Authors: Xin Zhang and Gang-Len Chang
Journal: Journal of Intelligent Transportation Systems: Technology, Planning, and Operations 2014
Abstract: In most metropolitan areas, an emergency evacuation may require a potentially large number of pedestrians to walk some distance to access their passenger cars or resort to transit systems. In this process, the massive number of pedestrians may place a tremendous burden on vehicles in the roadway network, especially at critical intersections. Thus, the effective road enforcement of the vehicle and pedestrian flows and the proper coordination between these two flows at critical intersections during a multi-modal evacuation process is a critical issue in evacuation planning. This article presents an integrated linear model for the design of optimized flow plans for massive mixed pedestrian–vehicle flows within an evacuation zone. The optimized flow can also be used to generate signal timing plans at critical intersections. In addition, the linear nature of the model can circumvent the computational burden to apply in large-scale networks. An illustrating example of the evacuation around the M&T Bank Stadium in downtown Baltimore, MD, is presented and used to demonstrate the model’s capability to address the complex interactions between vehicle and pedestrian flows within an evacuation zone. Results of simulation experiments verify the applicability of our model to a real-world scenario and further indicate that accounting for such conflicting movements will yield more reliable estimation of an evacuation’s required clearance time.
Authors: Yang (Carl) Lu, Xianfeng Yang, and Gang-Len Chang
Journal: Transportation Research Record 2014
Abstract: Although average effective vehicle length (AEVL) has been recognized as one of the most popular methods for detecting data errors, how to set proper thresholds so as to prevent false alarms and missed detections remains a challenging ongoing issue. This study proposed a sequential screening algorithm that employed multiple comparisons with the best statistics to compare concurrently the estimated AEVLs between lanes and stations for assessment of the data quality of a target detector. With both the temporal and spatial information, the proposed method can reliably generate a confidence interval and determine whether the target detector is malfunctioning or in need of calibration. The proposed algorithm was tested with 2 weeks of detector data from Ocean City, Maryland. The analysis results demonstrate the effectiveness of the proposed sequential screening algorithm and its potential for field applications.
Download (Algorithm-for-Detector-Error-Screening-on-Basis-of-Temporal-and-Spatial-Information.pdf)
Authors: Xin Zhang and Gang-Len Chang
Journal: Journal of Public Transportation, Vol. 17, No. 3, 2014
Abstract: This paper develops a decision-support model for transit-based evacuation planning occurring in metropolitan areas. The model consists of two modules executed in a sequential manner: the first deals with determining pick-up locations from candidate locations based on the spatial distribution of the evacuees, and the second plans for the route and schedule for each transit vehicle based on vehicle availability and evacuee demand pattern. An overlapping clustering algorithm is first adopted in allocating the demands to several nearby clusters. Then, an optimization model is proposed to allocate available buses from the depots to transport the assembled evacuees between the pick-up locations and different safety destinations and public shelters. A numerical example based on the city of Baltimore demonstrates the applicability of the proposed model and the advantages compared to state-of-the-art models with overly strict and unrealistic assumptions.
Download (A-Transit-Based-Evacuation-Model-for-Metropolitan-Areas.pdf)
Authors: Chien-Lun Lan and Gang-Len Chang
Journal: IEEE Transactions on Intelligent Transportation Systems 2014
Abstract: In response to the need for designing signal plans for congested intersections caused by heavy scooter–vehicle mixed flows, this paper presents our formulated model for optimizing both the cycle length and signal timings for isolated intersections. The proposed model accounts for the interactions between scooter and vehicle flows and reflects the maneuverability of scooters in the queue formation and discharging process. The robustness of the proposed formulations has been evaluated with field data and laboratory experiments. The signal optimization model, grounded on such formulations for scooter–vehicle mixed flows, has also been implemented at an intersection and assessed with a rigorous before-and-after field analysis. Our research concludes that incorporating the unique properties of scooter flows is essential for design and development of effective signal control strategies to contend with recurrent congestion caused by heavy mixed scooter–vehicle flows.
Authors: Zichuan Li, Gang-Len Chang, Suhasini Natarajan
Conference: the 11th International IEEE Conference on Intelligent Transportation System, October 2008 in Beijing, China
Status: Presented
Abstract:
This study presents a mixed integer model for an integrated control between off-ramp and arterial traffic flows. The proposed study intends to minimize the queue spillback from off-ramp to the freeway mainline that may significantly degrade the performance quality of the entire freeway system. In this study, the Cell Transmission Model (1, 2) is employed to capture the traffic propagation on both freeway an surface streets, and to capture the interactions between those two types of flows within the target control boundaries. An efficient solution method based on Genetic Algorithm is provided along with a numeric case study to demonstrate the benefit of this proposed model.
