Short-term Traffic Flow Prediction Research Articles

A Diverse Ensemble Deep Learning Method for Short-Term Traffic Flow Prediction Based on Spatiotemporal Correlations

In this paper, considering spatiotemporal correlations, we propose a novel short-term traffic flow prediction method that is based on diverse ensemble deep learning. First, a new measurement function of spatial correlation is established, quantifying the spatial correlation of traffic flow parameters. Second, the convolutional neural network model embedded long short-term memory (LSTM-CNN) is used, enhancing the learning ability of the spatiotemporal correlation of traffic flow parameters. Third, according to the network structure, ensemble rules of diverse LSTM-CNNs are constructed, combining a series of different and moderately accurate LSTM-CNN models and improving the robustness of the algorithm. Finally, a dynamic optimization method for the weight parameters of the ensemble elements is proposed to accommodate the changes in the actual road networks. The experiments show that the measuring method of spatial correlations, the ensemble combination rules of diverse LSTM-CNN and the dynamic optimization method of ensemble element weight parameters, which are proposed in this paper, have positive effects on improving the performance of the prediction algorithm. In addition, the short-term traffic flow prediction method proposed in this paper can adapt to changes in the traffic flow. It can obtain a better prediction effect even with a small sample, and the prediction performance is better than that of the other four classical short-term traffic flow prediction methods.

Forecasting citywide short-term turning traffic flow at intersections using an attention-based spatiotemporal deep learning model

Prediction of short-term traffic flow has been examined recently, but little attention has been paid to the prediction of citywide turning traffic flow at intersections. Based on an in-depth analysis of turning traffic flow patterns, we propose a novel attention-based spatiotemporal deep learning model to predict citywide short-term turning traffic flow at road intersections with high accuracy. First, we examine the spatiotemporal patterns of turning traffic flow. Then, an end-to-end deep learning structure with four components is designed to model turning traffic flow. In our model, graph convolutional network is revised to learn spatial dependencies and sparseness, and gate recurrent unit network with an attention mechanism is developed to learn temporal dependencies and fluctuations. Experiments were conducted in Wuhan, China, where taxicab trajectory data were used to train and validate our model. The results suggest that our model outperforms current state-of-the-art models with higher accuracy on estimating turning traffic flow.

Short-Term Traffic Flow Prediction Based on CNN-BILSTM with Multicomponent Information

Problem definition: The intelligent transportation system (ITS) plays a vital role in the construction of smart cities. For the past few years, traffic flow prediction has been a hot study topic in the field of transportation. Facing the rapid increase in the amount of traffic information, finding out how to use dynamic traffic information to accurately predict its flow has become a challenge. Methodology: Thus, to figure out this issue, this study put forward a multistep prediction model based on a convolutional neural network and bidirectional long short-term memory (BILSTM) model. The spatial characteristics of traffic data were considered as input of the BILSTM model to extract the time series characteristics of the traffic. Results: The experimental results validated that the BILSTM model improved the prediction accuracy in comparison to the support vector regression and gated recurring unit models. Furthermore, the proposed model was comparatively analyzed in terms of mean absolute error, mean absolute percentage error, and root mean square error, which were reduced by 30.4%, 32.2%, and 39.6%, respectively. Managerial implications: Our study provides useful insights into predicting the short-term traffic flow on highways and will improve the management of traffic flow optimization.

Spatial–temporal short-term traffic flow prediction model based on dynamical-learning graph convolution mechanism

Short-term traffic flow prediction is a core branch of intelligent traffic systems (ITS) and plays an important role in traffic management. The graph convolution network (GCN) is widely used in traffic prediction models to efficiently handle the graphical structural data of road networks. However, the influence weights among different road sections are usually distinct in real life and are difficult to analyze manually. The traditional GCN mechanism, which relies on a manually set adjacency matrix, is unable to dynamically learn such spatial patterns during training. To address this drawback, this study proposes a novel location graph convolutional network (location-GCN). The location-GCN solves this problem by adding a new learnable matrix to the GCN mechanism, using the absolute value of this matrix to represent the distinct influence levels among different nodes. Subsequently, long short-term memory (LSTM) is employed in the proposed traffic prediction model. Moreover, trigonometric function encoding was used in this study to enable the short-term input sequence to convey long-term periodic information. Finally, the proposed model was compared with the baseline models and evaluated on two real-world traffic flow datasets. The results show that our model is more accurate and robust than the other representative traffic prediction models.

