Abstract:A `trajectory' refers to a trace generated by a moving object in geographical spaces, usually represented by of a series of chronologically ordered points, where each point consists of a geo-spatial coordinate set and a timestamp. Rapid advancements in location sensing and wireless communication technology enabled us to collect and store a massive amount of trajectory data. As a result, many researchers use trajectory data to analyze mobility of various moving objects. In this dissertation, we focus on the `urban vehicle trajectory,' which refers to trajectories of vehicles in urban traffic networks, and we focus on `urban vehicle trajectory analytics.' The urban vehicle trajectory analytics offers unprecedented opportunities to understand vehicle movement patterns in urban traffic networks including both user-centric travel experiences and system-wide spatiotemporal patterns. The spatiotemporal features of urban vehicle trajectory data are structurally correlated with each other, and consequently, many previous researchers used various methods to understand this structure. Especially, deep-learning models are getting attentions of many researchers due to its powerful function approximation and feature representation abilities. As a result, the objective of this dissertation is to develop deep-learning based models for urban vehicle trajectory analytics to better understand the mobility patterns of urban traffic networks. Particularly, this dissertation focuses on two research topics, which has high necessity, importance and applicability: Next Location Prediction, and Synthetic Trajectory Generation. In this study, we propose various novel models for urban vehicle trajectory analytics using deep learning.

What problem does this paper attempt to address?

The problems that this paper attempts to solve are: How to use deep - learning techniques to analyze urban vehicle trajectory data in order to improve the understanding of vehicle movement patterns in urban traffic networks. Specifically, the author focuses on two main research topics: 1. **Next Location Prediction**: - This problem aims to predict the next location of a single vehicle based on historical trajectory data. - The author proposes a model based on Recurrent Neural Network (RNN) and further introduces Attention - based Recurrent Neural Network (ARNN) to combine traffic - state information within the network range, thereby improving prediction accuracy. 2. **Synthetic Trajectory Generation**: - This problem aims to generate virtual trajectories with realistic spatio - temporal movement patterns based on historical trajectory data. - The author develops the TrajGAIL model, which uses Generative Adversarial Imitation Learning (GAIL) and Partially Observable Markov Decision Process (POMDP) to generate virtual trajectories that conform to both individual vehicle characteristics and group patterns. Through these two research topics, the paper hopes to develop deep - learning methods that can better understand movement patterns in urban traffic networks, thereby providing users with more accurate location - prediction services and providing system operators with a more comprehensive view of traffic dynamics. This not only helps to improve the reliability of location services but also can solve problems such as data sparsity and privacy protection. ### Core Problem Summary - **Next Location Prediction**: Improve the accuracy of vehicle location prediction based on historical trajectory data. - **Synthetic Trajectory Generation**: Generate virtual trajectories with realistic spatio - temporal movement patterns to make up for the lack of actual data and protect privacy. These researches are of great significance for improving urban traffic management, optimizing travel experiences, and promoting the construction of smart cities.

Deep Learning based Urban Vehicle Trajectory Analytics

Dynamic-learning Spatial-Temporal Transformer Network for Vehicular Trajectory Prediction at Urban Intersections

Enhanced Multimodal Trajectory Prediction for Autonomous Vehicles Using Advanced Diffusion Model Techniques

TrajGAIL: Generating Urban Vehicle Trajectories using Generative Adversarial Imitation Learning

Trajectory Forecasting with Neural Networks: an Empirical Evaluation and A New Hybrid Model

Attention-based Recurrent Neural Network for Urban Vehicle Trajectory Prediction

Research on vehicle trajectory prediction methods in dense and heterogeneous urban traffic

Deep Kinematic Models for Kinematically Feasible Vehicle Trajectory Predictions

Deep Learning Techniques for Vehicle Trajectory Extraction in Mixed Traffic

Vehicle Trajectory Prediction for Automated Driving Based on Temporal Convolution Networks

A Deep Generative Model for Trajectory Modeling and Utilization.

A CNN-LSTM Based Model to Predict Trajectory of Human-Driven Vehicle

Human Trajectory Prediction using Spatially aware Deep Attention Models

Deep Learning for Trajectory Data Management and Mining: A Survey and Beyond

Vehicle Trajectory Prediction Using LSTMs With Spatial–Temporal Attention Mechanisms

Supervised vehicle trajectory prediction using orthogonal image map for urban automated driving

Kinematics-aware Trajectory Generation and Prediction with Latent Stochastic Differential Modeling

Multimodal Trajectory Predictions for Autonomous Driving using Deep Convolutional Networks

Characterized Diffusion and Spatial-Temporal Interaction Network for Trajectory Prediction in Autonomous Driving

Vehicle trajectory prediction based on cross-attention and multilevel spatio-temporal features