Trajectory Prediction Model Research Articles

Effects of CNN Backbone on Trajectory Prediction Models for Autonomous Vehicle

Effects of CNN Backbone on Trajectory Prediction Models for Autonomous Vehicle

Query-Informed Multi-Agent Motion Prediction.

In a dynamic environment, autonomous driving vehicles require accurate decision-making and trajectory planning. To achieve this, autonomous vehicles need to understand their surrounding environment and predict the behavior and future trajectories of other traffic participants. In recent years, vectorization methods have dominated the field of motion prediction due to their ability to capture complex interactions in traffic scenes. However, existing research using vectorization methods for scene encoding often overlooks important physical information about vehicles, such as speed and heading angle, relying solely on displacement to represent the physical attributes of agents. This approach is insufficient for accurate trajectory prediction models. Additionally, agents' future trajectories can be diverse, such as proceeding straight or making left or right turns at intersections. Therefore, the output of trajectory prediction models should be multimodal to account for these variations. Existing research has used multiple regression heads to output future trajectories and confidence, but the results have been suboptimal. To address these issues, we propose QINET, a method for accurate multimodal trajectory prediction for all agents in a scene. In the scene encoding part, we enhance the feature attributes of agent vehicles to better represent the physical information of agents in the scene. Our scene representation also possesses rotational and spatial invariance. In the decoder part, we use cross-attention and induce the generation of multimodal future trajectories by employing a self-learned query matrix. Experimental results demonstrate that QINET achieves state-of-the-art performance on the Argoverse motion prediction benchmark and is capable of fast multimodal trajectory prediction for multiple agents.

Construction of a Real-Time Ship Trajectory Prediction Model Based on Ship Automatic Identification System Data

The extraction of ship behavior patterns from Automatic Identification System (AIS) data and the subsequent prediction of travel routes play crucial roles in mitigating the risk of ship accidents. This study focuses on the Wuhan section of the dendritic river system in the middle reaches of the Yangtze River and the partial reticulated river system in the northern part of the Zhejiang Province as its primary investigation areas. Considering the structure and attributes of AIS data, we introduce a novel algorithm known as the Combination of DBSCAN and DTW (CDDTW) to identify regional navigation characteristics of ships. Subsequently, we develop a real-time ship trajectory prediction model (RSTPM) to facilitate real-time ship trajectory predictions. Experimental tests on two distinct types of river sections are conducted to assess the model’s reliability. The results indicate that the RSTPM exhibits superior prediction accuracy when compared to conventional trajectory prediction models, achieving an approximate 20 m prediction accuracy for ship trajectories on inland waterways. This showcases the advancements made by this model.

Vehicle Trajectory Prediction Based on Graph Convolutional Networks in Connected Vehicle Environment

Vehicle trajectory prediction is an important research basis for the decision making and path planning of the intelligent and connected vehicle. In the connected vehicle environment, vehicles share information and drive cooperatively, and the intelligent and connected vehicles are able to obtain more accurate and rich perception information, which provides a data basis for accurate prediction of vehicle trajectories. However, attaining accurate and effective vehicle trajectory predictions poses technical challenges due to insufficient extraction of vehicular spatial–temporal interaction features. In this paper, we propose a vehicle trajectory prediction model based on graph convolutional neural network (GCN) in a connected vehicle environment. Specifically, using the driving scene information obtained by the intelligent and connected vehicle, the spatial graph and temporal graph are constructed based on the spatial interaction coefficient (SIC) and self-attention mechanism, respectively. Then, the graph data are entered into the interaction extraction module, and the spatial interaction features and temporal interaction features are extracted separately using the graph convolutional networks, which are fused to obtain the spatial–temporal interaction information. Finally, the interaction features are learned based on the convolutional neural networks to output the future trajectory information of all vehicles in the scene by one forward operation rather than a step-by-step process. The ablation experiment results show that the method proposed in this study to model the spatiotemporal interaction among vehicles based on SIC and self-attention mechanism reduces the prediction error by 5% and 12%, respectively. The results from the model comparison experiment show that the proposed method engenders an 8% improvement in prediction accuracy over the state-of-the-art solution, providing technical and theoretical support for trajectory prediction research of intelligent and connected vehicles.

