Historical Traffic Patterns Research Articles

Traffic Flow Prediction in Smart City

Traffic flow prediction in smart cities plays a crucial role in enhancing urban mobility, reducing congestion, and optimizing transportation systems. In this study, we leverage deep learning techniques to develop accurate and reliable traffic flow prediction models. We collect and preprocess real-time and historical traffic data from various sources, including traffic sensors, GPS devices, and traffic management systems. Through feature engineering, we extract relevant spatiotemporal features such as time of day, day of week, weather conditions, and historical traffic patterns. We then design and train deep learning architectures, including recurrent neural networks (RNNs) such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), as well as convolutional neural networks (CNNs), to capture complex patterns and relationships in the data. Our models are evaluated using appropriate metrics such as mean absolute error (MAE) and root mean squared error (RMSE) on validation and test sets. Once deployed, the models contribute to proactive traffic management, informing decisionmaking processes for urban planners, transportation agencies, and commuters alike. Through continuous monitoring and maintenance, we ensure the scalability, reliability, and responsiveness of the prediction system to adapt to evolving traffic conditions and urban dynamics. Our findings demonstrate the effectiveness of deep learning in traffic flow prediction, contributing to the development of smarter and more sustainable cities.

Air Traffic Flow Prediction with Spatiotemporal Knowledge Distillation Network

Air Traffic Flow Prediction with Spatiotemporal Knowledge Distillation Network

A PSO-ANN Hybrid Model for Temporal Traffic Speed Forecasts in Urban Environments

Intelligent traffic systems (ITS) are being explored to make traffic systems more efficient based on data collection and analytics. One of the major applications to avoid congestions in traffic speed forecasting. Traffic speed forecasting using machine learning is a crucial aspect of intelligent transportation systems, aiming to enhance traffic management and optimize the overall efficiency of urban mobility. This process involves the utilization of various data sources, such as historical traffic patterns, real-time sensor data, and environmental factors, to predict future traffic speeds. Machine learning models play a pivotal role in this domain, as they can analyze complex patterns and relationships within the data to generate accurate predictions. This paper presents a Particle Swarm Optimization (PSO)-Neural network model for traffic speed forecasting. Contrary to conventional neural network models, the PSO is used to adaptively update the network weights. The results clearly indicate that the proposed approach outperforms existing baseline approaches in terms of MAPE and Forecasting Accuracy. Keywords: Intelligent Traffic System (ITS), Particle Swarm Optimization (PSO), Artificial Neural Network (ANN), Mean Absolute Percentage Error, Regression.

Models for forecasting the traffic flow within the city of Ljubljana

Efficient traffic management is essential in modern urban areas. The development of intelligent traffic flow prediction systems can help to reduce travel times and maximize road capacity utilization. However, accurately modeling complex spatiotemporal dependencies can be a difficult task, especially when real-time data collection is not possible. This study aims to tackle this challenge by proposing a solution that incorporates extensive feature engineering to combine historical traffic patterns with covariates such as weather data and public holidays. The proposed approach is assessed using a new real-world data set of traffic patterns collected in Ljubljana, Slovenia. The constructed models are evaluated for their accuracy and hyperparameter sensitivity, providing insights into their performance. By providing practical solutions for real-world scenarios, the proposed approach offers an effective means to improve traffic flow prediction without relying on real-time data.

Prediction model for short‐term traffic flow based on a K‐means‐gated recurrent unit combination

Short-term forecasting of traffic flow is an indispensable part of easing traffic pressure. Considering that different traffic flow patterns will affect the short-term traffic flow prediction results, a combined method based on the K-means clustering algorithm and gated recurrent unit (GRU) is proposed to build a short-term traffic flow prediction model to overcome the above problems. The K-means algorithm is used to cluster historical traffic flow data to establish different traffic flow pattern libraries. The K-nearest neighbour (KNN) classification algorithm is used to determine the historical traffic flow pattern most similar to the traffic flow change trend of the date to be predicted. All historical traffic flow data in this category is used training samples to make targeted predictions. The traffic flow data of performance measurement system (PeMS) in California, USA is used to verify the performance of the proposed model. Compared with the GRU network, stacked auto encoders (SAEs), random forest (RF), and support vector machine regression (SVR), the results show that the proposed combination model K-means-GRU considers the diversity of traffic flow patterns and improves the prediction accuracy, it can better solve the short-term traffic flow prediction problem.

