Bernhard Heilmann

Predicting Motorway Traffic Performance by Data Fusion of Local Sensor Data and Electronic Toll Collection Data

In this article data fusion from different sources in order to improve estimation and prediction accuracy of traffic states on motorways is proposed. This is demonstrated in two case studies on an intraurban and an interurban motorway section in Austria. Data fusion combines local detector data and speed data from the Electronic Toll Collection (ETC) system for heavy goods vehicles (HGV). A macroscopic model for open motorway sections has been used to estimate passenger car and HGV density, applying a standard state-space model and a linear Kalman filter. The resulting historical database of 4 months of speed-density patterns has been used as a basis for pattern recognition. A nonparametric kernel predictor with memory length of 9 and 18 hours has been used to predict HGV speed for a prediction horizon of 15 minutes to 2 hours. The results showed a good overall prediction accuracy. Correlation analysis analysis showed little bias of predicted speed for free flow and congested time intervals, whereas transition states between free flow and congestion were frequently biases. The prediction accuracy can be improved by applying a combination of different prediction methods. However, the computational performance of the predictor needs to be further improved prior to implementation into a traffic management center.

An assessment of VMS-rerouting and traffic signal planning with emission objectives in an urban network — A case study for the city of Graz

This paper discusses a case study evaluating the potential impact of ITS traffic management on CO2 and Black carbon tailpipe emissions. Results are based on extensive microsimulations performed using a calibrated VISSIM model in combination with the AIRE model for calculating the tailpipe emissions from simulated vehicle trajectories. The ITS traffic management options hereby consist of easily implementable actions such as the usage of a variable message sign (VMS) or the setting of fixed time signal plans. Our simulations show that in the current case shifting 5% of vehicles from one route to another one leads to an improvement in terms of emissions only if the VMS is complemented with an adaptation of the signal programs, while the VMS sign or the change of the signal plans alone do not yield benefits. This shows that it is not sufficient to evaluate single actions in a ceteris paribus analysis, but their joint network effects need to be taken into account.

A data-driven approach for travel time prediction on motorway sections

In the presented case study, travel times for passenger cars (PC) and heavy goods vehicles (HGV) were predicted with a data-driven, hybrid approach, using historical traffic data of the entire high-ranking Austrian road network. In case flow data were available, travel time was predicted with a Kernel predictor searching for similar speed-density patterns. In case of missing flow data, travel time was predicted with deviations from typical historical speed time series. The performed steps in pre-processing traffic data, the hybrid prediction method as well as the results for selected road sections are described and analysed.

Linking statistical mobility data with electrical distribution network infrastructure for generating an agent population for multi-agent simulation of electric vehicles with Markov chains

To create realistic travel chains Markov Chains and Monte Carlo method are used. This includes different travel purposes in different locations, information about departure/arrival times and the distance driven. Individual agents are then generated which represent the EV population of a certain area. Depending on the defined charging infrastructure in the specific grid and the chosen EV model, this agent population can be calibrated for low- and medium voltage grids of different areas as there are rural or urban areas. Different charging strategies are simulated and their effect to the power grid analyzed.

Detection of faulty flow measurements based on queueing models and measured delays at intersections

Results of a case study for two signalized intersections in the Viennese intraurban road network have shown that detectors delivering artificially introduced faulty flow measurements can be detected. The method compares the theoretical delay distribution based on a deterministic queuing model to the empirical delay distribution of measured flow and speed values. A prerequisite is that the measurements at the intersection cover both under- and over-saturated traffic situations. The described method can detect several types of errors, namely overcount, undercount and Gaussian noise errors. The sensitivity of the method depends on the intensity of the error, i.e. relative error and standard deviation. As the method has achieved good results for artificially introduced errors, similar errors in real data will also be detected, since the investigated error types are typical for urban flow measurements.