An interpretable machine learning model for trajectory prediction based on nonlinear dynamics mechanism constraints: applications for HVs

An interpretable machine learning model for trajectory prediction based on nonlinear dynamics mechanism constraints: applications for HVs

Cross-Camera Tracking Model and Method Based on Multi-Feature Fusion

Multi-camera video surveillance has been widely applied in crowd statistics and analysis in smart city scenarios. Most existing studies rely on appearance or motion features for cross-camera trajectory tracking, due to the changing asymmetric perspectives of multiple cameras and occlusions in crowded scenes, resulting in low accuracy and poor tracking performance. This paper proposes a tracking method that fuses appearance and motion features. An implicit social model is used to obtain motion features containing spatio-temporal information and social relations for trajectory prediction. The TransReID model is used to obtain appearance features for re-identification. Fused features are derived by integrating appearance features, spatio-temporal information and social relations. Based on the fused features, multi-round clustering is adopted to associate cross-camera objects. Exclusively employing robust pedestrian reidentification and trajectory prediction models, coupled with the real-time detector YOLOX, without any reliance on supplementary information, an IDF1 score of 70.64% is attained on typical datasets derived from AiCity2023.

TCN-Informer-Based Flight Trajectory Prediction for Aircraft in the Approach Phase

Trajectory prediction (TP) is a vital operation in air traffic control systems for flight monitoring and tracking. The approach phase of general aviation (GA) aircraft is more of a visual approach, which is related to the safety of the flight and whether to go around. Therefore, it is important to accurately predict the flight trajectory of the approach phase. Based on the historical flight trajectories of GA aircraft, a TP model is proposed with deep learning after feature extraction in this study, and the hybrid model combines a time convolution network and an improved transformer model. First, feature extraction of the spatiotemporal dimension is performed on the preprocessed flight data by using TCN; then, the extracted features are executed by adopting the Informer model for TP. The performance of the novel architecture is verified by experiments based on real flight trajectory data. The results show that the proposed TCN-Informer architecture performs better according to various evaluation metrics, which means that the prediction accuracies of the hybrid model are better than those of the typical prediction models widely used today. Moreover, it has been verified that the proposed method can provide valuable suggestions for decision-making regarding whether to go around during the approach.

A Dynamic Air Combat Situation Assessment Model Based on Situation Knowledge Extraction and Weight Optimization

Air combat situation assessment is the basis of target assignment and maneuver decisions. The current air combat situation assessment models, whether nonparametric or parametric, ignore the continuity and timing of situation changes, making the situation assessment results lose tactical significance. Aimed at the shortcomings of current air combat situation assessment, a dynamic air combat situation assessment model based on situation knowledge extraction and weight optimization was proposed by combining a multiple regression model of hidden logic process, a weight optimization model based on grey prospect theory, a weight mapping model based on autoencoder and extreme learning machine (AE-ELM) and an air combat situation characteristic parameter prediction model based on dynamic weight online extreme learning machine (DWOSELM). Firstly, considering the timing and continuity of air combat situation change, a hidden logic process multiple regression model was introduced to realize the segmentation of air combat situation time series data and the extraction of air combat situation primitives. Secondly, the weight optimization method based on grey prospect theory was used to obtain the weight of the evaluation index under different air combat situations. On this basis, the dynamic mapping model between air combat situation characteristic parameters and the weight of index was constructed by using AE-ELM. Then, the dynamic weighted extreme learning machine was used to build the target maneuver trajectory prediction model, and the future position information of the target was predicted. On this basis, the future situation information between the enemy and us was obtained. Finally, the time weight calculation model based on normal cumulative distribution was used to determine the importance of the situation at each time. The situation information at multiple times in the air combat process was fused to obtain the comprehensive air combat situation assessment results at the current time. The simulation results show that the model can fully exploit the influence of historical information, effectively integrate the air combat situation information at multiple moments, and generate the air combat situation assessment results with practical tactical significance according to the individual differences of different pilots.