Airport Passenger Traffic Forecast: An Exploratory Study

Sultan Hasanuddin International Airport has been an important hub airport in Indonesia for decades, connecting traffic between west and east Indonesia as well as functioned as international gate in East Indonesia along with Sam Ratulangi Airport. Analysing characteristics of historical traffic data pattern, determining factors affecting past behaviour and building the best-fit model to forecast future traffic are critical for airport operator. Several forecast techniques are employed including Holt-Winter, Decomposition Method and Econometric Model. Moreover, trend, seasonal event and irregular phenomena from past data are observed to analyse traffic behaviour. Passenger traffic data from 1995 up to 2015 is utilized to predict future traffic until 2020. Validation of selected forecast model is conducted by implementing backtesting method which shows that the model successful foretell annual passenger movement with estimation average deviation around 0.5%. Some potential risks and opportunities as well as potential route expansion are identified to fortify future challenges.

Nonintrusive Elevator System Fault Detection Using Learned Traffic Patterns

A new method for nonintrusive elevator fault detection is presented. A computationally efficient algorithm for implementing the method is also proposed. The method is employed to detect when the elevator has been stationary for an unusually long period of time compared to historical traffic load patterns. This information can be used for fault detection but also indirectly to monitor the condition of the doors. The traffic load on the elevator is modeled as a nonhomogeneous Poisson process, and a generalized linear model is used to describe how the intensity of the process varies over time. A statistical hypothesis test is then used to determine if the elevator has been stationary for an unusually long time. The application of the proposed method is illustrated by an example where the detected faults are compared with the elevator service log. All faults were detected long before the service company was notified by the facility owner. Furthermore, based on the evaluation of 30 weeks of data, the method achieves a precision of 0.82 at a recall probability of 0.80.

Short-Term travel speed prediction for urban expressways using convolutional neural network and tensor decomposition

Prediction is an important part of the traffic management system (TMS) which supports route planning, dynamic traffic control, and information provision. We developed a multi-dimensional learning machine for predicting the traffic speed. Proposed methodology considered both historical experience and near past observation of traffic data by combining a convolutional neural network (CNN) with tensor decomposition (TD) predictor. TD based method is treated as an effective traffic predictor considering temporal-spatial neighbourhood data. However, limited by learning mechanism, such a method cannot extract historical traffic pattern from large datasets. Our main contribution is converting the traffic prediction into a tensor imputation problem, which considers the historical pattern information from CNN by constructing input tensor. Besides, multiple low-rank choosing and weighted optimization are introduced to improve the accuracy of TD-based prediction. We validated our methodology using empirical detector data of urban expressway in Shanghai. Compared with single algorithms, our method has smaller absolute and relative error.

On-demand assessment of air traffic impact of blocking airspace

ABSTRACTThe Federal Aviation Administration often blocks strategically located airspace volumes to ensure safety during a variety of operations that are potentially hazardous to aircraft, such as space launches. As the frequency of these operations increases, there is a growing need to deepen collaboration and transparency between stakeholders regarding the use of airspace. This collaboration can be supported by models and capabilities to quickly assess the impact of airspace closures, up to 12 months into the future. This paper presents a technique to enable a ‘what-if’ analysis capability coupled with a prediction model, whereby changes in airspace dimension, location, and activation time are reflected instantaneously as measures of projected impact. The technique can also be used for quick post-operations analysis using historical traffic data and to develop air traffic impact assessment capabilities accessible to a broad range of users outside of the air traffic domain. This research has three key components: developing a model to predict air traffic demand up to 12 months into the future, modelling air traffic impact to the affected traffic, and reducing this information into a data structure that can support on-demand analysis. The focus of this paper is on new techniques to predict demand using a large set of historical track data and further encode these projections to support the quick assessment of the impact of blocking various airspace volumes. Initial results show that the proposed data reduction scheme accurately represented the traffic crossing an airspace and resulted in data size reduction by over 50%. The projection model performed well, the actual number of impacted flights were within the estimated range of approximately 80% of the time. Finally, the responsiveness of the web-based prototype developed to illustrate the concept demonstrated the model’s ability to support an on-demand assessment of the air traffic impact of blocking airspace. A significant limitation of the projection model is that it is based on the historical traffic pattern within the U.S. airspace; separate analysis is needed to adapt it to other geographical location.

