Applying multiple kernel learning and support vector machine for solving the multicriteria and nonlinearity problems of traffic flow prediction

This article proposes to develop a prediction model for traffic flow using kernel learning methods such as support vector machine (SVM) and multiple kernel learning (MKL). Traffic flow prediction is a dynamic problem owing to its complex nature of multicriteria and nonlinearity. Influential factors of traffic flow were firstly investigated; five‐point scale and entropy methods were employed to transfer the qualitative factors into quantitative ones and rank these factors, respectively. Then, SVM and MKL‐based prediction models were developed, with the influential factors and the traffic flow as the input and output variables. The prediction capability of MKL was compared with SVM through a case study. It is proved that both the SVM and MKL perform well in prediction with regard to the accuracy rate and efficiency, and MKL is more preferable with a higher accuracy rate when under proper parameters setting. Therefore, MKL can enhance the decision‐making of traffic flow prediction. Copyright © 2012 John Wiley & Sons, Ltd.     

A new methodology for incident detection and characterization on surface streets

In this paper, a new methodology is presented for real-time detection and characterization of incidents on surface streets. The proposed automatic incident detection approach is capable of detecting incidents promptly as well as characterizing incidents in terms of time-varying lane-changing fractions and queue lengths in blocked lanes, lanes blocked due to incidents, and incident duration. The architecture of the proposed incident detection approach consists of three sequential procedures: (1) Symptom Identification for identification of incident symptoms, (2) Signal Processing for real-time prediction of incident-related lane traffic characteristics and (3) Pattern Recognition for incident recognition. Lane traffic counts and occupancy are the only two major types of input data, which can be readily collected from point detectors. The primary techniques utilized in this paper include: (1) a discrete-time, nonlinear, stochastic system with boundary constraints to predict real-time lane-changing fractions and queue lengths and (2) a pattern-recognition-based algorithm employing modified sequential probability ratio tests (MSPRT) to detect incidents. Off-line tests based on simulated as well as video-based real data were conducted to assess the performance of the proposed algorithm. The test results have indicated the feasibility of achieving real-time incident detection using the proposed methodology.     

An ensemble approach for conflict detection in Free Flight by data mining

Airborne separation assurance is a key requirement for Free Flight operations. A variety of conflict detection (CD) and resolution algorithms have been developed for this task. A lack of rigorous evaluation and the existence of an infinite number of possible conflict geometries in Free Flight makes the choice of which algorithm to be placed in the cockpit a challenging task for the designers of future air traffic management (ATM) systems. In this paper, we propose an ensemble approach for CD in Free Flight. The ensemble consists of several CD algorithms, a rule set for each algorithm describing its learned behavior from its past performance and a switch mechanism to choose an appropriate CD algorithm given probe characteristics. A novel mechanism to evolve complex conflict scenarios, using genetic algorithms (GA), is developed and integrated in a fast time air traffic simulator to generate the performance data of CD algorithms. Data mining techniques are then employed to identify implicit patterns in the probe characteristics where the CD algorithms missed or falsely identified a conflict. These patterns are formulated as rule sets for each CD algorithm and are then used by a switch in the ensemble to route a probe for conflict prediction. Given probe characteristics, the CD algorithm, which is less likely to miss or falsely identify a conflict, is selected to evaluate the probe for potential conflict. The performance of the ensemble and of individual algorithms is evaluated by comparing the Pareto efficient set of solutions generated by them. The ensemble approach demonstrates a significant reduction in the number of missed detects and false alarms as compared to individual algorithms. The proposed methodology is capable of accommodating existing as well as new CD models and can be extended to other ATM concepts as well.     

