Search

Macroscopic traffic flow modeling is essential for describing and forecasting the characteristics of traffic flow. However, the classic Lighthill–Whitham–Richards (LWR) model only provides equilibrium values for steady‐state conditions and fails to capture common stochastic variabilities, which are a necessary component of accurate modeling of real‐time traffic management and control. In this paper, a stochastic LWR (SLWR) model that randomizes free‐flow speed is developed to account for the stochasticity incurred by the heterogeneity of drivers, while holding individual drivers' behavior constant. The SLWR model follows a conservation law of stochastic traffic density and flow and is formulated as a time‐dependent stochastic partial differential equation. The model is solved using a dynamically bi‐orthogonal (DyBO) method based on a spatial basis and stochastic basis. Various scenarios are simulated and compared with the Monte Carlo (MC) method, and the results show that the SLWR model can effectively describe dynamic traffic evolutions and reproduce some commonly observed traffic phenomena. Furthermore, the DyBO method shows significant computational advantages over the MC method. [ABSTRACT FROM AUTHOR]

Reinforcement learning (RL) algorithms have been widely applied in solving traffic signal control problems. Traffic environments, however, are intrinsically nonstationary, which creates a convergence problem that RL algorithms struggle to overcome. Basically, as a target problem for an RL algorithm, the Markov decision process (MDP) can be solved only when both the transition and reward functions do not vary. Unfortunately, the environment for traffic signal control is not stationary since the goal of traffic signal control varies according to congestion levels. For unsaturated traffic conditions, the objective of traffic signal control should be to minimize vehicle delay. On the other hand, the objective must be to maximize the throughput when traffic flow is saturated. A multiregime analysis is possible for varying conditions, but classifying the traffic regime creates another complex task. The present study provides a meta‐RL algorithm that embeds a latent vector to recognize the different contexts of an environment in order to automatically classify traffic regimes and apply a customized reward for each context. In simulation experiments, the proposed meta‐RL algorithm succeeded in differentiating rewards according to the saturation level of traffic conditions. [ABSTRACT FROM AUTHOR]

Observability and controllability are two critical requirements for a partially observable transportation system. This paper proposes a stepwise signal optimization framework with connected vehicle (CV) data as input to solve both challenges. First, a Bayesian deduction method based on low‐penetration CV data is established to estimate the traffic volume. Thereafter, an offline signal optimization model is constructed to simultaneously optimize the flexible lane settings and signal timings, which are set as the prior information for the third step. In the third step, an online deep recurrent Q‐learning (DRQN) signal optimization model dynamically adjusts signal settings based on real‐time traffic information. Numerical experiments demonstrate that the model outperforms the actuated control, the online DQRN model without offline filter, and the back‐pressure model by 9%–66% and 7%–29% in two networks. This study innovatively combines traffic state estimation and traffic signal control as an integrated process. It contributes to an improved understanding of traffic control in a CV environment and lays a solid foundation for future traffic control strategies. [ABSTRACT FROM AUTHOR]

This paper presents a methodology to control the trajectory of cooperative connected automated vehicles (CAVs) at roundabouts with a mixed fleet of CAVs and human‐driven vehicles (HVs). We formulate an optimization program in a two‐dimensional space for this purpose. A model predictive control‐based solution technique is developed to optimize the trajectories of CAVs at discretized time steps based on the estimated driving behavior of HVs, while the actual behavior of HVs is controlled by a microscopic traffic simulator. At each time step, the location and speed of vehicles are collected, and a decomposition‐based methodology optimizes CAV trajectories for a few time steps ahead of the system time. The optimization methodology has convexification, alternating direction method of multipliers, and cutting plane decomposition components to tackle the complexities of the problem. We tested the solution technique in a case study roundabout with different traffic demand flow rates and CAV market penetration rates. The results showed that increasing the CAV market penetration rate from 20% to 100% reduced total travel times by 2.8% to 35.8%. The analyses indicate that the presence of cooperative CAVs in roundabouts can lead to considerable improvements. [ABSTRACT FROM AUTHOR]