An autonomous coverage path planning algorithm for maritime search and rescue of persons-in-water based on deep reinforcement learning

The prevalence of maritime transportation and operations is increasing, leading to a gradual increase in drowning accidents at sea. In the context of maritime search and rescue (SAR), it is essential to develop effective search plans to improve the survival probability of persons-in-water (PIWs). However, conventional SAR search plans typically use predetermined patterns to ensure complete coverage of the search area, disregarding the varying probabilities associated with the PIW distribution. To address this issue, this study has proposed a maritime SAR vessel coverage path planning framework (SARCPPF) suitable for multiple PIWs. This framework comprises three modules, namely, drift trajectory prediction, the establishment of a multilevel search area environment model, and coverage search. First, sea area-scale drift trajectory prediction models were employed using the random particle simulation method to forecast drift trajectories. A hierarchical probability environment map model was established to guide the SAR of multiple SAR units. Subsequently, we integrated deep reinforcement learning with a reward function that encompasses multiple variables to guide the navigation behavior of ship agents. We developed a coverage path planning algorithm aimed at maximizing the success rates within a limited timeframe. The experimental results have demonstrated that our model enables vessel agents to prioritize high-probability regions while avoiding repeated coverage.

Experimental and numerical analysis study on the low and medium speed bird strike

The bird impact process of civil aircraft mostly occurs in the takeoff and landing phases. The movement speed of the aircraft in these phases belong to the category of low and medium speed, and the bird will not disintegrate in this velocity range. The speed and direction of the bird after impacting with the air intake directly affects the damage to the aero-engine. According to the preliminary experimental data, we developed a bird strike trajectory prediction model based on the quasi-linear viscoelastic model. This model reproduced the trajectory change law in the low and medium speed bird impact experiment combining with the simulation analysis under multiple conditions. We designed a simulated air intake model for the real air intake structure characteristics, and processed the simulated air intake parts based on this model. The accuracy of the model was further verified by comparing the simulated and experimental bird impact trajectories, and a predictive modeling method was obtained for the simulation of low and medium speed bird impact trajectories.

A context-aware hybrid deep learning model for the prediction of tropical cyclone trajectories

Tropical cyclones (TCs) are powerful natural disasters that lead to the loss of human lives and destruction of nature. Accurate prediction of the TC trajectory over a long period is crucial for timely warning and safe evacuation. Given the complexity of TCs and the geographical contexts that affects their movements, robust tools are required to process historical TC data and predict their locations. In this study, three deep learning models with flexible topology, namely, the multilayer perceptron (MLP) model, long short-term memory (LSTM) model, and a new data-driven model (MLP-LSTM) for TC trajectory prediction, were developed. They were evaluated using the North Atlantic Ocean TC dataset and contextual information such as wind speed, wind direction, and air pressure. The results showed that the hybrid MLP-LSTM model outperformed the MLP and LSTM models, especially when contextual information was considered. The average prediction distance errors for the next three hours were calculated as 52.73 km, 20.65 km, and 19.54 km for the MLP, LSTM, and MLP-LSTM models, respectively. The prediction enhancements for the next 24 h when contextual information was considered were 166 km, 203 km, and 208 km for the MLP, LSTM, and MLP-LSTM models, respectively.

A novel ship trajectory clustering analysis and anomaly detection method based on AIS data

The increasing volume of ship traffic has resulted in new challenges for the supervision of maritime safety administration. The conventional manual monitoring approach for maritime traffic is inefficient and lacks specifics, particularly for supervising ships with abnormal trajectories. To address this issue, this study proposes utilizing the minimum description length criterion to extract features from ship trajectory data provided by the automatic identification system (AIS). This approach simplifies the compression of ship trajectories. Additionally, the dynamic time warping trajectory similarity measurement algorithm is employed to optimize the density-based spatial clustering of applications with noise algorithm. This optimization enables the clustering of ship trajectories and the acquisition of normalized ship motion trajectories. Furthermore, a ship trajectory prediction method based on a transformer model is proposed, and the normalized motion trajectory is used as the training set for model training. The trained ship trajectory prediction model is subsequently utilized to predict the target ship trajectory. The AIS ship trajectory data in the vicinity of Yantai Port were used for experimental verification. The results demonstrate the effectiveness of the proposed approach in identifying abnormal ship trajectories.

