Future Travel Times Research Articles

Periodic Attention-based Stacked Sequence to Sequence framework for long-term travel time prediction

Travel time analysis and prediction are keystones for building intelligent transportation systems in the new era, which has gained wide attention from the research community. Over the last few years, deep learning-based approaches have demonstrated the capability to predict travel time with big data. However, the existing works focused on predicting short-term travel time for paths and roads. In fact, knowing a period of long-term future travel time is a demand with important applications like traffic management and schedule routing planning. Nevertheless, studies on long-term traffic prediction are still very limited due to the complicated underlying factors. In this paper, we propose a novel deep learning-based framework named Periodic Attention-based Stacked Sequence to Sequence (PASS2S), which aims to address the long-term traffic prediction problem. PASS2S consists of two main components, namely periodic segment generation and attention-based stacked sequential prediction. To extract periodic information, we design a periodic segment generation component to capture historical periodic segments from the traffic data. To reduce the error propagation and improve long-term prediction accuracy, we propose an attention-based stacked prediction component to model long short-term and short-term dependencies from the periodic segments. We conducted a series of experiments on a real-world travel time dataset and the experimental results show that our proposed approach outperforms the state-of-the-art competing methods in terms of various metrics like MAE, RMSE, and SMAPE. To the best of our knowledge, this is the first work that considers attentive periodic historical information for solving the long-term travel time prediction with a period of the future, which has not been well studied in the research community.

Constructing Cooperative Intelligent Transport Systems for Travel Time Prediction With Deep Learning Approaches

Drivers and traffic system planners require accurate forecasting of future travel times. To alleviate traffic congestion on Taiwan’s Provincial Highway No. 61, this study considers temporal (weekdays, weekends or continuous holidays) and spatial characteristics (road types), and establishes short-term and long-term travel time prediction models. Data pre-processing is accomplished using the Google Maps API and floating car method to verify travel time comparisons, finding that post-processing travel times are reliable, credible and reasonable. This study develops a collaborative intelligent transportation system (CITS) based on 9 different algorithms for the prediction of current and future travel-time. The results show that <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula> NN-R provides the most accurate short-term predictions, with average prediction error within 20 seconds per kilometer. For long-term forecasting, SARIMAX and fbProphet provide the most accurate results for weekday and continuous holiday modes. Travel time prediction can assist traffic management agencies in the timely implementation of appropriate traffic management measures. CITS also provides real-time traffic query and travel time prediction functions to help drivers avoid traffic congestion.

Prediction of Travel Time Reliability on Interstates Using Linear Quantile Mixed Models

Under the Moving Ahead for Progress in the 21st Century Act (MAP-21), state Departments of Transportation (DOTs) are responsible for reporting travel time reliability and also setting targets and showing progress toward those targets. To know how to improve travel time reliability and what to expect from investments in transportation infrastructure, state DOTs need a better understanding of the factors that affect travel time reliability and methods to predict future travel time reliability. This paper proposes linear quantile mixed models (LQMMs) to quantify travel time reliability impact factors and predict selected reliability measures (level of travel time reliability [LOTTR] and the 90th percentile) to address these needs. The method was demonstrated using probe vehicle data from interstate segments in Virginia that had been partitioned into approximately homogeneous clusters based on the similarity of their cumulative distribution functions (CDFs) of travel times. Using clustered data meant that LQMMs were only necessary for a limited number of clusters rather than for hundreds of individual segments, thus making the process more efficient and manageable. The LQMMs showed that frequencies of non-recurrent events, such as incidents and weather, were correlated with higher travel time percentiles. The prediction performance of LQMMs was compared with trend line predictions, a common method used in practice. The results showed that LQMMs significantly improved prediction accuracy.

An Adaptive Neural Fuzzy Inference System model for freeway travel time estimation based on existing detector facilities

While volume-counting detectors are currently installed in many freeways and have the capabilities of collecting high-resolution traffic information, travel time estimations are usually only based on spot speeds collected by the detectors, which could only produce rough estimates of travel time during peak hours when roadways are congested. In addition to spot speeds, these detectors provide directional volume counts and occupancy, which are useful for a better travel time estimation. This research used data from volume counting detectors and made Adaptive Neural Fuzzy Inference System (ANFIS) to automatically estimate real-time travel times. ANFIS’s ability to learn traffic patterns allows accurate and reliable real-time travel time estimation even with missing or corrupted data. ANFIS also proves to be a powerful tool for estimating future travel times on freeways. For easier use by practitioners and researchers, the method used in this study was implemented into a software package, named FTTE (Freeway Travel Time Estimator). Finally, a case study was conducted using the new approach and the results were compared for both congested and uncongested traffic conditions.