Abstract: Connected vehicle technology, the Internet of Things, and other advanced communication technologies create possibilities to facilitate the movement of vehicles through transportation networks and reduce their travel time. Harmonizing the speed of vehicles in different network links not only yields a more efficient network capacity utilization, but also regulates the movement of vehicles to achieve a “smoother” flow of traffic. This study develops a mathematical nonlinear formulation for dynamic speed harmonization in urban street networks aiming at improving traffic operations. We have converted the nonlinear problem into a linear program utilizing the fundamental flow–density relationship and developed a model predictive control approach to account for stochastic changes in traffic demand and further improve the efficiency of the developed solution algorithm. Results showed that the algorithm efficiently found dynamic optimal advisory speeds on various network links, and speed harmonization significantly reduced the travel time (up to 5.4%), speed variance (19.8%–29.4%), and the number of stops (8.3%–18.5%), while increasing the average speed (up to 5.9%) and the number of completed trips (up to 4%) in our case study network under all tested demand patterns. [ABSTRACT FROM AUTHOR]

Connected vehicles (CVs), be they autonomous vehicles or a fleet of cargo carriers or Uber, are a matter of when they become a reality and not if. It is not unreasonable to think that CV technology may have a far-reaching impact, even to the genesis of a completely new traffic pattern. To this end, the literature has yet to address the routing behavior of the CVs, namely traffic assignment problem (TAP) (perhaps it is assumed, they ought to follow the traditional shortest possible paths, known as user equilibrium [UE]). It is possible that real-time data could be derived from the vehicles' communications that in turn could be used to achieve a better traffic circulation. In this article, we propose a mathematical formulation to ensure the CVs are seeking the system optimal (SO) principles, while the remainder continue to pursue the old-fashioned UE pattern. The model is formulated as a nonlinear complementarity problem (NCP). This article contributes to the literature in three distinct ways: (i) mathematical formulation for the CVs' routing, stated as a mixed UE-SO traffic pattern, is proposed; (ii) a variety of realistic features are explicitly considered in the solution to the TAP including road capacity, elastic demand, multiclass and asymmetric travel time; and (iii) formal proof of the existence and uniqueness of the solutions are also presented. The proposed methodology is applied to the networks of Sioux-Falls and Melbourne. [ABSTRACT FROM AUTHOR]

This article proposes a prototype of an urban traffic control system based on a prediction-after-classification approach. In an off-line phase, a repository of traffic control strategies for a set of (dynamic) traffic patterns is constructed. The core of this stage is the k-means algorithm for daily traffic pattern identification. The clustering method uses the input attributes flow, speed, and occupancy and it transforms the dynamic traffic data at network level in a pseudo-covariance matrix, which collects the dynamic correlations between the road links. A desirable number of traffic patterns is provided by Bayesian Information Criterion and the ratio of change in dispersion measurements. In an on-line phase, the current daily traffic pattern is predicted within the repository and its associated control strategy is implemented in the traffic network. The dynamic prediction scheme is constructed on the basis of an existing static prediction method by accumulating the trials on set of patterns in the repository. This proposal has been assessed in synthetic and real networks testing its effectiveness as a data mining tool for the analysis of traffic patterns. The approach promises to effectively detect the current daily traffic pattern and is open to being used in intelligent traffic management systems. [ABSTRACT FROM AUTHOR]