A novel hybrid model combining a fuzzy inference system and a deep learning method for short-term traffic flow prediction

Deep learning techniques have been widely used in traffic flow prediction. They can perform much better than shallow models. However, most existing deep learning models only focus on deterministic data, ignoring the fact that traffic flow contains a large amount of uncertain data. Therefore, this paper proposes a novel hybrid model called FGRU combining a fuzzy inference system (FIS) and a gated recurrent unit (GRU) neural network to predict short-term traffic flows. The GRU model is applied to capture the temporal dependencies within traffic flow data, and the FIS makes up for the shortcomings of deep learning by lessening the influence of uncertain data. In addition, a temporal feature enhancement mechanism is proposed to calculate the appropriate time intervals as model inputs. The most appropriate model structure and parameters are explored by performing comparative experiments. Finally, the simulation results show that the mean absolute error of FGRU is 7.75% and 3.05% lower than ARIMA and the state-of-the-art traffic flow prediction model based on deep learning.

Hybrid Model for Method for Short-Term Traffic Flow Prediction Based on Secondary Decomposition Technique and ELM

Strong nonstationary and nonlinearity are the main characteristics in the short-term traffic flow data, which frustrates traditional methods (e.g., autoregressive integrated moving average and deep belief network) to provide a satisfactory prediction. To address the above problem, a novel forecasting method, which is composed of a secondary decomposition technique and extreme learning machine, is proposed in this study. This developed technique is a hybrid of time-varying filtering-empirical mode decomposition (TVF-EMD) and local mean decomposition (LMD), which not only can effectively handle the above complex data features by decomposing them into several regular subsets but also produce the smoother subseries that is beneficial to prediction. To verify the effectiveness of the proposed method, a case study based on two groups of actual traffic flow data with different characteristics is performed. Meanwhile, several single models and hybrid models based on the other decomposition methods (e.g., EMD and variational mode decomposition) are considered benchmark models. The experimental results reveal that the proposed model presents the best performance. For example, compared with the TVF-EMD-based method, the improvement by the proposed approach reaches 33.3% in terms of the evaluation criterion of mean absolute percentage error.

A Hybrid Deep Learning Model for Short-Term Traffic Flow Pre-Diction Considering Spatiotemporal Features

Traffic flow prediction is one of the basic, key problems with developing an intelligent transportation system since accurate and timely traffic flow prediction can provide information support and decision support for traffic control and guidance. However, due to the complex characteristics of traffic information, it is still a challenging task. This paper proposes a novel hybrid deep learning model for short-term traffic flow prediction by considering the inherent features of traffic data. The proposed model consists of three components: the recent, daily and weekly components. The recent component is integrated with an improved graph convolutional network (GCN) and bi-directional LSTM (Bi-LSTM). It is designed to capture spatiotemporal features. The remaining two components are built by multi-layer Bi-LSTM. They are developed to extract the periodic features. The proposed model focus on the important information by using an attention mechanism. We tested the performance of our model with a real-world traffic dataset and the experimental results indicate that our model has better prediction performance than those developed previously.

Interval Short-Term Traffic Flow Prediction Method Based on CEEMDAN-SE Nosie Reduction and LSTM Optimized by GWO

With rapid economic growth and urbanization, the accelerated increase in car ownership has brought massive pressure on urban traffic, and accurate traffic flow prediction information can provide an important basis for urban traffic dynamic planning. The existing methods have problems such as low efficiency, large error, and inability to adapt to short-term traffic changes. To solve the above problems, the CEEMDAN-SE-GWO-LSTM method was proposed in this paper. First, the traffic flow data is processed for outliers and missing values. The Complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) method is used to decompose the traffic flow data, and sample entropy (SE) is used to reconstruct the subsequence, which is used to improve the quality of the input data. Then, the Grey Wolf Optimizer (GWO) is used to optimize the parameters of the long-short-term memory (LSTM) in order to improve the prediction accuracy and prevent the model from falling into a local optimum. Three models are used to compare with the ensemble model proposed in this paper, including back propagation neural network (BPNN), LSTM, and long-short-term memory optimized by Grey Wolf Optimizer (GWO-LSTM). Root mean square error (RMSE) is reduced by 40.9% to 66.7%;R2score is improved by 1.5% to 7.1%. The experimental results show that CEEMDAN-SE-GWO-LSTM has a higher prediction accuracy than the existing traffic flow prediction models. Finally, this paper uses the model prediction error to establish an interval prediction model based on the kernel density estimation theory, which enhances the generalization of the model and the practical application value.

