Research on Intelligent Fault Diagnosis Sensor Data Fusion Algorithm in Avionics Maintenance

This paper presents a kind of intelligent fault diagnosis algorithm, which combines Kalman filter, particle filter and deep-learning technique, to integrate and analyze sensor data in avionic system in real time. Experimental results show that in the fault diagnosis of typical avionics systems (such as flight control systems), the diagnostic accuracy of the Kalman filter algorithm is 91.5%, the particle filter algorithm is 94.3%, and the fusion algorithm combined with deep learning achieves a fault diagnosis accuracy of 97.8%. In addition, the error of deep learning combined with data fusion algorithm in fault location is controlled within 3.2 meters, which reduces the positioning error by about 25% compared with traditional methods, and improves the computational efficiency by about 15%. This study proves the application effect of data fusion algorithm in avionics maintenance, demonstrates its advantages in improving fault diagnosis accuracy and real-time performance, and provides innovative ideas and methods for the future development of aviation maintenance technology.