Hybrid time series forecasting methods for travel time prediction

Providing accurate information about travel time to passengers is important in public transportation. In this aspect, the travel time of buses between two consecutive stops can be handled as time series. Then, the future travel time can be predicted using time series forecasting methods. In this study, we propose a novel method with three-layer architecture to predict bus travel time between two stops. In the first layer of the proposed method, initial prediction is made by processing measured data. In the second layer, residuals are predicted in the specified depth. In the third layer, the final prediction is made by integrating the results of two previous layers with three different approach. The experiments were performed on the data, which were obtained from public transportation of Istanbul, using various time series forecasting methods in form of traditional and proposed architecture. The results show that proposed method outperforms traditional approach with approximately MAPE of 6.

Traffic congestion and travel time prediction based on historical congestion maps and identification of consensual days

In this paper, a new practice-ready method for the real-time estimation of traffic conditions and travel times on highways is introduced. First, after a principal component analysis, observation days of a historical dataset are clustered. Two different methods are compared: a Gaussian Mixture Model and a k-means algorithm. The clustering results reveal that congestion maps of days of the same group have substantial similarity in their traffic conditions and dynamic. Such a map is a binary visualization of the congestion propagation on the freeway, giving more importance to the traffic dynamics. Second, a consensus day is identified in each cluster as the most representative day of the community according to the congestion maps. Third, this information obtained from the historical data is used to predict traffic congestion propagation and travel times. Thus, the first measurements of a new day are used to determine which consensual day is the closest to this new day. The past observations recorded for that consensual day are then used to predict future traffic conditions and travel times. This method is tested using ten months of data collected on a French freeway and shows very encouraging results.

Probe data-driven travel time forecasting for urban expressways by matching similar spatiotemporal traffic patterns

Travel time is an effective measure of roadway traffic conditions. The provision of accurate travel time information enables travelers to make smart decisions about departure time, route choice and congestion avoidance. Based on a vast amount of probe vehicle data, this study proposes a simple but efficient pattern-matching method for travel time forecasting. Unlike previous approaches that directly employ travel time as the input variable, the proposed approach resorts to matching large-scale spatiotemporal traffic patterns for multi-step travel time forecasting. Specifically, the Gray-Level Co-occurrence Matrix (GLCM) is first employed to extract spatiotemporal traffic features. The Normalized Squared Differences (NSD) between the GLCMs of current and historical datasets serve as a basis for distance measurements of similar traffic patterns. Then, a screening process with a time constraint window is implemented for the selection of the best-matched candidates. Finally, future travel times are forecasted as a negative exponential weighted combination of each candidate’s experienced travel time for a given departure. The proposed approach is tested on Ring 2, which is a 32km urban expressway in Beijing, China. The intermediate procedures of the methodology are visualized by providing an in-depth quantitative analysis on the speed pattern matching and examples of matched speed contour plots. The prediction results confirm the desirable performance of the proposed approach and its robustness and effectiveness in various traffic conditions.

Dynamic Routing: Anticipation of Emission-Sensitive Traffic Management

Abstract: Delivery companies are affected by emission-sensitive traffic management systems. These systems are installed in cities to react instantly to emission hot-spots through adapting traffic light programs at intersections. This results in a change of the travel times for the delivery vehicle. In this paper, we model the problem in a dynamic vehicle routing problem with stochastic transition of deterministic travel time matrices. To solve this problem, we apply approximate value iteration, a method of approximate dynamic programming, to anticipate future travel time matrix changes in dynamic routing decisions. We vary the approach in the level of information about the state of the traffic management system. This allows to distinguish the required information for the routing decisions. Further, we compare the dynamic and anticipatory routing policies with a static a priori routing. Computational studies show an improved tour duration of routing with traffic management information over the a priori routing by up to 6.5%. We further show that an efficient representation of the traffic management system status in the approximate value iteration approach is mandatory to achieve sufficient anticipation.