A Two-Stage Decomposition-Reinforcement Learning Optimal Combined Short-Time Traffic Flow Prediction Model Considering Multiple Factors

Accurate short-term traffic flow prediction is a prerequisite for achieving an intelligent transportation system to proactively alleviate traffic congestion. Considering the complex and variable traffic environment, so that the traffic flow contains a large number of non-linear characteristics, which makes it difficult to improve the prediction accuracy, a combined prediction model that reduces the unsteadiness of traffic flow and fully extracts the traffic flow features is proposed. Firstly, decompose the traffic flow data into multiple components by the seasonal and trend decomposition using loess (STL); these components contain different features, and the optimized variational modal decomposition (VMD) is used for the second decomposition of the component with large fluctuation frequencies, and then the components are reconstructed according to the fuzzy entropy and Lempel-Ziv complexity index and the Pearson correlation coefficient is used to filter the traffic flow features. Then light gradient boosting machine (LightGBM), long short-term memory with attention mechanism (LA), and kernel extreme learning machine with genetic algorithm optimization (GA-KELM) are built for prediction. Finally, we use reinforcement learning to integrate the advantages of each model, and the weights of each model are determined to obtain the best prediction results. The case study shows that the model established in this paper is better than other models in predicting urban road traffic flow, with an average absolute error of 2.622 and a root mean square error of 3.479, both of which are lower than the prediction errors of other models, indicating that the model can fully extract the features in complex traffic flow.

Research Based on vehicle exhaust emission estimation

With the development and progress of society, short-term traffic flow prediction has become an increasingly important topic. Accurate prediction of traffic flow can not only provide necessary guidance for the transportation department, but also bring convenience to travel, but also play an important role in the prevention and control of urban air pollution. This paper mainly studies the prediction of vehicle flow, and then estimates the vehicle exhaust through the calculated vehicle flow, trying to establish the estimation model of vehicle flow and vehicle exhaust. Through Gaode map interface and WGS-84 (GPS) coordinates of each bayonet provided, the road name corresponding to the bayonet equipment is obtained according to the distance vector, and the data is preprocessed. For the interval road motor vehicle flow, the dynamic prediction can be realized by using the interdependence between the time series observations. According to the time dimension of prediction, traffic flow prediction can generally be divided into long-term traffic flow prediction (in year), medium-term traffic flow prediction (in month and day) and short-term traffic flow prediction (in hour and minute). According to the data provided, due to the weak regularity of short-term traffic flow data of interval roads, strong random error interference, and high uncertainty, it is difficult to predict accurately, The long-term and medium-term traffic flow data have strong periodicity and weak random interference. Considering the accuracy and data provided, this paper mainly selects the prediction method based on endogenous variables represented by autoregressive moving average model ARIMA and time cycle neural network LSTM. Because the use conditions of the two models are different, it is necessary to improve the model and modify the use scenario of the model. The advantage of this method is that the data acquisition cost is low and easy to implement. Finally, the prediction effect of traffic flow is evaluated by cellular automata to improve the performance of the model.

Short‐term traffic flow prediction approach incorporating vehicle functions from RFID‐ELP data for urban road sections

Short-term traffic flow prediction is particularly important for urban traffic control and congestion management. To improve the accuracy of traffic flow prediction, a new traffic flow prediction approach is proposed from the perspective of vehicle functions by using Radio Frequency Identification (RFID) electronic license plate (ELP) data. First, the improved Wavelet Neural Network (WNN) is introduced by adding three momentum terms to avoid local minimum and overfitting. The Long Short-Term Memory Neural Network (LSTM) is constructed to train the time series with long time lags. Second, a novel short-term traffic flow prediction method is developed by combining the improved WNN and LSTM in an urban road section. The proposed method performs a weighted combination to obtain a new prediction value according to whether the difference between the prediction values of the two methods satisfy the defined threshold, and then selects the one with the smallest error among the three prediction values as the final prediction result. Third, the experiments explore the prediction performance of the weekdays and weekends with 5 min-interval, 15 min-interval and 30 min-interval via using RFID-ELP data from Chongqing, China. Compared with existing prediction methods, the proposed approach can promote prediction performance and meet real-time requirements.

Improved genetic algorithm optimized LSTM model and its application in short-term traffic flow prediction.

Considering that the road short-term traffic flow has strong time series correlation characteristics, a new long-term and short-term memory neural network (LSTM)-based prediction model optimized by the improved genetic algorithm (IGA) is proposed to improve the prediction accuracy of road traffic flow. Firstly, an improved genetic algorithm (IGA) is proposed by dynamically adjusting the mutation rate and crossover rate of standard GA. Secondly, the parameters of the LSTM, such as the number of hidden units, training times, gradient threshold and learning rate, are optimized by the IGA. Therefore, the optimal parameters are obtained. In the analysis stage, 5-min short-term traffic flow data are used to demonstrate the superiority of the proposed method over the existing neural network algorithms. Finally, the results show that the Root Mean Square Error achieved by the proposed algorithm is lower than that achieved by the other neural network methods in both the weekday and weekend data sets. This verifies that the algorithm can adapt well to different kinds of data and achieve higher prediction accuracy.

