Optimizing Trajectory Prediction for Autonomous Vehicles: A Comparative Study of Kalman Filtering and Machine Learning Approaches

With the development of technology, autonomous driving technology has become a popular research direction. Currently, there are still some unresolved or optimization problems in autonomous driving technology, such as large trajectory prediction deviations. Therefore, this study aims to improve the accuracy of autonomous driving technology. To solve this problem, we achieve this goal by optimizing algorithms. This study used Kalman filtering algorithm and linear regression, decision Tree and random forest regression in machine learning to predict trajectories and obtain results by processing various sensor data. The obtained results were then compared and analyzed, and the optimal algorithm was finally obtained. It can be concluded that when using a combination of IMU, GNSS, and LiDAR, random forest regression has the highest accuracy and can improve the safety and reliability of autonomous driving technology.