隨著車輛智慧化的發展，開發自駕車各種功能也成為現代熱門研究方向，目前自駕車環周感知技術已大幅提升，在行駛於複雜車流環境時若能進一步了解其他用路人（例如行人與車輛）的意圖，便能採取更安全的因應策略，因此自駕車環周感知能力具備用路人的軌跡預測功能，對於自動駕駛安全性與可靠度扮演重要的角色。因此，本論文針對用路者移動軌跡預測提出一種混合式預測架構，此架構結合長短期記憶(Long Short-Term Memory, LSTM)編碼-解碼器網路與卡爾曼濾波器(Kalman Filter, KF)，其中KF可以穩定的預測用路人直行與轉彎移動的軌跡，LSTM編碼-解碼器能夠依據軌跡的資訊提早判斷用路人轉彎的趨勢，為了加強所提出的架構於不同移動軌跡的適應力，本論文設計適應性即時權重機制，根據兩個模型的預測誤差調整輸出權重加乘的比例，除此之外也使用LSTM編碼-解碼器的部分預測結果來強化KF針對用路人轉彎移動的預測精準度。目前本論文所開發的軌跡預測技術能夠應用於車輛、摩托車、及行人三種類別的用路人，為了驗證所提出方法的有效性與正確性，本論文除了透過Waymo開放資料集來訓練與測試模型之外，並在校園環境及一般市區道路行駛的自駕巴士平台進行資料蒐集與預測效能驗證。

The research and development on autonomous vehicles (AVs) have been a primary topic based on rapid improvements of automotive electronics. AVs have to understand the intent of other road users (pedestrians and vehicles) while driving to adopt complementary strategies. Therefore, trajectory prediction of surrounding targets is an integral part of AVs in order to enhance the safety and efficiency of autonomous driving. To this end, this thesis proposes a hybrid trajectory prediction architecture that combines Long Short-Term Memory (LSTM)-based encoder-decoder network and Kalman Filter (KF) for surrounding traffic agents. KF can be stable to predict the motions of the surrounding traffic agents, while the LSTM encoder-decoder network can judge the turning situation early based on the trajectory information. The prediction error of the model adjusts the ratio of output weight multiplication. In addition, the proposed predictor uses part of the prediction results of the LSTM encoder-decoder network to assist KF in acquiring the high accuracy of turning motion prediction. Initially, an analysis of prediction evaluation of our model through the Waymo Open Dataset is conducted with cars, motorcycles, and pedestrians. Finally, the experiments present the multiple case studies for the real traffic scenarios on the driverless shuttles.

[1] SAE 自動駕駛分類標準，資料來源https://www.sae.org/
[2] S. Qiao, D. Shen, X. Wang, N. Han and W. Zhu, "A Self-Adaptive Parameter Selection Trajectory Prediction Approach via Hidden Markov Models," IEEE Transactions on Intelligent Transportation Systems, vol. 16, no. 1, pp. 284-296, Feb. 2015, doi: 10.1109/TITS.2014.2331758.
[3] J. V. Dueholm, M. S. Kristoffersen, R. K. Satzoda, T. B. Moeslund and M. M. Trivedi, "Trajectories and Maneuvers of Surrounding Vehicles With Panoramic Camera Arrays," IEEE Transactions on Intelligent Vehicles, vol. 1, no. 2, pp. 203-214, June 2016, doi: 10.1109/TIV.2016.2622921.
[4] C. G. Prevost, A. Desbiens and E. Gagnon, "Extended Kalman Filter for State Estimation and Trajectory Prediction of a Moving Object Detected by an Unmanned Aerial Vehicle," 2007 American Control Conference, pp. 1805-1810, 2007, doi: 10.1109/ACC.2007.4282823.
[5] N. Deo, A. Rangesh and M. M. Trivedi, "How Would Surround Vehicles Move? A Unified Framework for Maneuver Classification and Motion Prediction," IEEE Transactions on Intelligent Vehicles, vol. 3, no. 2, pp. 129-140, June 2018, doi: 10.1109/TIV.2018.2804159.
[6] G. Xie, H. Gao, L. Qian, B. Huang, K. Li and J. Wang, "Vehicle Trajectory Prediction by Integrating Physics- and Maneuver-Based Approaches Using Interactive Multiple Models," IEEE Transactions on Industrial Electronics, vol. 65, no. 7, pp. 5999-6008, 2018, doi: 10.1109/TIE.2017.2782236.