The fuel consumption of ground vehicles is significantly affected by how they are driven. The fuel-optimized vehicular automation technique can improve fuel economy for the host vehicle, but their effectiveness on a platoon of vehicles is still unknown. This article studies the performance of a well-known fuel-optimized vehicle automation strategy, i.e., Pulse-and-Glide (PnG) operation, on traffic smoothness and fuel economy in a mixed traffic flow. The mixed traffic flow is assumed to be a single-lane highway on flat road consisting of both driverless and manually driven vehicles. The driverless vehicles are equipped with fuel economy-oriented automated controller using the PnG strategy. The manually driven vehicles are simulated using the Intelligent Driver Models (IDM) to mimic the average car-following behavior of human drivers in naturalistic traffics. A series of simulations are conducted with three scenarios, i.e., a single car, a car section, and a car platoon. The simulation results show that the PnG strategy can significantly improve the fuel economy of individual vehicles. For traffic flows, the fuel economy and traffic smoothness vary significantly under the PnG strategy. [ABSTRACT FROM AUTHOR]

The objective of this article is to propose an analytical traffic model adapted to freeway weaving sections. To this aim, a theoretical weaving section is considered at the macroscopic level as the superposition of two merges and two diverges. The model endogenously incorporates the capacity drop related to the weaving lane changes. It only depends on six parameters: three parameters of the fundamental diagram (the free-flow speed, the jam density, and the wave speed in congestion), the vehicles' acceleration, the relaxation factor, and the length of the anticipation zone. A sensitivity analysis proves that the acceleration rate and the relaxation parameter most influence the capacity drop. The analytical model agrees well with microsimulation results and empirical data collected on a weaving section in France. [ABSTRACT FROM AUTHOR]

Real time traffic flow simulation models are used to provide traffic information for dynamic traffic management systems. Those simulation models are supplied by traffic data in order to estimate and predict traffic conditions in unobserved sections of a traffic network. In general, most of recent real time traffic simulators are based on the macroscopic model because the macroscopic model replicates the average traffic behavior in terms of observable variables such as (time-space) flow and speed at a relatively fast computational time. Like other simulation models, an important aspect of the real time macroscopic simulator is to calibrate the model parameters online. The most conventional way of the online calibration is to add a random walk to the parameters to constitute an augmentation of the traffic variables and the model parameters to be estimated. Actually, this method allows the parameters to vary at every time step and, therefore, describes the adaptation of the model to the prevailing traffic conditions. However, it has been reported that the use of the random walk results in a loss of information and an increase of the covariance of parameters, which consequently leads to posteriors that are far more diffuse than the theoretical posteriors for the true parameters. To this end, this article puts forward a Kernel density estimation technique in the calibration process to handle the covariance issue and to avoid the information loss. The Kernel density estimation technique is embedded in the particle filter algorithm, which is extended to the calibration problems. The proposed framework is investigated using real-life data collected in a freeway in England. [ABSTRACT FROM AUTHOR]

We present an advanced interpolation method for estimating smooth spatiotemporal profiles for local highway traffic variables such as flow, speed and density. The method is based on the 'adaptive smoothing method' which takes as input stationary detector data as typically collected by traffic control centers. We generalize this method to allow for fusion with floating car data or other traffic information. The resulting profiles display transitions between free and congested traffic in great detail, as well as fine structures such as stop-and-go waves. We establish the accuracy and robustness of the method and demonstrate three potential applications: (1) compensation for gaps in data caused by detector failure; (2) separation of noise from dynamic traffic information; and (3) the fusion of floating car data with stationary detector data. [ABSTRACT FROM AUTHOR]

Fusing freeway traffic data such as spot speeds and travel times from a variety of traffic sensors (loops, cameras, automated vehicle identification systems) into a coherent, consistent, and reliable picture of the prevailing traffic conditions (e.g., speeds, flows) is a critical task in any off- or online traffic management or data archival system. This task is challenging as such data differ in terms of spatial and temporal resolution, accuracy, reliability, and most importantly in terms of spatiotemporal semantics. In this article, we propose a data fusion algorithm (the extended generalized Treiber-Helbing filter [the EGTF]) which, although heuristic in nature, uses basic notions from traffic flow theory and is generic in the sense that it does not impose any restrictions on the way the data are structured in a temporal or spatial way. This implies that the data can stem from any data source, given they provide a means to distinguish between free flowing and congested traffic. On the basis of (ground truth and sensor) data from a micro-simulation tool, we demonstrate that the EGTF method results in accurate reconstructed traffic conditions and is robust to increasing degrees of data corruption. Further research should focus on validating the approach on real data. The method can be straightforwardly implemented in any traffic data archiving system or application which requires consistent and coherent traffic data from traffic sensors as inputs. [ABSTRACT FROM AUTHOR]

