Fault diagnosis of flight control system based on BWO-VMD-DNCNN
Abstract The flight control system of commercial aircraft is the core system of the aircraft and is directly related to whether the aircraft can fly safely. There are difficulties in fault diagnosis of flight control systems, such as noise and complex signals in the collected signals. This paper pro...
| Published in: | Journal of Physics: Conference Series |
|---|---|
| Main Authors: | , , , , , , |
| Format: | Article in Journal/Newspaper |
| Language: | unknown |
| Published: |
IOP Publishing
2024
|
| Subjects: | |
| Online Access: | http://dx.doi.org/10.1088/1742-6596/2820/1/012072 https://iopscience.iop.org/article/10.1088/1742-6596/2820/1/012072 https://iopscience.iop.org/article/10.1088/1742-6596/2820/1/012072/pdf |
| Summary: | Abstract The flight control system of commercial aircraft is the core system of the aircraft and is directly related to whether the aircraft can fly safely. There are difficulties in fault diagnosis of flight control systems, such as noise and complex signals in the collected signals. This paper proposes a fault diagnosis method to solve this problem, using the Beluga Whale Optimization (BWO) and the denoise convolutional neural network (DNCNN) improved variational mode decomposition combined with the support vector machine optimized convolutional neural network. Firstly, the signal undergoes decomposition using variational mode decomposition, followed by the introduction of Beluga Whale Optimization. The minimum envelope entropy serves as the fitness function, and the determination of the number of decomposition layers and quadratic level of variational mode decomposition relies on the fitness value, incorporating a penalty factor. Secondly, enhancements are made to the denoising convolutional neural network to adapt it for one-dimensional signal denoising. The denoising convolutional neural network is used to denoise each mode after variational mode decomposition, and the denoised modes are reconstructed to obtain the denoised signal. Finally, the convolutional neural network (CNN) is used to extract data features, and the support vector machine (SVM) is used to replace the Softmax classifier in the convolutional neural network to realize fault diagnosis of the flight control system. According to the results of the experiment, the method is universally applicable, demonstrating strong diagnostic ability and high diagnostic accuracy. |
|---|