Bus arrival time prediction based on mixed model

How to predict the bus arrival time accurately is a crucial problem to be solved in Internet of Vehicle. Existed methods cannot solve the problem effectively for ignoring the traffic delay jitter. In this paper, a three-stage mixed model is proposed for bus arrival time prediction. The first stage is pattern training. In this stage, the traffic delay jitter patterns (TDJP) are mined by K nearest neighbor and K-means in the historical traffic time data. The second stage is the single-step prediction, which is based on real-time adjusted Kalman filter with a modification of historical TDJP. In the third stage, as the influence of historical law is increasing in long distance prediction, we combine the single-step prediction dynamically with Markov historical transfer model to conduct the multi-step prediction. The experimental results show that the proposed single-step prediction model performs better in accuracy and efficiency than short-term traffic flow prediction and dynamic Kalman filter. The multi-step prediction provides a higher level veracity and reliability in travel time forecasting than short-term traffic flow and historical traffic pattern prediction models.

Improvement of Search Strategy With K-Nearest Neighbors Approach for Traffic State Prediction

Having access to the future traffic state information is crucial in maintaining successful intelligent transportation systems (ITS). However, predicting the future traffic state is a challenging research subject involving prediction reliability issues. Predictive performance measures, including the accuracy, efficiency, and stability, are generally considered as the most important priorities in the evaluation of prediction modules. Researchers have developed various K-nearest-neighbors-based searching algorithms that find the future state from the historical traffic patterns. Interestingly, there has not been sufficient effort made for improving the performance. For the emerging big data era, incorporating an efficient search strategy has become increasingly important since the applicability of the prediction module in ITS heavily relies on the efficiency of the searching method used. This paper develops a novel sequential search strategy for traffic state predictions. The proposed sequential strategy is found to be outperforming the conventional single-level search approach in terms of prediction measures, which are prediction accuracy, efficiency, and stability. Compared with the conventional approach, the proposed sequential method yields significantly more accurate results via internal hierarchical improvements across sublevels while maintaining excellent efficiency and stability.

Data Mining and Pattern Matching for Dynamic Origin–Destination Demand Estimation

Historical traffic data are widely used in the estimation of origin– destination (O-D) demand patterns in simulation-based dynamic traffic assignment models, in the prediction of traffic states, and as a basis for defining traffic management scenarios. This study investigated the determination of historical traffic patterns by applying a classical clustering algorithm to a very large data set of sensor observations over an extended period and identified appropriate patterns for use in the application of traffic estimation and prediction systems. Systematic identification of similarity and dissimilarity of traffic flow data can lead to a systematic process for defining critical demand scenarios for traffic state prediction. The objective was to explore the impact of various demand scenarios on real-time traffic estimation and prediction. A detailed procedure for clustering traffic flow data that is based on the K-means clustering algorithm is presented. The procedure was applied to a subnetwork in Salt Lake City, Utah. An O-D estimation procedure was applied to each cluster separately; this procedure repeatedly solved a bilevel optimization problem in which the goal was to minimize the deviation from observations and historical demand. For the evaluation of the effect of selecting a best-matching initial matrix on traffic pattern estimation and prediction quality, a comparative application of initial matrix choices was conducted with an online traffic estimation and prediction system. The results indicate that the clustering process results in better starting matrices matched to the day's unfolding traffic and weather conditions.

Travel Time Prediction of Road Network Based on Multi-Source Data Fusion

This paper is concerned with the task of travel time pre-diction of urban roadway. For improving the travel time predication ac-curacy, a travel time predication model based multi-source data fusion is proposed. The prediction procedure is divided into two phases, the estimation phase and the prediction phase The method is combined the historical traffic patterns with real-time traffic data as a linear. The resulting model is tested with realistic traffic data, and is found to perform well.