The existing well-known short-term traffic forecasting algorithms require large traffic flow data sets, including information on current traffic scenarios to predict the future traffic conditions. This article proposes a random process traffic volume model that enables estimation and prediction of traffic volume at sites where such large and continuous data sets of traffic condition related information are unavailable. The proposed model is based on a combination of wavelet analysis (WA) and Bayesian hierarchical methodology (BHM). The average daily 'trend' of urban traffic flow observations can be reliably modeled using discrete WA. The remaining fluctuating parts of the traffic volume observations are modeled using BHM. This BHM modeling considers that the variance of the urban traffic flow observations from an intersection vary with the time-of-the-day. A case study has been performed at two busy junctions at the city-centre of Dublin to validate the effectiveness of the strategy. [ABSTRACT FROM AUTHOR]

A safety and security issue of concern for traffic managers and homeland security is the monitoring of flows on “planar surfaces,” such as ports and border areas. This article addresses the problem of locating surveillance radars to cover a given target surface that may have barriers through which radar signals cannot penetrate. The area of coverage of a radar is assumed to be a disc, or a partial disc when there are barriers, with a known radius. The article shows that the corresponding location problems relate to two well studied problems: the set-covering model and the maximal covering problem. In the first problem, the minimum number of radars is to be located to completely cover the target area; in the second problem a given number M of radars are to be located to cover the target area as much as possible. Based on a discrete representation of the target area, a Lagrangian heuristic and a two-stage procedure with a conquer-and-divide scaling are developed to solve the above two models. The computational experiences reported demonstrate that the developed method solves well the radar location problems formulated here. [ABSTRACT FROM AUTHOR]

Origin-destination (OD) matrices are essential for various analyses in the field of traffic planning, and they are often estimated from link flow observations. We compare methods for allocating link flow detectors to a traffic network with respect to the quality of the estimated OD-matrix. First, an overview of allocation methods proposed in the literature is presented. Second, we construct a controlled experimental environment where any allocation method can be evaluated, and compared to others, in terms of the quality of the estimated OD-matrix. Third, this environment is used to evaluate and compare three fundamental allocation methods. Studies are made on the Sioux Falls network and on a network modeling the city of Linköping. Our conclusion is, that the most commonly studied approach for detector allocation, maximizing the coverage of OD-pairs, seems to be unfavorable for the quality of the estimated OD-matrix. [ABSTRACT FROM AUTHOR]

This article presents an evaluation of the system performance of a proposed self-organizing, distributed traffic information system based on vehicle-to-vehicle information-sharing architecture. Using microsimulation, several information applications derived from this system are analyzed relative to the effectiveness and efficiency of the system to estimate traffic conditions along each individual path in the network, to identify possible incidents in the traffic network, and to provide rerouting strategies for vehicles to escape congested spots in the network. A subset of vehicles in the traffic network is equipped with specific intervehicle communication devices capable of autonomous traffic surveillance, peer-to-peer information sharing, and self-data processing. A self-organizing traffic information overlay on the existing vehicular roadway network assists their independent evaluation of route information, detection of traffic incidents, and dynamic rerouting in the network based both on historical information stored in an in-vehicle database and on real-time information disseminated through intervehicle communications. A path-based microsimulation model is developed for these information applications and the proposed distributed traffic information system is tested in a large-scale real-world network. Based on simulation study results, potential benefits both for travelers with such equipment as well as for the traffic system as a whole are demonstrated. [ABSTRACT FROM AUTHOR]

