Fault Diagnosis Method of Aircraft Anti-skid Brake System Based on GWO-PNN

The health of the aircraft's anti-skid braking system is critical to the flight of the aircraft. The failure of the aircraft's anti-skid brake will affect the system efficiency and flight safety. Therefore, this paper proposes an aircraft anti-skid brake system fault diagnosis method based on Grey Wolf Optimizer (GWO) and Probabilistic Neural Networks (PNN). First, preprocess the acquired data of five parameters of a certain type of aircraft anti-skid brake system: wheel speed, aircraft speed, braking time, brake pressure, and brake servo valve control current to create a PNN fault diagnosis model. Aiming at the shortcomings of PNN network smoothing factor selection based on experience, the GWO optimization algorithm is proposed to optimize the PNN network to find the optimal smoothing factor. The validity of the optimal GWO-PNN fault diagnosis model created is verified by experiments with the relevant parameter data of the brake system. The results show that the GWO-PNN-based aircraft anti-skid brake system fault diagnosis method proposed in this paper can effectively solve the problem of poor fault diagnosis effect caused by the artificial setting of the smoothing factor of the probabilistic neural network, avoiding the interference and influence of human factors. It has a good fault diagnosis performance. The fault diagnosis accuracy rate of the GWO-PNN model is as high as 95%, which is better than the fault diagnosis of the PNN and Back propagation (BP) diagnostic models.