Automated vision inspection of rail surface cracks: A double-layer data-driven framework

A double-layer data-driven framework for the automated vision inspection of the rail surface cracks is proposed in this paper. Based on images of rails, the proposed framework is capable to detect the location of cracks firstly and next automatically obtain the boundary of cracks via a feature-based linear iterative crack aggregation (FLICA). Extended Haar-like features are applied to develop significant features for identifying cracks in images. Built on extended Haar-like features, a cascading classifier ensemble integrating three single cascading classifiers with a major voting scheme is proposed to decide the presence of cracks in the image. Each single cascading classifier is composed of a sequence of stage classifiers trained by the LogitBoost algorithm. A scalable sliding window carrying the cascading classifier ensemble is applied to scan images of rail tracks, which is identified by the Otsu’s method, and detect cracks. After completing the crack registration in the first layer, the FLICA is developed to discover boundaries of cracks. The effectiveness of the proposed data-driven framework for identifying rail surface cracks is validated with the rail images provided by the China Railway Corporation and Hong Kong Mass Transit Railway (MRT). Six benchmarking methods, the Otsu’s method, mean shift, the visual detection system, the geometrical approach, fully convolutional networks and the U-net, are utilized to prove advantages of the proposed framework. Results of the validation and comparative analyses demonstrate that the proposed framework is most effective in the rail surface crack detection.     

Fuzzy-logic-based incident detection for signalized diamond interchanges

This paper documents a fuzzy-logic-based incident detection algorithm for signalized urban diamond interchanges. The model is capable of detecting lane-blocking incidents whose effects are manifested by patterns of deterioration in traffic conditions that require adjustments in signal control strategies. As a component of a real-time traffic adaptive control system for signalized diamond interchanges, the algorithm feeds an incident report (i.e., the time, location, and severity of the incident) to the system's optimization manager, which uses that information to determine the appropriate signal control strategy.The performance of the model was studied using a simulation of an actual diamond interchange. The simulation study evaluated the model's performance in terms of detection rate, false alarm rate, and mean time to detect. The model's performance was encouraging, and the fuzzy-logic-based approach is considered promising.     

Discrete dynamic models for freeway incident detection systems

This paper considers the problem of freeway incident detection within the general framework of computer‐based freeway surveillance and control. A new approach to the detection of freeway traffic incidents is presented based on a discrete‐time stochastic model of the form ARIMA (0, 1, 3) that describes the dynamics of traffic occupancy observations. This approach utilizes real‐time estimates of the variability in traffic occupancies as detection thresholds, thus eliminating the need for threshold calibration and lessening the problem of false‐alarms. Because the moving average parameters of the ARIMA (0, 1, 3) model change over time, these parameters can be updated occasionally. The performance of the developed detection algorithm has been evaluated in terms of detection rate, false‐alarm rate, and average time‐lag to detection, using a total of 1692 minutes of occupancy observations recorded during 50 representative traffic incidents.     

Characterization of incidents on an urban arterial road

The early warning of incidents on urban arterial roads in a congested city can reduce delay, accidents and pollutant emission. Freeway incident detection systems implemented in recent years may not be suitable for arterial incidents. Arterial incident detection is more difficult. The traffic flow on an arterial road is not conserved from the upstream end of a road link to the downstream end because urban traffic does turn in and out of side‐streets, car‐parks and local residences. Roadside friction such as kerbside parking and shopping traffic also tends to create apparent incidents which are in fact frequent and normal events. This paper develops a definition for an arterial incident and describes a case study on an arterial road in Melbourne, Australia. The study shows that detectors upstream of an incident are more useful for incident detection than downstream detectors. It also identifies occupancy and speed as the appropriate parameters to characterise and detect arterial incidents.     

A computational model for the probit‐based dynamic stochastic user optimal traffic assignment problem

This paper focuses on computational model development for the probit‐based dynamic stochastic user optimal (P‐DSUO) traffic assignment problem. We first examine a general fixed‐point formulation for the P‐DSUO traffic assignment problem, and subsequently propose a computational model that can find an approximated solution of the interest problem. The computational model includes four components: a strategy to determine a set of the prevailing routes between each origin–destination pair, a method to estimate the covariance of perceived travel time for any two prevailing routes, a cell transmission model‐based traffic performance model to calculate the actual route travel time used by the probit‐based dynamic stochastic network loading procedure, and an iterative solution algorithm solving the customized fixed‐point model. The Ishikawa algorithm is proposed to solve the computational model. A comparison study is carried out to investigate the efficiency and accuracy of the proposed algorithm with the method of successive averages. Two numerical examples are used to assess the computational model and the algorithm proposed. Results show that Ishikawa algorithm has better accuracy for smaller network despite requiring longer computational time. Nevertheless, it could not converge for larger network. Copyright © 2010 John Wiley & Sons, Ltd.     