Recognizing temporal patterns in traffic flow has been an important consideration in short-term traffic forecasting research. However, little work has been conducted on identifying and associating traffic pattern occurrence with prevailing traffic conditions. We propose a multilayer strategy that first identifies patterns of traffic based on their structure and evolution in time and then clusters the pattern-based evolution of traffic flow with respect to prevailing traffic flow conditions. Temporal pattern identification is based on the statistical treatment of the recurrent behavior of jointly considered volume and occupancy series; clustering is done via a two-level neural network approach. Results on urban signalized arterial 90-second traffic volume and occupancy data indicate that traffic pattern propagation exhibits variability with respect to its statistical characteristics such as deterministic structure and nonlinear evolution. Further, traffic pattern clustering uncovers four distinct classes of traffic pattern evolution, whereas transitional traffic conditions can be straightforwardly identified. [ABSTRACT FROM AUTHOR]

Systematic tabu search (TS)-based heuristic methods are put forward in this article and applied for the design of public transportation networks with variable demand. A multi-objective nonlinear mixed integer model is formulated. Solution methodologies are proposed, which consist of three main components: an initial candidate route set generation procedure (ICRSGP) that generates all feasible routes incorporating practical bus transit industry guidelines; a network analysis procedure (NAP) that decides transit demand matrix, assigns transit trips, determines service frequencies, and computes performance measures; and a Tabu search method (TSM) that combines these two parts, guides the candidate solution generation process, and selects an optimal set of routes from the huge solution space. Comprehensive tests are conducted and sensitivity analyses are performed. Characteristics analyses are undertaken and solution qualities from different algorithms are compared. Numerical results clearly indicate that the preferred TSM outperforms the genetic algorithm used as a benchmark for the optimal bus transit route network design problem without zone demand aggregation. [ABSTRACT FROM AUTHOR]

This article addresses the problem of the accuracy of short-term traffic flow forecasting in the complex case of urban signalized arterial networks. A new, artificial intelligence (AI)-based approach is suggested for improving the accuracy of traffic predictions through suitably combining the forecasts derived from a set of individual predictors. This approach employs a fuzzy rule-based system (FRBS), which is augmented with an appropriate metaheuristic (direct search) technique to automate the tuning of the system parameters within an online adaptive rolling horizon framework. The proposed hybrid FRBS is used to nonlinearly combine traffic flow forecasts resulting from an online adaptive Kalman filter (KF) and an artificial neural network (ANN) model. The empirical results obtained from the model implementation into a real-world urban signalized arterial demonstrate the ability of the proposed approach to considerably overperform the given individual traffic predictors. [ABSTRACT FROM AUTHOR]

This article deals with the problem of estimating and updating the origin-destination matrix and link flows from traffic counts and its optimal location. A combination (bi-level) of an OD-pair matrix estimation model based on Bayesian networks, and a Wardrop-minimum-variance model, which identifies origins and destinations of link flows, is used to estimate OD-pair and unobserved link flows based on some observations of links and/or OD-pair flows. The Bayesian network model is also used to select the optimal number and locations of the links counters based on maximum correlation. Finally, the proposed methods are illustrated by their application to the Nguyen–Dupuis and the Ciudad Real networks. [ABSTRACT FROM AUTHOR]

A polymorphic dynamic network loading (PDNL) model is developed and discretized to integrate a variety of macroscopic traffic flow and node models. The polymorphism, realized through a general node-link interface and proper discretization, offers several prominent advantages. First of all, PDNL allows road facilities in the same network to be represented by different traffic flow models based on the tradeoff of efficiency and realism and/or the characteristics of the targeted problem. Second, new macroscopic link/node models can be easily plugged into the framework and compared against existing ones. Third, PDNL decouples links and nodes in network loading, and thus opens the door to parallel computing. Finally, PDNL keeps track of individual vehicular quanta of arbitrary size, which makes it possible to replicate analytical loading results as closely as desired. PDNL, thus, offers an ideal platform for studying both analytical dynamic traffic assignment problems of different kinds and macroscopic traffic simulation. [ABSTRACT FROM AUTHOR]