Short-Term Traffic Flow Prediction via Improved Mode Decomposition and Self-Attention Mechanism Based Deep Learning Approach

Short-term traffic flow prediction (STFP) is one of the key technologies in Intelligence Transportation System (ITS). With the development of artificial intelligence technology, deep learning has been employed to STFP and certain achievements have been obtained. However, further improvement of STFP accuracy is constrained by the presence of noise, modal aliasing in the modal components and insufficient feature extraction of neural networks. Therefore, to more accurately capture the changing trend of traffic flow in different time periods, and increase the reliability and interpretability of the prediction models, improved variational mode decomposition (IVMD) and self-attention mechanism (SAM) based hybrid convolutional neural network (CNN) and long short term memory network (LSTM) are proposed for STFP in this study, benefited from the modal decomposition of IVMD and redistribution of the neural weights in CNN-LSTM networks by SAM, the proposed IVMD-CNN-LSTM-SAM effectively suppresses the influence of randomness and volatility of traffic flow, and accurately predict the short-term lane occupancy in the designated area. Furthermore, this study analyses the mechanism of SAM and IVMD in STFP, and proves the superiority of IVMD and SAM in signal decomposition with physical significance and weight allocation respectively. Therefore, the interpretability of the proposed neural network model is significantly improved. Eleven competitive neural networks forecasting methods are used as benchmarks and numerical studies based on actual traffic flow validate the accuracy and stability of the proposed IVMD-CNN-LSTM-SAM. Mean absolute percentage error and goodness of fit of the proposed model are 0.81% and 0.9968, respectively.

Highway Traffic Flow Prediction Model Construction Based on the Gray Theory and BP Neural Network.

The short-term traffic flow prediction and modeling of highways are the core content and important foundation of highway management decision-making support systems. It is of great significance to improving the level of highway management. Based on the macrodynamic traffic flow model, this article establishes a method for establishing a highway traffic flow prediction model based on the BP neural network theory and gray theory. We collected and carry out modeling and prediction of highway traffic flow data near a certain station. It is learned from the prediction results that the traffic flow prediction model based on the BP neural network and gray theory has a high degree of reliability.

Using support vector regression and K-nearest neighbors for short-term traffic flow prediction based on maximal information coefficient

The prediction of short-term traffic flow is critical for improving service levels for drivers and passengers as well as enhancing the efficiency of traffic management in the urban transportation system. For transportation departments, the issue remains of how to efficiently utilize the spatial and temporal information of traffic data for better prediction performance. As a means of improving traffic prediction accuracy, this paper proposes a method for screening spatial time-delayed traffic series based on the maximal information coefficient. The selected time-delayed traffic series are transformed into traffic state vectors, from which traffic flow is predicted by adopting the combination of support vector regression method and k-nearest neighbors method. We employ the proposed framework for real-world traffic flow prediction. Root Mean Squared Error (RMSE) and Mean Absolute Percent Error (MAPE) validate the superior performance of the proposed model compared to traditional methods. This new approach reduces the RMSE by 23.448% and the MAPE by 14.726% of the predicted results.

Entropy-Weight-Method-Based Integrated Models for Short-Term Intersection Traffic Flow Prediction.

Three different types of entropy weight methods (EWMs), i.e., EWM-A, EWM-B, and EWM-C, have been used by previous studies for integrating prediction models. These three methods use very different ideas on determining the weights of individual models for integration. To evaluate the performances of these three EWMs, this study applied them to developing integrated short-term traffic flow prediction models for signalized intersections. At first, two individual models, i.e., a k-nearest neighbors (KNN)-algorithm-based model and a neural-network-based model (Elman), were developed as individual models to be integrated using EWMs. These two models were selected because they have been widely used for traffic flow prediction and have been approved to be able to achieve good performance. After that, three integrated models were developed by using the three different types of EWMs. The performances of the three integrated models, as well as the individual KNN and Elman models, were compared. We found that the traffic flow predicted with the EWM-C model is the most accurate prediction for most of the days. Based on the model evaluation results, the advantages of using the EWM-C method were deliberated and the problems with the EWM-A and EWM-B methods were also discussed.