Taxis as a Recourse Option for Ridesharing Services

Ridesharing services commonly offered at airports and to mobility-impaired persons exist worldwide. Unlike operators in conventional public transit, ridesharing operators maintain flexible routes and schedules that are used in providing door-to-door service. These routes and schedules are typically created by using the assumption of perfect foresight about future travel times. In practice, however, vehicles may encounter traffic congestion, be delayed because of road work or an accident, or even break down. Thus, the routes and schedules may adhere to quality-of-service constraints in planning when perfect conditions are assumed, but operationally they often violate such constraints. This paper investigates the effectiveness of routes and schedules created under a deterministic travel time assumption. The paper evaluates practical solutions that force the schedules to include excess slack. Finally, the paper proposes and assesses a recourse option for use in actual operations of assigning to taxis those passengers who are likely to be picked up late. The evaluation was completed on a real-world case study in Richmond, Virginia.

Dynamic Short-Term Projections of Travel Time Distributions in Urban Signalized Networks utilizing Composite Information of Traffic Characteristics

Abstract: The estimation of experienced travel-time in road networks stands for an important feature in Advanced Traveler Information Systems (ATIS). In the era of data availability, the dissemination of accurate traffic information to travelers could have a huge impact on their trip choices and thus in systems’ performance. The scope of this paper is time travel modeling and prediction based on alternative configurations of mixed duration and time series models. The proposed framework enables the dynamic probabilistic estimation of travel time between a predefined set of Origin-Destinations (O-D) locations, by taking into account available spatial data. For achieving that, suitable formulated hazard models are combined with time series input, forming a hybrid methodological framework and compared to two different theoretical duration models. The resulted mechanism provides valuable probabilistic estimations for travel time for O-D pairs, able to cover extensive urban road networks. In particular, the proposed modeling approach uses past travel time observations as inputs and provides distributions of future travel times. The proposed application is tested on a realistic road network, namely that of Cyprus. The aim of the study is to provide reliable travel information to travelers in a methodologically sound, straightforward and comprehensible way.

On Intention-Propagation-Based Prediction in Autonomously Self-adapting Navigation

It is widely believed that road traffic as a whole self-adapts to the current situation to make travel times shorter by avoiding congestions, if the autonomously operating navigation devices exploit real-time traffic information. The classical theoretical models do not have definite answer if car navigation based on real-time data is able to self-adapt and produce better traffic, or not. The novel theoretical approach to study this belief is the online routing game model. Current commercial car navigation systems are modelled with the class of simple naive online routing games. It is already proved that simple naive online routing games may show undesirable phenomena. One of the approaches to improve car navigation is intention-propagation-based prediction, where agents share their intention, and can forecast future travel times. In this paper we prove that in spite of exploiting this type of prediction in online routing games, the phenomena studied in simple naive online routing games are still possible, although in a different way. With these theoretical results we point out hitherto hidden pivotal phenomena of intention-propagation-based prediction in collective adaptive systems.

A space–time diurnal method for short-term freeway travel time prediction

A number of short-term travel time prediction approaches have been developed in the past decade. However, few studies take into account spatial and temporal travel time information simultaneously in the prediction approach. In this study, we proposed a space–time diurnal (ST-D) method, which merges the spatial and temporal travel time information to obtain accurate short-term travel time prediction for freeway corridors under different traffic conditions. The proposed approach can take into account important characteristics of travel times: spatial and temporal correlation, diurnal pattern, and the nonnegativity of the travel time. We use two distributions to model the 5-min average travel time: a truncated normal distribution and a lognormal distribution. Contrary to the most existing methods that yield a point prediction of short-term travel time, this probabilistic modeling approach can overcome the drawbacks of the point prediction by fitting a probability distribution to describe the uncertainty of the future travel times, and from which prediction intervals can be calculated. We use minimum continuous ranked probability score (CRPS) estimation to numerically estimate the parameters in the prediction models. The ST-D method is examined using the travel time data collected on a segment along the US-290 in Houston, Texas. The proposed method provides prediction of travel time over 5-min intervals for up to 1h in advance. It was found that travel time data from neighboring links along the freeway corridors can be efficiently used to obtain reliable short-term prediction of travel time. The study results suggest the ST-D method is more robust than the traditional vector autoregressive models.