Authors: Xin Zhang and Gang-Len Chang
Conference: TRB 2010 Annual Meeting, January 12, 2010
Status: Presented
Abstract:
This paper presents an integrated model for design of signal plans for massive mixed pedestrian-vehicle flows within the evacuation zone. The proposed model with its embedded formulations for pedestrians and vehicles in the same network can effectively take into their potential conflicts during the evacuation, and generate the optimal routing strategies for guiding evacuees moving toward either the pick-up locations or their parking areas. The core formulations is based on the cell-transmission concept, but the proposed model has been enhanced with the notion of sub cells proposed mainly to capture the complex movements in the pedestrian flows so that it can concurrently optimize both the signals for pedestrian-vehicle flows and the movement directions for evacuees. An illustrating example concerning the evacuation around the M&T stadium area has been used to demonstrate the promising properties of our model.
Authors: Xin Zhang, Gang-Len Chang, and Alvin Marquess
Abstract: Evacuating large municipal areas during emergencies in an efficient manner is one of the critical concerns of most responsible management agencies. Previous studies focus mainly on strategic evacuation plans or controls for the passenger cars, giving inadequate attention to those pedestrians community with transit systems or other modes especially in metropolitan areas. This study has developed an evacuation planning module for pedestrians in Washington D.C. and integrates it with the system developed for passenger-car evacuation. The proposed module first guides people to the nearest metro-stations, and then applies the knowledge-based method to choose proper evacuation routes for shuttles to pick up evacuees. It also includes the plans for guiding pedestrians to the nearest evacuation route, and for dispatching shuttles need for their evacuation.
Authors: Zichuan Li, Gang-Len Chang, and Suhasini Natarajan
Conference: The 11th International IEEE Conference on Intelligent Transportation System, October 2008 in Beijing, China
Status: Presented
Abstract:
Ramp metering is a very important method to improve freeway traffic system performance, but current researches emphasize only on the entering flow control, and not much attention has been paid to the exiting volume at offramp. In some situations, the exiting queue will decrease the overall system performance by a large magnitude. This is demonstrated by simulation experiments and field observation. In this study, a mix integer model is proposed to optimize the arterial and off-ramp control signal timing based on cell transmission traffic propagation model. A Genetic Algorithm based solution algorithm is proposed along with a numerical case study to demonstrate the benefits of this model.
Authors: Jie Yu, Yue Liu, Gang-Len Chang, Xiao-guang Yang
Conference: Transportation Research Board 2008
Status: Presented
Abstract:
Choosing proper locations of urban transit hubs has always been one of the critical concerns facing urban transportation planning agencies in China. This study proposes a mixed integer optimal location model for urban transit hubs, with the objective to minimize the demand-weighted total travel time, when explicitly taking into account Traffic Analysis Zones (TAZs) as demand origins or destinations in a target urban area. An Integer Non-linear Programming (INLP) reformulation was developed to significantly reduce the number of variables. Bilinear constraints in the proposed INLP formulation were then re-modeled into linear functions to ensure global optimal solutions obtained. The model was successfully applied to optimize the hub locations in Suzhou Industrial Park, China, with the result of significantly improved system performance. The impacts of several critical factors, such as the number of hubs and the travel time discount coefficient on the system performance are also investigated.
Authors: Jie Yu, Gang-Len Chang, H.W. Ho and Yue Liu
Conference: The 11th International IEEE Conference on Intelligent Transportation System, October 2008 in Beijing, China
Status: Presented
Abstract:
This paper proposes a variation-based online travel time prediction approach using clustered Neural Networks with traffic vectors extracted from raw detector data as the input variables. Different from previous studies, the proposed approach decomposes the corridor travel time into two parts: 1) the base term, which is predicted by a fuzzy membership-value-weighted average of the clustered historical data to reflect the primary traffic pattern in the corridor; and 2) the variation term, which is predicted through the calibrated cluster-based artificial neural network model to capture the actual traffic fluctuation. To evaluate the effectiveness of the proposed approach, this paper has conducted intensive numerical experiments with simulated data from the microscopic simulator CORSIM. Experimental results under various traffic volume levels have revealed the potentials for the proposed method to be applied in online corridor travel time prediction.