Short-Term Traffic Flow Prediction of Expressway Considering Spatial Influences

Real-time and accurate short-term traffic flow prediction is important for the operation and management of expressways. This paper presents a hybrid model that can be used to discover the spatiotemporal dependencies of traffic flows and, thus, achieve a more accurate traffic flow forecast. This model stacks a full connection (FC) layer, two-layer long short-term memory (LSTM), and a middle mean pooling layer, denoted by FC-LSTM, to expand the ability of LSTM to capture spatial correlations and too long-term temporal dependencies of traffic flows. Traffic data of 51 toll stations (accounting for 15% of all stations) in the Guizhou expressway network in China in January 2016 were used to evaluate the validation of FC-LSTM. The results showed that the traffic flows at adjacent tollgates were interactional and correlated, that FC-LSTM could capture this spatial dependency and the temporal correlations of traffic flows, and, thus, that it was superior to other baselines with a low prediction error, high precision, and high fitting degree. Moreover, FC-LSTM is interpretable and robust owing to its explicit input and is suitable for traffic flow prediction for most tollgates under the same parameters.

A Bidirectional Context-Aware and Multi-Scale Fusion Hybrid Network for Short-Term Traffic Flow Prediction

Short-term traffic flow prediction is to automatically predict the traffic flow changes in a period of future time based on the extraction of the spatiotemporal features in the road network. For governments, timely and accurate traffic flow prediction is crucial to plan road manage-ment and improve traffic efficiency. Recent advances in deep learning have shown their dominance on short-term traffic flow prediction. However, previous methods based on deep learning are mainly limited to temporal features and have so far failed to predict the bidirectional con-textual spatiotemporal relationship correctly. Besides, the precision and the practicality are limited by the road network scale and the single time scale. To remedy these issues, a Bidirectional Context-aware and Multi-scale fusion hybrid Network (BCM-Net) is proposed, which is a novel short-term traffic flow prediction framework to predict timely and accurate traffic flow changes. In BCM-Net, the Bidirectional Context-aware (BCM) block is added to the feature extraction structure to effective-ly integrate spatiotemporal features. The Interpolation Back Propagation sub-network is used to merge multi-scale information, which further improves the robustness of the model. Experiment results on diverse datasets demonstrated that the proposed method outperformed the state-of-the-art methods.

A Multifeature Fusion Short-Term Traffic Flow Prediction Model Based on Deep Learnings

Short-term traffic flow prediction is an important component of intelligent transportation systems, which can support traffic trip planning and traffic management. Although existing predicting methods have been applied in the field of traffic flow prediction, they cannot capture the complex multifeatures of traffic flows resulting in unsatisfactory short-term traffic flow prediction results. In this paper, a multifeature fusion model based on deep learning methods is proposed, which consists of three modules, namely, a CNN-Bidirectional GRU module with an attention mechanism (CNN-BiGRU-attention) and two Bidirectional GRU modules with an attention mechanism (BiGRU-attention). The CNN-BiGRU-attention module is used to extract local trend features and long-term dependent features of the traffic flow, and the two BiGRU-attention modules are used to extract daily and weekly periodic features of the traffic flow. Moreover, a feature fusion layer in the model is used to fuse the features extracted by each module. And then, the number of neurons in the model, the loss function, and other parameters such as the optimization algorithm are discussed and set up through simulation experiments. Finally, the multifeature fusion model is trained and tested based on the training and test sets from the data collected from the field. And the results indicate that the proposed model can better achieve traffic flow prediction and has good robustness. Furthermore, the multifeature fusion model is compared and analyzed against the baseline models with the same dataset, and the experimental results show that the multifeature fusion model has superior predictive performance compared to the baseline models.

Short-Term Traffic Flow Prediction with Structure-Optimized Deep Belief Network

This paper proposes a structure-optimized deep belief network method for short-term traffic flow forecast, which is used to solve the problems of too simple training data in deep learning short-term traffic flow forecast and random selection of model structure construction parameters. We constructed a deep belief network short-term traffic flow forecast model that can simultaneously train three types of traffic data related to the predicted node traffic volume, enhance the spatiotemporal correlation of predictions, and overcome the shortcomings of too single training data. At the same time, we optimize the short-term traffic flow prediction model structure of the deep belief network; and use the T-PSO algorithm to optimize the hidden layer structure parameters of the prediction model. It can avoid the decrease in the practicability of the model caused by the random selection of the model structure construction parameters. The experimental results show that the method of structure-optimized deep belief network is feasible and effective, and the prediction accuracy is better than the classic deep learning prediction model.