Algorithm for correcting trajectory anomalies by integrating GeoHash grid division and long short-term memory network prediction

With the rapid development of intelligent transportation and positioning technology, the accuracy of vehicle trajectory data is crucial. However, due to the influence of complex environmental factors, anomalies often appear in the trajectory, which seriously restricts the reliability of related applications. This paper proposes a fusion algorithm of GeoHash grid division and Long Short-Term Memory Network (LSTM) prediction for trajectory anomaly correction. This algorithm uses GeoHash algorithm to divide trajectory points into grids, achieving efficient data organization and map matching; and utilizes LSTM's powerful time series processing ability to accurately predict the real coordinates of anomalies. Experiments have shown that this algorithm significantly improves the accuracy and efficiency of trajectory anomaly point correction, providing reliable technical support for traffic management, autonomous driving, and other fields.