Estimating vehicle speed with embedded inertial sensors

Pavements were instrumented with inertial sensors, and the possibility of estimating the speed of a passing vehicle was investigated numerically and experimentally from the measurements of two embedded accelerometers. The sensors were spaced apart in the travel direction, and subsequently the speed was directly related to the time delay between the received signals. No assumption was made regarding the vehicle and pavement properties. Model accelerations were presented, studied, and contrasted against field measurements; the latter were shown to be dominated by random vibration sources. Two calculation techniques were offered and applied to handle the noisy data. The first was based on time-centroids, and the second was based on cross-correlation with kernel presmoothing. The overall concept is deemed promising not only for inferring speeds but also for extracting additional traffic characteristics such as axle spacing and relative axle load distributions.     

Length-based vehicle classification using event-based loop detector data

Length-based vehicle classification is an important topic in traffic engineering, because estimation of traffic speed from single loop detectors usually requires the knowledge of vehicle length. In this paper, we present an algorithm that can classify vehicles passing by a loop detector into two categories: long vehicles and regular cars. The proposed algorithm takes advantage of event-based loop detector data that contains every vehicle detector actuation and de-actuation “event”, therefore time gaps between consecutive vehicles and detector occupation time for each vehicle can be easily derived. The proposed algorithm is based on an intuitive observation that, for a vehicle platoon, longer vehicles in the platoon will have relatively longer detector occupation time. Therefore, we can identify longer vehicles by examining the changes of occupation time in a vehicle platoon. The method was tested using the event-based data collected from Trunk Highway 55 in Minnesota, which is a high speed arterial corridor controlled by semi-actuated coordinated traffic signals. The result shows that the proposed method can correctly classify most of the vehicles passing by a single loop detector.     

Automated parking surveys from a LIDAR equipped vehicle

Parking surveys provide quantitative data describing the spatial and temporal utilization of parking spaces within an area of interest. These surveys are important tools for parking supply management and infrastructure planning. Parking studies have typically been performed by tabulating observations by hand, limiting temporal resolution due to high labor cost. This paper investigates the possibility of automating the data gathering and information extraction in a proof of concept study using a two-dimensional scanning Light Detection and Ranging (LIDAR) sensor mounted on a vehicle, though the work is compatible with other ranging sensors, e.g., stereo vision. This study examines parallel parking in the opposing direction of travel. The ranging measurements are processed to estimate the location of the curb and the presence of objects in the road. Occlusion and location reasoning are then applied to determine which of the objects are vehicles, and whether a given vehicle is parked or is in the traffic-stream. The occupancy of the parking area, vehicle size, and vehicle-to-vehicle gaps are then measured. The algorithm was applied to an area with unmarked, on-street parking near a large university campus. Vehicle counts from 29 trips over 4 years were compared against concurrent ground truth with favorable results. The approach can also be applied to monitor parking in the direction of travel, eliminating the possibility of occlusions and simplifying the processing.     

Estimating historical hourly traffic volumes via machine learning and vehicle probe data: A Maryland case study

This paper focuses on the problem of estimating historical traffic volumes between sparsely-located traffic sensors, which transportation agencies need to accurately compute statewide performance measures. To this end, the paper examines applications of vehicle probe data, automatic traffic recorder counts, and neural network models to estimate hourly volumes in the Maryland highway network, and proposes a novel approach that combines neural networks with an existing profiling method. On average, the proposed approach yields 24% more accurate estimates than volume profiles, which are currently used by transportation agencies across the US to compute statewide performance measures. The paper also quantifies the value of using vehicle probe data in estimating hourly traffic volumes, which provides important managerial insights to transportation agencies interested in acquiring this type of data. For example, results show that volumes can be estimated with a mean absolute percent error of about 21% at locations where average number of observed probes is between 30 and 47 vehicles/h, which provides a useful guideline for assessing the value of probe vehicle data from different vendors.     

SIV-DSS: Smart In-Vehicle Decision Support System for driving at signalized intersections with V2I communication

In this paper, we present a Smart In-Vehicle Decision Support System (SIV-DSS) to help making better stop/go decisions in the indecision zone as a vehicle is approaching a signalized intersection. Supported by the Vehicle-to-Infrastructure (V2I) communications, the system integrates and utilizes the information from both vehicle and intersection. The effective decision support models of SIV-DSS are realized with the probabilistic sequential decision making process with the capability of combining a variety of advantages gained from a set of decision rules, where each decision rule is responsible to specific situations for making right decisions even without complete information. The decision rules are either extracted from the existing parametric models of the indecision zone problem, or designed as novel ones based on physical models utilizing the integrated information containing the key inputs from vehicle motion, vehicle-driver characteristics, intersection geometry and topology, signal phase and timings, and the definitions of red-light running (RLR). In SIV-DSS, the generality is reached through physical models utilizing a large number of accurate physical parameters, and the heterogeneity is treated by including a few behavioral parameters in driver characteristics. The performance of SIV-DSS is evaluated with systematic simulation experiments. The results show that the system can not only ensure traffic safety by greatly reducing the RLR probability, but also improve mobility by significantly reducing unnecessary stops at the intersection. Finally, we briefly discuss some relevant aspects and implications for SIV-DSS in practical implementations.     

