicle was studied and the quaternion kinematics differential equation was established. Through the first-order RungeKutta method, the equation was solved. The real-time quaternion reversal Euler angle was used to represent the flight attitude information, so the attitude process could be controlled easily.The real-time attitude data was collected by the inertial sensor MPU9250, and the actual flight attitude was obtained by Mahony complementary filtering algorithm fusion.The flight control of micro four-axis vehicle was realized by cascade PID control algorithm. Finally, the simulation parameters of the cascade PID were adjusted by MATLAB software, and the actual debugging on the inner diagonal velocity loop and outer Euler angle loop of the main control aircraft was carried out on STM32F411. By equipped with intelligent flame recognition camera, the aircraft can automatically lock the key fire areas during area search and monitoring, and improve the detection efficiency.

Key words: fire, detection and monitoring, aircraft, quaternion method, Mahony complementary filtering, Rungekutta method