GATraj: A graph- and attention-based multi-agent trajectory prediction model

Trajectory prediction has been a long-standing problem in intelligent systems like autonomous driving and robot navigation. Models trained on large-scale benchmarks have made significant progress in improving prediction accuracy. However, the importance on efficiency for real-time applications has been less emphasized. This paper proposes an attention-based graph model, named GATraj, which achieves a good balance of prediction accuracy and inference speed. We use attention mechanisms to model the spatial–temporal dynamics of agents, such as pedestrians or vehicles, and a graph convolutional network to model their interactions. Additionally, a Laplacian mixture decoder is implemented to mitigate mode collapse and generate diverse multimodal predictions for each agent. GATraj achieves state-of-the-art prediction performance at a much higher speed when tested on the ETH/UCY datasets for pedestrian trajectories, and good performance at about 100Hz inference speed when tested on the nuScenes dataset for autonomous driving. We conduct extensive experiments to analyze the probability estimation of the Laplacian mixture decoder and compare it with a Gaussian mixture decoder for predicting different multimodalities. Furthermore, comprehensive ablation studies demonstrate the effectiveness of each proposed module in GATraj.

A federated pedestrian trajectory prediction model with data privacy protection

Pedestrian trajectory prediction is essential for self-driving vehicles, social robots, and intelligent monitoring applications. Diverse trajectory data is critical for high-accuracy trajectory prediction. However, the trajectory data is captured in scattered scenes, which can cause the problem of data island. Furthermore, artificial aggregation of trajectory data suffers from the risk of data leakage, ignoring the rule of privacy protection. We propose a multi-scene federated trajectory prediction (Fed-TP) method to solve the above problems. As our key contribution, a destination-oriented LSTM (Long Short-Term Memory)-based trajectory prediction (DO-TP) network is proposed in each scene to forecast future trajectories in an encoder-decoder manner. The independent training using trajectory data in each scene can prevent data leakage and achieves high privacy security. As another key contribution, a federated learning framework is introduced to break the scene limitation by conducting distributed collaborative training. The performance of different federated learning methods is compared on public datasets, including ETH, UCY, and Stanford Drone Dataset (SDD). Compared with FedAvg and FedProx, FedAtt is more suitable for pedestrian trajectory prediction. Experimental results demonstrate that the proposed method has better data privacy security than directly training on multiple scenes and superior prediction performance than training on a single scene.

Prediction of ship trajectory based on deep learning

The rapid development of computer technology strongly promotes the study of maritime traffic safety. The application of artificial intelligence technology makes the ship’s trajectory prediction not limited to complex physical models, and improves the generality of ship’s track prediction. To address the problems of relatively poor prediction accuracy in existing ship trajectory prediction research and high coupling of algorithms during ship navigation while the prediction model needs to be improved, this paper combines grey forecast prediction model and long and short-term memory (LSTM) neural network to establish a neural network model for ship trajectory prediction. This model can effectively increase the number of messages in LSTM model, and reduce the loss value of the model and improve the accuracy of ship track prediction. The prediction method in this paper is verified by AIS data of a ship. The results show that compared with other time series prediction algorithms, this algorithm has higher prediction accuracy.