Vehicle scheduling under stochastic trip times: An approximate dynamic programming approach

Due to unexpected demand surge and supply disruptions, road traffic conditions could exhibit substantial uncertainty, which often makes bus travelers encounter start delays of service trips and substantially degrades the performance of an urban transit system. Meanwhile, rapid advances of information and communication technologies have presented tremendous opportunities for intelligently scheduling a bus fleet. With the full consideration of delay propagation effects, this paper is devoted to formulating the stochastic dynamic vehicle scheduling problem, which dynamically schedules an urban bus fleet to tackle the trip time stochasticity, reduce the delay and minimize the total costs of a transit system. To address the challenge of “curse of dimensionality”, we adopt an approximate dynamic programming approach (ADP) where the value function is approximated through a three-layer feed-forward neural network so that we are capable of stepping forward to make decisions and solving the Bellman’s equation through sequentially solving multiple mixed integer linear programs. Numerical examples based on the realistic operations dataset of bus lines in Beijing have demonstrated that the proposed neural-network-based ADP approach not only exhibits a good learning behavior but also significantly outperforms both myopic and static polices, especially when trip time stochasticity is high.     

Vehicle routing with roaming delivery locations

We propose the vehicle routing problem with roaming delivery locations (VRPRDL) to model an innovation in last-mile delivery where a customer’s order is delivered to the trunk of his car. We develop construction and improvement heuristics for the VRPRDL based on two problem-specific techniques: (1) efficiently optimizing the delivery locations for a fixed customer delivery sequence and (2) efficiently switching a predecessor’s or successor’s delivery location during the insertion or deletion of a customer in a route. Furthermore, we conduct an extensive computation study to assess and quantify the benefits of trunk delivery in a variety of settings. The study reveals that a significant reduction in total distance travelled can be achieved, especially when trunk delivery is combined with traditional home delivery, which has both economic and environmental benefits.     

Specification and calibration of a microscopic model for pedestrian dynamic simulation at signalized intersections: A hybrid approach

Macroscopic pedestrian models for bidirectional flow analysis encounter limitations in describing microscopic dynamics at crosswalks. Pedestrian behavior at crosswalks is typically characterized by the evasive effect with conflicting pedestrians and vehicles and the following effect with leading pedestrians. This study proposes a hybrid approach (i.e., route search and social force-based approach) for modeling of pedestrian movement at signalized crosswalks. The key influential factors, i.e., leading pedestrians, conflict with opposite pedestrians, collision avoidance with vehicles, and compromise with traffic lights, are considered. Aerial video data collected at one intersection in Beijing, China were recorded and extracted. A new calibration approach based on a genetic algorithm is proposed that enables optimization of the relative error of pedestrian trajectory in two dimensions, i.e., moving distance and angle. Model validation is conducted by comparison with the observed trajectories in five typical cases of pedestrian crossing with or without conflict between pedestrians and vehicles. The characteristics of pedestrian flow, speed, acceleration, pedestrian-vehicle conflict, and the lane formation phenomenon were compared with those from two competitive models, thus demonstrating the advantage of the proposed model.     

Customized bus service design for jointly optimizing passenger-to-vehicle assignment and vehicle routing

Emerging transportation network services, such as customized buses, hold the promise of expanding overall traveler accessibility in congested metropolitan areas. A number of internet-based customized bus services have been planned and deployed for major origin-destination (OD) pairs to/from inner cities with limited physical road infrastructure. In this research, we aim to develop a joint optimization model for addressing a number of practical challenges for providing flexible public transportation services. First, how to maintain minimum loading rate requirements and increase the number of customers per bus for the bus operators to reach long-term profitability. Second, how to optimize detailed bus routing and timetabling plans to satisfy a wide range of specific user constraints, such as passengers’ pickup and delivery locations with preferred time windows, through flexible decision for matching passengers to bus routes. From a space-time network modeling perspective, this paper develops a multi-commodity network flow-based optimization model to formulate a customized bus service network design problem so as to optimize the utilization of the vehicle capacity while satisfying individual demand requests defined through space-time windows. We further develop a solution algorithm based on the Lagrangian decomposition for the primal problem and a space-time prism based method to reduce the solution search space. Case studies using both the illustrative and real-world large-scale transportation networks are conducted to demonstrate the effectiveness of the proposed algorithm and its sensitivity under different practical operating conditions.     

CO2 emissions correction of light-duty vehicles based on an energy deviation analysis method

In present study, a total of 3 types of light-duty vehicles were tested for 7 types of driving modes by using a vehicle dynamics-based simulation method. Deviations between the target and actual test vehicle speed and inaccurate chassis dynamometer road load settings are selected as the two main factors affecting the CO₂ emission measurement test. 126 cases of simulations were conducted to analyze the effects of the deviation between the target and actual test vehicle speed on CO₂ emissions and 63 cases of simulation were conducted to analyze the effectiveness of inaccurate road load test conditions on CO₂ emissions. Then, the CO₂ correlation method proposed by World Forum for Harmonization of Vehicle Regulations (WP.29, UN) was applied for simulated results in original simulation cases, and a correction to performance was validated by comparative analysis. From the simulation results, it is revealed that velocity and road load deviation are related to distinct changes in fuel efficiency and CO₂ emission rates of test vehicles. Also, it is revealed that CO₂ correction method show good performance to reduce the relative error lower than 2% in all test cases.