Improving short-term travel time prediction for on-line car navigation by linearly transforming historical traffic patterns to fit the current traffic conditions

This research is focused on the problem of travel time prediction for a personal on-line car navigation system. The aim of this study is to improve the short-term travel time prediction quality by creating a dynamic model that utilises the real-time GPS floating car data (from the users of a car navigation system), assuming a static black box model of historical traffic patterns is given as a base. A novel model is introduced for this task. It combines two methods: the first one (applied on the city level) is based on linearly transforming traffic patterns to fit the current traffic conditions by solving a weighted and regularised regression problem. The second method (applied on short road segments) is based on exponential smoothing. The models quality is evaluated through extensive experiments on real data, by measuring the squared prediction error on the chosen observations that has not previously influenced the examined model. The results show significant improvement over the static historical traffic patterns model.

Urban traffic signal control using reinforcement learning agents

This study presents a distributed multi-agent-based traffic signal control for optimising green timing in an urban arterial road network to reduce the total travel time and delay experienced by vehicles. The proposed multi-agent architecture uses traffic data collected by sensors at each intersection, stored historical traffic patterns and data communicated from agents in adjacent intersections to compute green time for a phase. The parameters like weights, threshold values used in computing the green time is fine tuned by online reinforcement learning with an objective to reduce overall delay. PARAMICS software was used as a platform to simulate 29 signalised intersection at Central Business District of Singapore and test the performance of proposed multi-agent traffic signal control for different traffic scenarios. The proposed multi-agent reinforcement learning (RLA) signal control showed significant improvement in mean time delay and speed in comparison to other traffic control system like hierarchical multi-agent system (HMS), cooperative ensemble (CE) and actuated control.

Travel Time Prediction

As reported in the literature for the applications of intelligent transportation systems with traffic detectors, various missing data patterns are frequently observed in such systems and may dramatically degrade their performance. This study presents two imputation approaches for contending with the missing data issues in travel time prediction. The first model is based on the concept of multiple imputation technique to predict directly the travel times under various missing data patterns. The second model that serves as the supplemental component is to estimate the missing detector values using neighboring detector data and historical traffic patterns. Both models have been incorporated with reliability indicators so as to assess the quality of imputed data and its applicability for use in prediction. The numerical example based on 10 roadside detectors on I-70 in Maryland has demonstrated that both developed models outperformed existing methods and offer the potential for field implementation.

Mutational bias suggests that replication termination occurs near the dif site, not at Ter sites: what's the Dif?

In this issue of Molecular Microbiology, Hendrickson and Lawrence analyse the sequence of bacterial genomes to map the historical traffic pattern of chromosome replication. Their surprising conclusion is that most forks terminate at the dif site rather than at the Tus/Ter sites where most investigators have concluded termination occurs most frequently. What make this analysis novel are the methods and the revisionist hypotheses for how and why forks might stop at dif.

Surveillance strategies for a class of adaptive-routing algorithms in circuit-switched DNHR communications networks: a simulation study

A large, circuit-switched, communications network engineered for historical traffic patterns using dynamic nonhierarchical routing (DNHR) is considered. During overloads, the offered traffic many not follow the historical patterns. In such situations, it is possible to increase network throughput by augmenting the routing tables in near-realtime to utilize instantaneous spare capacity. The surveillance strategy determines how often network data is collected and how frequently additional paths are added to the routing tables. Six surveillance strategies for adaptive routing are examined: (1) 2.5-min surveillance interval; (2) 5-min surveillance interval; (3) 10-min surveillance interval; (4) 15-min surveillance interval; (5) modify the routing every other 5 min; and (6) modify the routing every 5 min with a fixed 10-min lag before the routing tables are changed. Simulation is used to compare the network performance to the baseline performance without adaptive routing for each of the strategies. It is found that the more adaptive strategies, (1), (2), and (6), perform better during periods with variable traffic, and that the less adaptive strategies perform better during periods with stable traffic. During heavy overloads, almost as important as the surveillance strategy is the number of problems, overflowing node pairs, and the algorithm attempts to relieve. >