Modeling the destination choice is a difficult task and very often it represents the weakest step in travel demand modeling. This weakness is mainly due to the high number of potential alternatives and to the very limited number of available attributes. Indeed, the alternative destinations in a distribution model are generally all the (hundreds in a medium large size city) traffic zones identified in the zoning phase. Moreover, the most widely adopted specification of random utility (RU) destination choice models introduces just two categories of attributes: attractiveness attributes of the destination zone and “impedance” attributes reproducing the origin-destination generalized cost. Through this kind of attribute alone it is quite difficult reproducing the real choice context faced by the decision maker, who generally knows only a limited part of the study area with sufficient detail to evaluate its attractiveness and the generalized transport costs of reaching it. In this regard, our article proposes two sets of new dummy-like attributes to be used within the destination choice models to identify, within the whole choice set, a smaller subset of zones (those with nonzero value of dummy-like attributes) more/less likely to be perceived. The former are generated by extending and applying the concept of dominance among alternatives to the framework of RU theory and can be used to identify a set of alternatives less likely to be perceived (exclusion variables) whose systematic utility will be penalized as a function of these variables. The latter are spatial variables reproducing better knowledge of zones with a privileged spatial position and can be used to identify a set of alternatives more likely to be perceived (selection variables) whose systematic utility will be improved as a function of these variables. These new attributes are tested on empirical data related to nonsystematic trips in Rome, Italy. It is also important to underline that the proposed dominance variables can be conveniently used in any other choice context. [ABSTRACT FROM AUTHOR]

When modeling the acceleration and deceleration of drivers, there are three characteristic time constants that influence the dynamics and stability of traffic flow: The reaction time of the drivers, the velocity adaptation time needed to accelerate to a new desired velocity, and the numerical update time. By means of numerical simulations with a time-continuous car-following model, we investigate how these times interplay with each other and effectively influence the longitudinal instability mechanisms for a platoon of vehicles. The long-wavelength string instability is mainly driven by the velocity adaptation time while short-wavelength local instabilities arise for sufficiently high reaction and update times. Furthermore, we investigate the relation between large update time steps and finite reaction times as they both introduce delays in the reaction to the traffic situation. Remarkably, the numerical update time is dynamically equivalent to about half the reaction time, which clarifies the meaning of the time step in models formulated as iterated maps such as the Newell and the Gipps model. Furthermore, with respect to stability, there is an optimal adaptation time as a function of the reaction time. [ABSTRACT FROM AUTHOR]

This article investigates the application of Kalman filter with discrete wavelet analysis in short-term traffic volume forecasting. Short-term traffic volume data are often corrupted by local noises, which may significantly affect the prediction accuracy of short-term traffic volumes. Discrete wavelet decomposition analysis is used to divide the original data into several approximate and detailed data such that the Kalman filter model can then be applied to the denoised data and the prediction accuracy can be improved. Two types of wavelet Kalman filter models based on Daubechies 4 and Haar mother wavelets are investigated. Traffic volume data collected from four different locations are used for comparison in this study. The test results show that both proposed wavelet Kalman filter models outperform the direct Kalman filter model in terms of mean absolute percentage error and root mean square error. [ABSTRACT FROM AUTHOR]

Current interest in short-term traffic volume forecasting focuses on incorporating temporal and spatial volume characteristics in the forecasting process. This article addresses the problem of integrating and optimizing predictive information from multiple locations of an urban signalized arterial and proposes a modular neural predictor consisting of temporal genetically optimized structures of multilayer perceptrons (MLP) that are fed with volume data from sequential locations to improve the accuracy of short-term forecasts. The results show that the proposed methodology provides more accurate forecasts compared to the conventional statistical methodologies applied, as well as to the static forms of neural networks. [ABSTRACT FROM AUTHOR]