A Robust Scenario Approach for the Vehicle Routing Problem with Uncertain Travel Times

We consider a vehicle routing problem with uncertain travel times in which a penalty is incurred for each vehicle that exceeds a given time limit. A traditional stochastic programming approach would require precise knowledge of the underlying probability distributions of random data. In a novel approach presented here, we assume that only rough information on future travel times is available, leading to the multiple range forecasts of travel times and the probabilities of each range being realized. In this setting, we replace the point estimates of travel times on a scenario by range estimates. For each scenario, we then find the robust routes that protect the solution against the worst case within the given ranges, and finally we find the routes with the minimum expected cost. We propose a branch-and-cut algorithm to solve the problem and report computational results on both randomly generated and the well-known Solomon's instances. The results demonstrate that our approach is a favorable one when exact information of probability distributions is not available.

Experienced travel time prediction for congested freeways

Travel time is an important performance measure for transportation systems, and dissemination of travel time information can help travelers make reliable travel decisions such as route choice or departure time. Since the traffic data collected in real time reflects the past or current conditions on the roadway, a predictive travel time methodology should be used to obtain the information to be disseminated. However, an important part of the literature either uses instantaneous travel time assumption, and sums the travel time of roadway segments at the starting time of the trip, or uses statistical forecasting algorithms to predict the future travel time. This study benefits from the available traffic flow fundamentals (e.g. shockwave analysis and bottleneck identification), and makes use of both historical and real time traffic information to provide travel time prediction. The methodological framework of this approach sequentially includes a bottleneck identification algorithm, clustering of traffic data in traffic regimes with similar characteristics, development of stochastic congestion maps for clustered data and an online congestion search algorithm, which combines historical data analysis and real-time data to predict experienced travel times at the starting time of the trip. The experimental results based on the loop detector data on Californian freeways indicate that the proposed method provides promising travel time predictions under varying traffic conditions.

Short-Term Prediction of Travel Time Using DSRC on Highway

현재까지 통행시간 예측과 관련된 다양한 연구들이 수행되었지만, 한국고속도로 특성에 맞는 예측방법론에 대한 실증연구는 부족한 실정이다. 이에 본 연구에서는 실제 통행시간을 기반으로한 DSRC 자료를 바탕으로 한국고속도로에 적절한 예측방법론을 도출한다. 경부고속도로 안성 JC~오산IC 구간의 24시간 DSRC 자료를 이용하며 단주기 통행시간 예측 및 비선형 관계에서 높은 정확도를 보이는 인공신경망 기법을 적용한다. 이어서 랜덤난수를 이용한 통행시간 예측결과의 정확도 검증을 실시한다. 통행시간 예측결과 오차율이 약 4%로 우수한 예측력을 보였으며, 이는 패턴기반 인공신경망 예측시 이력자료의 전처리 과정과 최적의 입력층 및 은닉층의 선정으로 인한 결과로 판단된다. 통행시간 예측결과의 검증을 위해서 랜덤난수를 이용하였으며, 랜덤난수가 이력자료 패턴에 포함되지 않은 경우 실측치와의 오차율이 18.98%로 높게 도출되었다. 이는 인공신경망을 이용한 통행시간 예측시 패턴DB가 예측의 정확도에 주요하게 작용한 결과로 판단된다. 본 연구의 결과를 통해서 한국고속도로 특성에 맞는 통행시간 예측 및 정보제공이 가능할 것으로 판단된다. This paper develops a travel time prediction algorithm that can be used for real-time application. The algorithm searches for the most similar pattern in historical travel time database as soon as a series of real-time data become available. Artificial neural network approach is then taken to forecast travel time in the near future. To examine the performance of this algorithm, travel time data from Gyungbu Highway were obtained and the algorithm is applied. The evaluation shows that the algorithm could predict travel time within 4% error range if comparable patterns are available in the historical travel time database. This paper documents the detailed algorithm and validation procedure, thereby furnishing a key to generating future travel time information.

Integrated Traffic and Communication Performance Evaluation of an Intelligent Vehicle Infrastructure Integration (VII) System for Online Travel-Time Prediction

This paper presents a framework for online highway travel-time prediction using traffic measurements that are likely to be available from vehicle infrastructure integration (VII) systems, in which vehicle and infrastructure devices communicate to improve mobility and safety. In the proposed intelligent VII system, two artificial intelligence (AI) paradigms, i.e., artificial neural networks (ANNs) and support vector regression (SVR), are used to determine future travel time based on such information as the current travel time and VII-enabled vehicles' flow and density. The development and performance evaluation of the VII-ANN and VII-SVR frameworks, in both the traffic and communications domains, were conducted using an integrated simulation platform for a highway network in Greenville, SC. In particular, the simulation platform allows for implementing traffic surveillance and management methods in the traffic simulator PARAMICS and for evaluating different communication protocols and network parameters in the communication network simulator, Network Simulator version 2 (ns-2). This paper's findings reveal that the designed communications system can support the travel-time prediction functionality. The findings also demonstrate that the travel-time prediction accuracy of the VII-AI framework was superior to a baseline instantaneous travel-time prediction algorithm, with the VII-SVR model slightly outperforming the VII-ANN model. Moreover, the VII-AI framework was shown to perform reasonably well during nonrecurrent congestion scenarios, which have traditionally challenged sensor-based highway travel-time prediction methods.

