Topics for master students
The list below gives some ideas on what could be relevant and useful projects for students who want to do a master thesis with me. Feel free to propose other subjects yourself, this is just to have some input. The list is based on the scientific challenges I currently see arising.
Lane changing
Combination of lane changing and speed changes
On the motorway, lane changes can be related to speed changes. This can be analysed from remote sensing data. When do people change lanes, and when do they accelerate? Are these coupled? (leaflet here)Incentives of lane changing (with ACC?)
Lane changes are possibly induced by a desire to go faster, or not change speeds. What can be derived as explanatory variables for a lane change. This is probably best studied using a driving simulator. How does this change when the vehicle is equiped with a cruise control, or a cruise control which keeps distance automatically? (leaflet here) What is the influence of the speed limit and the check on the speed limit on lane distribution and headways
In particular in zones with average speed check, the variation in speed goes down, and driver seem to use the left lane less (leaflet here)Mixed traffic (bicycles and car traffic)
What is the impact of bicycles on the traffic flow?
Can we quantify this effect? What dependencies does it have? Can we express a pce-value for bikes, and is this different for different roads and/or speeds? (leaflet here)Network fundamental diagram for mixed traffic
Does the Network Fundamental Diagram also hold for mixed traffic? What are the dependencies of car speed on bike traffic and vice versa?(leaflet here)Network fundamental diagram
Calibrate an zone based traffic model for a real life situation
A dynamic traffic model, the Network Transmission Model, is developed based on principles of the Network Fundamental Diagram. It is not clear though how accurate the model is. In this study, you model a city (The Hague or Rotterdam) by the Network Transmission Model. By adapting the parameters in the model, you calibrate the model. You will aslo validate the model. (leaflet here) Assess the inpact of the chaning route length within a zone on the Network Fundamental Diagram
Earlier work showed that there is a strict relation between the flow in an area and the outflow out of an area. The requirement for this relation to hold is that the trip length is unchanged. In this study, this is further studied. (leaflet here) Test the Network Transmission Model for a large area -- e.g, the Randstad area, in stead of a static model
In this study, you model a large area by the Network Transmission Model. This area is larger than the area the model is developed for; examples can be the Netherlands or the Randstad area. Can the model replace the static models for representing a typical day? (leaflet here)Control of traffic in large regional road networks
Different control strategies for roads are available. If complexity increases, the interactions become too complex to describe the consequences of a control action in detail. Can control based on zones be developed? (leaflet here)Combine a Network Transmission Model with the Cell Transmission Model
Other traffic flow theory topics
Features of the capacity drop
How much is the capacity drop, and which dependencies does that have? Does it depend on the upstream speed? On the type of bottleneck (fixed, wide moving jam?). Does it depend on the lane configuration upstream, or downstream. Many different research questions can be posed here. This also allows for measurements of individual drivers (going up in the air in a helicopter :-)Check the quality of time mean speeds
Based on Individial loop data: can space mean speeds be approximated by the harmonic averages of time mean speeds per laneImpact of sags on driving speed
If drivers go downhill followed by an uphill gradient, we see congestion. This is caused by drivers that do not react sufficiently on the changed gradient. What is the exact behavour? How can control be applied, for instance main line metering? Goal of a control thesis: get the measures optimized with an optimal control (Model Predictive Control) schemeImprovement of the ASM filter
The adaptive smoothing method gives a prediction of speed for all points in space and time. It does not correct for the aggregation time. Can this be implemented. The thesis work consists of developing the algorithm and testing this against real data (preferably floating car data).