Competing risk mixture model and text analysis for sequential incident duration prediction

Predicting the duration of traffic incidents sequentially during the incident clearance period is helpful in deploying efficient measures and minimizing traffic congestion related to such incidents. This study proposes a competing risk mixture hazard-based model to analyze the effect of various factors on traffic incident duration and predict the duration sequentially. First, topic modeling, a text analysis technique, is used to process the textual features of the traffic incident to extract time-dependent topics. Given four specific clearance methods and the uncertainty of these methods when used during traffic incidents, the proposed mixture model uses the multinomial logistic model and parametric hazard-based model to assess the influence of covariates on the probability of clearance methods and on the duration of the incident. Subsequently, the performance of estimated mixture model in sequentially predicting the incident duration is compared with that of the non-mixture model. The prediction results show that the presented mixture model outperforms the non-mixture model.     

Improved waveform‐feature‐based vehicle classification using a single‐point magnetic sensor

Vehicle classification systems have important roles in applications related to real‐time traffic management. They also provide essential data and necessary information for traffic planning, pavement design, and maintenance. Among various classification techniques, the length‐based classification technique is widely used at present. However, the undesirable speed estimates provided by conventional data aggregation make it impossible to collect reliable length data from a single‐point sensor during real‐time operations. In this paper, an innovative approach of vehicle classification will be proposed, which achieved very satisfactory results on a single‐point sensor. This method has two essential parts. The first concerns with the procedure of smart feature extraction and selection according to the proposed filter–filter–wrapper model. The model of filter–filter–wrapper is adopted to make an evaluation on the extracted feature subsets. Meanwhile, the model will determine a nonredundant feature subset, which can make a complete reflection on the differences of various types of vehicles. In the second part, an algorithm for vehicle classification according to the theoretical basis of clustering support vector machines (C‐SVMs) was established with the selected optimal feature subset. The paper also uses particle swarm optimization (PSO), with the purpose of searching for an optimal kernel parameter and the slack penalty parameter in C‐SVMs. A total of 460 samples were tested through cross validation, and the result turned out that the classification accuracy was over 99%. In summary, the test results demonstrated that our vehicle classification method could enhance the efficiency of machine‐learning‐based data mining and the accuracy of vehicle classification. Copyright © 2014 John Wiley & Sons, Ltd.     

Reprint of: Modelling the impact of traffic incidents on travel time reliability

Traffic incidents are recognised as one of the key sources of non-recurrent congestion that often leads to reduction in travel time reliability (TTR), a key metric of roadway performance. A method is proposed here to quantify the impacts of traffic incidents on TTR on freeways. The method uses historical data to establish recurrent speed profiles and identifies non-recurrent congestion based on their negative impacts on speeds. The locations and times of incidents are used to identify incidents among non-recurrent congestion events. Buffer time is employed to measure TTR. Extra buffer time is defined as the extra delay caused by traffic incidents. This reliability measure indicates how much extra travel time is required by travellers to arrive at their destination on time with 95% certainty in the case of an incident, over and above the travel time that would have been required under recurrent conditions. An extra buffer time index (EBTI) is defined as the ratio of extra buffer time to recurrent travel time, with zero being the best case (no delay). A Tobit model is used to identify and quantify factors that affect EBTI using a selected freeway segment in the Southeast Queensland, Australia network. Both fixed and random parameter Tobit specifications are tested. The estimation results reveal that models with random parameters offer a superior statistical fit for all types of incidents, suggesting the presence of unobserved heterogeneity across segments. What factors influence EBTI depends on the type of incident. In addition, changes in TTR as a result of traffic incidents are related to the characteristics of the incidents (multiple vehicles involved, incident duration, major incidents, etc.) and traffic characteristics.     

