Enhanced feedback analysis of vertical load reliability parameters for airplane landing gear using an improved generative adversarial network and explainable artificial intelligence techniques

Effective feedback analysis of critical equipment data is essential for improving performance and optimizing design parameters in aviation systems. This study presents a novel framework that integrates an improved generative adversarial network (GAN) with explainable artificial intelligence techniques (XAI) to evaluate the reliability of the vertical load for airplane landing gear. By utilizing limited data from the Quick Access Recorder (QAR), the improved GAN generates extensive synthetic data to expand the dataset and strengthen the analysis. Each parameter's importance and influence on vertical load reliability are then evaluated through the Shapley Additive Explanations (SHAP) method, a key approach in XAI. Validation using landing gear data from a typical civil airplane demonstrates the effectiveness of this method and confirms the viability of explainable artificial intelligence for parametric feedback analysis. The results highlight the impact of each parameter on vertical load reliability, providing valuable insights to support enhanced design and operational efficiency of landing gear.