Simulation-based Aircraft Attitude Estimation Using CNN RNN Approach

Recent advances in deep learning have led to the widespread use of convolutional neural networks (CNNs) and recurrent neural networks (RNNs) across various fields. For aircraft attitude estimation, CNNs can effectively extract spatial features from images, whereas RNNs model temporal continuity for more accurate predictions. This study proposes a hybrid method that combines CNN and RNN architectures to improve the accuracy and stability of real-time aircraft attitude estimation. CNN extracts spatial information from each frame, and the RNN captures the temporal dependencies to predict attitude changes. A dataset from the X-Plane 11 simulator collected under diverse environmental conditions was used to train the model. The proposed approach achieved high accuracy based on performance metrics, such as the mean absolute error, root mean square error, and R-squared, confirming its effectiveness. The simulation results demonstrated reliable attitude estimation even in complex environments, making it suitable for real-time autonomous flight systems. These findings suggest that the proposed model can significantly enhance the safety and reliability of such systems.