Most automatic incident detection algorithms were successfully developed using loop-detector-based traffic measurements collected from their own localities. But their detection performances were not satisfactory when applied on data collected using a video-based detector system. The video-based detector system is gaining popularity as it was reported to be cost-effective, less prone to damage compared to loop detectors embedded in road pavement, and possesses surveillance capability. It is able to provide the homogeneity of traffic measurements with greater reliability in non-incident situations. In this study, a simple detection rule was used to develop algorithms that use video-based data for detecting lane-blocking incidents. A set of 96 incidents from Singapore's Central Expressway was used for calibrating these algorithms, with another 64 incidents for validation. Two single-station algorithms, named dual-variable (DV) and flow-based DV algorithms were developed. They have similar detection logic, but the latter includes a pre-incident traffic flow condition in its detection framework. On average, the flow-based DV algorithm outperformed the DV algorithm, and both proved to be effective techniques when compared to some existing loop-detector-based algorithms. [ABSTRACT FROM AUTHOR]

Accurate and timely forecasting of traffic flow is of paramount importance for effective management of traffic congestion in intelligent transportation systems. A detailed understanding of the properties of traffic flow is essential for building a reliable forecasting model. The discrete wavelet packet transform (DWPT) provides more coefficients than the conventional discrete wavelet transform (DWT), representing additional subtle details of a signal. In wavelet multiresolution analysis, an important decision is the selection of the decomposition level. In this research, the statistical autocorrelation function (ACF) is proposed for the selection of the decomposition level in wavelet multiresolution analysis of traffic flow time series. A hybrid wavelet packet-ACF method is proposed for analysis of traffic flow time series and determining its self-similar, singular, and fractal properties. A DWPT-based approach combined with a wavelet coefficients penalization scheme and soft thresholding is presented for denoising the traffic flow. The proposed methodology provides a powerful tool in removing the noise and identifying singularities in the traffic flow. The methods created in this research are of value in developing accurate traffic-forecasting models. [ABSTRACT FROM AUTHOR]

This article explores the ability of imprecise computations to address real–time computational requirements in infrastructure control and management systems. The research in this area focuses on the development of nonparametric regression as a means to forecast traffic flow rates for transportation management systems. Nonparametric regression is a forecasting technique based on nearest neighbor searching, in which forecasts are derived from past observations that are similar to current conditions. A key concern regarding nonparametric regression is the significant time required to search for nearest neighbors in large databases. The results presented in this article indicate that approximate nearest neighbors, which are imprecise computations as applied to nonparametric regression, may be used to adequately speed the execution time of nonparametric regression, with acceptable degradations in forecast accuracy. The article concludes with a demonstration of the use of genetic algorithms as a design aid for real–time algorithms employing imprecise computations. [ABSTRACT FROM AUTHOR]

Real-time traffic management is a promising approach for alleviating congestion. This approach uses realtime and predicted traffic information to develop routing strategies that optimize the performance of highway networks. This article explores the potential for using case-based reasoning (CBR), an emerging artificial intelligence (AI]) paradigm, to overcome the limitations of existing traffic-management decision support systems. To illustrate the feasibility of the approach, the article develops and evaluates a prototype CBR routing system for a real-world network in Hampton Roads, Virginia. Cases for building the system's case base are generated using a heuristic dynamic traffic assignment (DTA) model specifically designed for the region. Using a set of 25 new independent cases, the performance of the prototype system is evaluated by comparing its solutions with those of the DTA model. The evaluation results demonstrate the feasibility of the CBR approach. The prototype system was capable of running in real time and produced high-quality solutions using case bases of reasonable size. [ABSTRACT FROM AUTHOR]