A Vectorized Representation Model for Trajectory Prediction of Intelligent Vehicles in Challenging Scenarios

A Vectorized Representation Model for Trajectory Prediction of Intelligent Vehicles in Challenging Scenarios

MPC‐Bi‐LSTM based control strategy for connected and automated vehicles platoon oriented to cyberattacks

AbstractAny technological innovation will be accompanied by new challenges and risks, and the connected and automated vehicles (CAVs) are no exception. Among them, the argument that cooperating platoons may fall victim to cyberattacks through wireless communication has emerged as a significant issue. Therefore, this paper designs a communication topology anomaly response system (CTARS) to ensure platoon safety, which consists of a trigger module and a control module. The primary objective of the trigger module is to distinguish abnormal vehicle behavior based on time to collision (TTC) indicators, and the control module combines the model predictive control (MPC) and bidirectional long short‐term memory (Bi‐LSTM) to achieve accurate trajectory prediction of and optimal control, working in tandem with the trigger module. Subsequently, the real dataset HISTORIC is used to calibrate the multiple vehicle intelligent driver model (IDM) and train the trajectory prediction model. Furthermore, comparative simulations are conducted, encompassing various forms of cyberattacks, in order to examine the evolution characteristics of CAVs platoons (CAVP) and evaluate the performance of CTARS. The results demonstrate the remarkable effectiveness of CTARS in safeguarding the security of CAVP during cyberattacks, thus confirming its exceptional performance.

A generalized driving risk assessment on high-speed highways using field theory

This study presents a new safety measure derived from field theory. It evaluates the risk arising from the various concurrent conflicts within a platoon that can occur on high-speed highway driving situations, such as car-following, yielding, and lane changing. We defined the risk field as a finite scalar field produced by traveling vehicles on the road, and we defined the conflict field as the overlapping risk field between any vehicles in proximity on the roadway. The study used a probabilistic trajectory prediction model to construct risk fields and an approximation method to reduce the computational time for real-time applications. To demonstrate the applicability of the proposed new measure, we applied it to real-world trajectory data (NGSIM data from US Highway 101). We compared the results with three traditional conflict-based safety measures: post-encroachment time (PET), modified time-to-collision (MTTC), and deceleration rate to avoid a crash (DRAC). The new measure produced seamless and continuous risk estimations even during time windows when the other measures could not estimate the risk between vehicles. This is a major advantage over traditional measures. The study also developed visual displays of the estimated conflict fields to provide safety analysts with an intuitive and fast understanding of the results of the safety assessments made using the conflict field measure. We conclude that the proposed new safety measure provides a robust, reliable, and improved assessment of the risk involved in expected future mixed-traffic environments that involve both human-driven vehicles and automated vehicles in the future.

A Gated-Recurrent-Unit-Based Interacting Multiple Model Method for Small Bird Tracking on Lidar System.

Lidar presents a promising solution for bird surveillance in airport environments. However, the low observation refresh rate of Lidar poses challenges for tracking bird targets. To address this problem, we propose a gated recurrent unit (GRU)-based interacting multiple model (IMM) approach for tracking bird targets at low sampling frequencies. The proposed method constructs various GRU-based motion models to extract different motion patterns and to give different predictions of target trajectory in place of traditional target moving models and uses an interacting multiple model mechanism to dynamically select the most suitable GRU-based motion model for trajectory prediction and tracking. In order to fuse the GRU-based motion model and IMM, the approximation state transfer matrix method is proposed to transform the prediction of GRU-based network into an explicit state transfer model, which enables the calculation of the models' probability. The simulation carried out on an open bird trajectory dataset proves that our method outperforms classical tracking methods at low refresh rates with at least 26% improvement in tracking error. The results show that the proposed method is effective for tracking small bird targets based on Lidar systems, as well as for other low-refresh-rate tracking systems.

An Interacting Multiple Model for Trajectory Prediction of Intelligent Vehicles in Typical Road Traffic Scenario.

This article presents an interacting multiple model (IMM) for short-term prediction and long-term trajectory prediction of an intelligent vehicle. This model is based on vehicle's physics model and maneuver recognition model. The long-term trajectory prediction is challenging due to the dynamical nature of the system and large uncertainties. The vehicle physics model is composed of kinematics and dynamics models, which could guarantee the accuracy of short-term prediction. The maneuver recognition model is realized by means of hidden Markov model, which could guarantee the accuracy of long-term prediction, and an IMM is adopted to guarantee the accuracy of both short-term prediction and long-term prediction. The experiment results of a real vehicle are presented to show the effectiveness of the prediction method.