[7] S. Su, F. Guo, Z. Chen and H. Huang, "IAT: Pedestrian Intention and Trajectory Prediction," 2020 Chinese Automation Congress (CAC), pp. 4182-4185, 2020, doi: 10.1109/CAC51589.2020.9327076.
[8] D. Jeong, M. Baek and S. -S. Lee, "Long-term prediction of vehicle trajectory based on a deep neural network," 2017 International Conference on Information and Communication Technology Convergence (ICTC), pp. 725-727, 2017, doi: 10.1109/ICTC.2017.8190764.
[9] S. Dai, L. Li and Z. Li, "Modeling Vehicle Interactions via Modified LSTM Models for Trajectory Prediction," IEEE Access, vol. 7, pp. 38287-38296, 2019, doi: 10.1109/ACCESS.2019.2907000.
[10] Y. Jeong, S. Kim and K. Yi, "Surround Vehicle Motion Prediction Using LSTM-RNN for Motion Planning of Autonomous Vehicles at Multi-Lane Turn Intersections," IEEE Open Journal of Intelligent Transportation Systems, vol. 1, pp. 2-14, 2020, doi: 10.1109/OJITS.2020.2965969.
[11] N. Deo and M. M. Trivedi, "Multi-Modal Trajectory Prediction of Surrounding Vehicles with Maneuver based LSTMs," 2018 IEEE Intelligent Vehicles Symposium (IV), pp. 1179-1184, 2018, doi: 10.1109/IVS.2018.8500493.
[12] S. H. Park, B. Kim, C. M. Kang, C. C. Chung and J. W. Choi, "Sequence-to-Sequence Prediction of Vehicle Trajectory via LSTM Encoder-Decoder Architecture," 2018 IEEE Intelligent Vehicles Symposium (IV), pp. 1672-1678, 2018, doi: 10.1109/IVS.2018.8500658.
[13] K. Saleh, M. Hossny and S. Nahavandi, "Long-Term Recurrent Predictive Model for Intent Prediction of Pedestrians via Inverse Reinforcement Learning," 2018 Digital Image Computing: Techniques and Applications (DICTA), pp. 1-8, 2018, doi: 10.1109/DICTA.2018.8615854.
[14] Y. Li, X.-Y. Lu, J. Wang and K. Li, "Pedestrian Trajectory Prediction Combining Probabilistic Reasoning and Sequence Learning, "IEEE Transactions on Intelligent Vehicles, " vol. 5, no. 3, pp. 461-474, Sept. 2020, doi: 10.1109/TIV.2020.2966117.
[15] S. Choi, N. Kweon, et al., "DSA-GAN: Driving Style Attention Generative Adversarial Network for Vehicle Trajectory Prediction," 2021 IEEE International Intelligent Transportation Systems Conference (ITSC), pp. 1515-1520, 2021, doi: 10.1109/ITSC48978.2021.9564674.
[16] P. Sun, H. Kretzschmar, et al., "Scalability in Perception for Autonomous Driving: Waymo Open Dataset," Conference on Computer Vision and Pattern Recognition (CVPR), 2020.
[17] J. Redmon and A. Farhadi, "YOLOv3: An Incremental Improvement," Conference on Computer Vision and Pattern Recognition (CVPR), 2018
[18] Y. Yan, Y. Mao and B. Li, "SECOND: Sparsely Embedded Convolutional Detection," Sensors, vol.18, issue 10, 3337, 2018.
[19] Y. Zhou and O. Tuzel, "VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection," 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4490-4499, 2018, doi: 10.1109/CVPR.2018.00472.
[20] H. Gunes and M. Piccardi, "Affect recognition from face and body: early fusion vs. late fusion," 2005 IEEE International Conference on Systems, Man and Cybernetics, vol. 4, pp. 3437-3443, 2005, doi: 10.1109/ICSMC.2005.1571679.
[21] C. G. M. Snoek, M. Worring, and A. Smeulders, "Early versus late fusion in semantic video analysis," 13th annual ACM international conference on Multimedia, pp. 399-402, 2005.
[22] S. Pang, D. Morris and H. Radha, "CLOCs: Camera-LiDAR Object Candidates Fusion for 3D Object Detection," 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 10386-10393, 2020, doi: 10.1109/IROS45743.2020.9341791.
[23] S. Hochreiter and J. Schmidhuber, "Long Short-Term Memory," Neural Comput, vol.9, issue 8, 1735–1780, 1997.