Urban traffic flow prediction: a spatio‐temporal variable selection‐based approach

Short‐term traffic flow prediction in urban area remains a difficult yet important problem in intelligent transportation systems. Current spatio‐temporal‐based urban traffic flow prediction techniques trend aims to discover the relationship between adjacent upstream and downstream road segments using specific models, while in this paper, we advocate to exploit the spatial and temporal information from all available road segments in a partial road network. However, the available traffic states can be high dimensional for high‐density road networks. Therefore, we propose a spatio‐temporal variable selection‐based support vector regression (VS‐SVR) model fed with the high‐dimensional traffic data collected from all available road segments. Our prediction model can be presented as a two‐stage framework. In the first stage, we employ the multivariate adaptive regression splines model to select a set of predictors most related to the target one from the high‐dimensional spatio‐temporal variables, and different weights are assigned to the selected predictors. In the second stage, the kernel learning method, support vector regression, is trained on the weighted variables. The experimental results on the real‐world traffic volume collected from a sub‐area of Shanghai, China, demonstrate that the proposed spatio‐temporal VS‐SVR model outperforms the state‐of‐the‐art. Copyright © 2015 John Wiley & Sons, Ltd.     

Fine-tuning ADAS algorithm parameters for optimizing traffic safety and mobility in connected vehicle environment

Under the Connected Vehicle environment where vehicles and road-side infrastructure can communicate wirelessly, the Advanced Driver Assistance Systems (ADAS) can be adopted as an actuator for achieving traffic safety and mobility optimization at highway facilities. In this regard, the traffic management centers need to identify the optimal ADAS algorithm parameter set that leads to the optimization of the traffic safety and mobility performance, and broadcast the optimal parameter set wirelessly to individual ADAS-equipped vehicles. Once the ADAS-equipped drivers implement the optimal parameter set, they become active agents that work cooperatively to prevent traffic conflicts, and suppress the development of traffic oscillations into heavy traffic jams. Measuring systematic effectiveness of this traffic management requires am analytic capability to capture the quantified impact of the ADAS on individual drivers’ behaviors and the aggregated traffic safety and mobility improvement due to such an impact. To this end, this research proposes a synthetic methodology that incorporates the ADAS-affected driving behavior modeling and state-of-the-art microscopic traffic flow modeling into a virtually simulated environment. Building on such an environment, the optimal ADAS algorithm parameter set is identified through a multi-objective optimization approach that uses the Genetic Algorithm. The developed methodology is tested at a freeway facility under low, medium and high ADAS market penetration rate scenarios. The case study reveals that fine-tuning the ADAS algorithm parameter can significantly improve the throughput and reduce the traffic delay and conflicts at the study site in the medium and high penetration scenarios. In these scenarios, the ADAS algorithm parameter optimization is necessary. Otherwise the ADAS will intensify the behavior heterogeneity among drivers, resulting in little traffic safety improvement and negative mobility impact. In the high penetration rate scenario, the identified optimal ADAS algorithm parameter set can be used to support different control objectives (e.g., safety improvement has priority vs. mobility improvement has priority).     

Hybrid model for prediction of bus arrival times at next station

Effective prediction of bus arrival times is important to advanced traveler information systems (ATIS). Here a hybrid model, based on support vector machine (SVM) and Kalman filtering technique, is presented to predict bus arrival times. In the model, the SVM model predicts the baseline travel times on the basic of historical trips occurring data at given time‐of‐day, weather conditions, route segment, the travel times on the current segment, and the latest travel times on the predicted segment; the Kalman filtering‐based dynamic algorithm uses the latest bus arrival information, together with estimated baseline travel times, to predict arrival times at the next point. The predicted bus arrival times are examined by data of bus no. 7 in a satellite town of Dalian in China. Results show that the hybrid model proposed in this paper is feasible and applicable in bus arrival time forecasting area, and generally provides better performance than artificial neural network (ANN)–based methods. Copyright © 2010 John Wiley & Sons, Ltd.     

Development of an integrated system based vehicle tracking algorithm with shadow removal and occlusion handling methods

Video image processing system (VIPS) is more efficient than other detecting systems. However, VIPS involves outdoor images and is very sensitive to the external environment, which could greatly decrease its accuracy according to rapid environmental changes. To obtain accurate traffic data accordingly, VIPS must address the problems such as growing shadows in transition; distortion of images due to the headlights at night; noises caused by the rain, snow or fog; and occlusions. This study intends to accurately calculate traffic data while addressing the shadow and occlusion problems, which are the most difficult tasks for the image‐detector‐based traffic data system. In this study, an algorithm for the individual vehicle tracking collection was developed to address the occlusion problem and to eliminate the noises or shadows caused by external environmental factors. A traffic data collection system was also proposed in order to accurately track individual vehicles that pass through the detection region. In addition, establishing an integrated system with shadow removal and occlusion handling using an image processing was also proposed. Copyright © 2010 John Wiley & Sons, Ltd.     

