Research On Information Fusion Algorithm Of Autonomous Anticollision UAV Sensor

UAV shows outstanding advantages and combat effectiveness in modern battlefield, thatmakes all the countries in the world which has appeared the climax in its development andapplication, and the application field has been extended from military to civil area, such asterrain exploration, public security fire, communication etc, and its using features of full timeand multi-airspace can not be matched by manned aircraft. But, with the widely application ofthe UAV, the security problem also gradually revealed, according to a survey, When flying atlow altitude, the UAV is susceptible to the irregular terrain of mountain and canyon, thusresulting in accidents fell to the ground, and it has become the research focus that how tocontrol the UAV autonomous collision.In this paper, took the multi-sensor system of UAV as the starting point, through researchand analysis on multi-sensor measurement technology, concluded that the height and pitchparameters was the important parameters when UAV flying close to ground, and as a clues tofurther studied the measurement principle and error sources of height and pitch angle sensors,and proposed the program of information fusion and collision avoidance of UAV multiplesensors on the basis of in-depth researched on sensor information fusion theory. And throughin-depth studied the algorithm of multi-sensor information fusion, designed fusion algorithmof kalman filter and adaptive weighted average for height parameters of UAV, and designedthe fusion algorithm of BP and RBF neural network to realize the fusion of pitch angleparameters of UAV, and through the simulation to contrast and verify the validity of thedesigned fusion algorithm.Finally, this paper combined the data of accuracy height and pitch angle of UAV afterfusion, and designed the path of UAV autonomous anti-collision with high slope as anexample, so as to realize the efficiency of UAV autonomous anti-collision.