The realization of road traffic prediction not only provides real-time and effective information for travelers, but also helps them select the optimal route to reduce travel time. Road traffic...

Our digital world is full of time series and graphs which capture the various aspects of many complex systems. Traditionally, there are respective methods in processing these two different types of data, e.g., Recurrent Neural Network (RNN) and Graph Neural Network (GNN), while in recent years, time series could be mapped to graphs by using the techniques such as Visibility Graph (VG), which simultaneously captures relevant aspects of both local and global dynamics in an easy way, so that researchers can use graph algorithms to mine the knowledge in time series and gain special latent graph representation features. Such mapping methods establish a bridge between time series and graphs, and have high potential to facilitate the analysis of various real-world time series. However, the VG method and its variants are just based on fixed rules and thus lack of flexibility, largely limiting their application in reality. In this paper, we propose an Adaptive Visibility Graph (AVG) algorithm that can adaptively map time series into graphs, based on which we further establish an end-to-end classification framework AvgNet, by utilizing GNN model DiffPool as the classifier. We then adopt AvgNet for radio signal modulation classification which is an important task in the field of wireless communication. The simulations validate that AvgNet outperforms a series of advanced deep learning methods, achieving the state-of-the-art performance in this task.

The real-time and accurate prediction of road traffic states is the basis of information service for road traffic participants. An algorithm based on kernel K-nearest neighbors (kernel-KNN) is presented to predict road traffic states in time series in this paper. First, representative road traffic state data are extracted to build the road traffic running characteristics reference sequences. Then, kernel function of the road traffic state data sequence in time series is constructed. The current and referenced road traffic state data sequences are matched, based on which k nearest referenced road traffic states are selected and the road traffic states are predicted. Several typical road links in Beijing are considered for a series of case studies. The final experiments results prove that the road traffic states prediction approach based on kernel-KNN presented herein is feasible and can achieve a high level of accuracy.

Vehicle trajectory prediction is a challenging problem in the field of autonomous driving, witch is of great significance to the safety of autonomous driving and traffic roads. In view of the interaction between surrounding vehicles and target vehicle and its own trajectory, we propose a new graph network model to predict future vehicle trajectory. First, the correlation network of vehicles at each time is constructed based on the complex network method. In order to make up for the lack of real spatial relevance caused by the fixed graph, we propose an adaptive parameter matrix to coordinate and optimize the global spatio-temporal graph. Second, the global spatio-temporal features of vehicle historical trajectory data are extracted by stacked graph convolution module. Finally, the obtained graph features are coded based on seq2seq network, and the trajectory prediction of road vehicles at different times in the future is realized. Our model has been trained and verified on the published NGSIM US-101 and I-80 data sets. Compared with other advanced schemes, our model has more accurate results in the future time of 5 seconds. In predicting the future group trajectory of vehicles on the road, the accuracy of long-term prediction is 11% higher than that of the most advanced scheme.

Road traffic state estimation is an essential component of intelligent transportation systems (ITSs). However, road traffic state data collected by traffic detectors are often incomplete, which can cause problems across a variety of transportation applications, such as traffic state prediction and pattern recognition. We present GA-GAN (Graph Aggregate Generative Adversarial Network), consisting of graph sample and aggregate (GraphSAGE) and a generative adversarial network (GAN), to impute missing road traffic state data. Instead of using the original road network structure, which presents the spatial information to process a graph operation, we reconstruct the road network according to the correlation coefficients of road historical data. We utilize GraphSAGE to aggregate the temporal-spatial information from the neighbors of each road in the reconstructed road network. GAN is used to generate complete traffic state data from the extracted temporal-spatial information to achieve traffic state data imputation. To illustrate the efficient performance of the model, experiments are conducted on traffic data collected from California and Seattle, Washington, showing that the proposed model outperforms state-of-the-art methods.

The future trajectory prediction of heterogeneous traffic-agents for autonomous vehicles in mixed traffic scene is of great significance for safe and reliable driving. Thus, we propose the Multi-View Adaptive Hierarchical Spatial Graph Convolution Network (MVHGN) to predict the future trajectories of heterogeneous traffic-agents. Firstly, multiple logical correlations are obtained based on the time series data of traffic-agents and a multi-view logical network is constructed. The multi-view logical feature extraction is realized based on the graph convolution module. Then, combining the multi-view logical features and the adaptive spatial topology network, the logical-physical features at the micro level are mined through the graph convolution module; based on the logical-physical features at the micro level and the regional clustering network at the macro level, the global logical-physical features are obtained. Finally, the model predicts the future trajectories of traffic-agents based on the encoder-decoder structure of the GRU. For the Apolloscape trajectory data set, the performance of our proposed method MVHGN is better than that of the comparison models.