A fuzzy logic‐based variable speed limit controller

Variable speed limit (VSL) is an emerging intelligent transportation system (ITS) measure to improve operational and safety performance of motorway systems. Rule‐based algorithms have been widely used in VSL applications because of their comprehensibility and ease of application. However, most of the algorithms proposed in the literature under this category are rather rough for the speed control. Pre‐specified rules show some difficulties in appropriately activating/deactivating control actions in real time because of non‐stationary and nonlinear nature of the traffic system. This paper proposes a fuzzy logic‐based VSL control algorithm as an alternative to the existing VSL control algorithms. The proposed algorithm uses fuzzy sets instead of crisp sets to allow the separation of attribute domains into several overlapping intervals. The discretization using fuzzy sets can help to overcome the sensitivity problem caused by crisp discretization used in the existing VSL algorithms. The proposed algorithm is assessed for a test bed in Auckland using AIMSUN micro‐simulator and verified against a well‐known VSL algorithm. The simulation results show that the proposed algorithm outperforms the existing one to improve the efficiency performance of the motorway system with the critical bottleneck capacity increased by 6.42% and total travel time reduced by 12.39% when compared to a no‐control scenario. Copyright © 2015 John Wiley & Sons, Ltd.     

Decomposing travel times measured by probe-based traffic monitoring systems to individual road segments

In probe-based traffic monitoring systems, traffic conditions can be inferred based on the position data of a set of periodically polled probe vehicles. In such systems, the two consecutive polled positions do not necessarily correspond to the end points of individual links. Obtaining estimates of travel time at the individual link level requires the total traversal time (which is equal to the polling interval duration) be decomposed. This paper presents an algorithm for solving the problem of decomposing the traversal time to times taken to traverse individual road segments on the route. The proposed algorithm assumes minimal information about the network, namely network topography (i.e. links and nodes) and the free flow speed of each link. Unlike existing deterministic methods, the proposed solution algorithm defines a likelihood function that is maximized to solve for the most likely travel time for each road segment on the traversed route. The proposed scheme is evaluated using simulated data and compared to a benchmark deterministic method. The evaluation results suggest that the proposed method outperforms the bench mark method and on average improves the accuracy of the estimated link travel times by up to 90%.     

An application of the CUSUM algorithm to freeway incident detection based on two contiguous detectors

The CUSUM (cumulative sum of log‐likelihood ratio) algorithm is a detection algorithm that shows potential for the improvement of incident detection algorithms because it is designed to minimize the mean detect delay for a given false alarm constraint and it can also detect changes with different patterns. In this study, the CUSUM algorithm was applied to freeway incident detection by integrating traffic measurements from two contiguous loop detectors and the non‐stationarity of traffic flows. The developed algorithm was tested based on incident data from the PATH program, with consideration given to the impact of different geometric conditions on algorithm performance. It was also compared with two existing algorithms, in order to address the influence of traffic patterns. The evaluation results show that the CUSUM incident detection algorithm can perform equally well in comparison with the selected algorithms.     

An improved headway-based holding strategy for bus transit

Abstract

This paper presents an improved headway-based holding strategy integrating bus transit travel and dwelling time prediction. A support vector machine-based (SVM) model is developed to predict the baseline travel and dwell times of buses based on recent data. In order to reduce prediction errors, an adaptive algorithm is used together with real-time bus operational information and estimated baseline times from SVM models. The objective of the improved holding strategy is to minimize the total waiting times of passengers at the current stop and at successive stops. Considering the time-varying features of bus running, a ‘forgetting factor’ is introduced to weight the most recent data and reduce the disturbance from unexpected incidents. Finally, the improved holding strategy proposed in this study is illustrated using the microscopic simulation model Paramics and some conclusions are drawn.