MACHINE LEARNING ON CONGESTION ANALYSIS BASED REAL-TIME NAVIGATION SYSTEM

In the metropolitan region, most congestion or traffic jams are caused by the uneven distribution of traffic flow that creates bottleneck points where the traffic volume exceeds the road capacity. Additionally, unexpected incidents are the next most probable cause of these bottleneck regions. Moreover, most drivers are driving based on their empirical experience without awareness of real-time traffic situations. This unintelligent traffic behavior can make the congestion problem worse. Prediction based route guidance systems show great improvements in solving the inefficient diversion strategy problem by estimating future travel time when calculating accurate travel time is difficult. However, performances of machine learning based prediction models that are based on the historical data set degrade sharply during a congestion situation. This paper develops a new navigation system for reducing travel time of an individual driver and distributing the flow of urban traffic efficiently in order to reduce the occurrence of congestion. Compared with previous route guidance systems, the results reveal that our system, applying the advanced multi-lane prediction based real-time fastest path (AMPRFP) algorithm, can significantly reduce the travel time especially when drivers travel in a complex route environment and face frequent congestion problems. Unlike the previous system,1 it can be applied either for single lane or multi-lane urban traffic networks where the reason for congestion is significantly complex. We also demonstrate the advantages of this system and verify the results using real highway traffic data and a synthetic experiment.

Updating Paths in Time-Varying Networks Given Arc Weight Changes

Many transportation applications, including applications in intelligent transportation systems, require the solution of a series of shortest path problems in which only the travel time along a set of arcs of the network change from one problem instance to the next. One could use an existing path algorithm to solve each problem instance independently as it arises. However, significant savings in computation time can often be achieved through the use of a reoptimization algorithm that would begin from the prior solution in determining the updated optimal solution for the given arc travel-time changes. Such quick solution is critical for providing routing instructions to travelers in real time as travel-time information is retrieved from the traffic network. Numerous works have presented reoptimization techniques for use in updating shortest path trees in deterministic and static networks; however, it appears that no reoptimization technique exists in the literature for updating paths where future travel times in time-varying networks change. In this paper, such procedures are proposed. The proposed techniques can provide updated solutions given simultaneous and arbitrary changes (increasing and decreasing in value) in any number of network arcs. Further, this technique can be extended for use in stochastic networks.

A Robust Solution Approach to the Dynamic Vehicle Scheduling Problem

This paper presents a solution approach to the dynamic vehicle scheduling problem. This approach consists of solving a sequence of optimization problems, where we take into account different scenarios for future travel times. We discuss the potential benefit of our approach compared to the traditional one, where the vehicle scheduling problem is solved only once for a whole period and the travel times are assumed to be fixed. Because in the multiple-depot case we cannot solve the problem exactly within reasonable computation time, we use a “cluster-reschedule” heuristic where we first assign trips to depots by solving the static problem and then solve dynamic single-depot problems. We use new mathematical formulations of these problems that allow fast solution by standard optimization software. Results of a computational study with real-life data are presented, in which we compare different variants of our approach and perform a sensitivity analysis with respect to deviations of the actual travel times from estimated ones.

A Simple and Effective Method for Predicting Travel Times on Freeways

We present a method to predict the time that will be needed to traverse a given section of a freeway when the departure is at a given time in the future. The prediction is done on the basis of the current traffic situation in combination with historical data. We argue that, for our purposes, the current traffic situation of a section of a freeway is well summarized by the current status travel time. This is the travel time that would result if one were to depart immediately and no significant changes in traffic would occur. This current status travel time can be estimated from single- or double-loop detectors, video data, probe vehicles, or any other means. Our prediction method arises from the empirical observation that there exists a linear relationship between any future travel time and the current status travel time. The slope and intercept of this relationship may change subject to the time of day and the time until departure, but linearity persists. This observation leads to a prediction scheme by means of linear regression with time-varying coefficients.