members & address research industrial partners teaching publications gallery home page of the group
    clickable logo

Combinatorial Optimization & Graph Algorithms

TU logo

contents
.

department
 .  group
 .  .  members & address
 .  .  research
 .  .  .  network flows
 .  .  .  . dynamic flows
 .  .  .  .  k-split flows
 .  .  .  .  path-based flows
 .  .  .  .  route guidance
 .  .  publications
 .  .  cooperation with industry
 .  .  teaching
 .  .  project gallery
 .  .  events
 .  .  internals
 .  .  search

Dynamic Flows

Time plays an important role in transport:

  • People use cars to move, often a vast number of cars at the same time on the same route. Still they want to reach their destinations fast.
  • Huge amounts of data run through communication networks. Network capacities limit the throughput and data sent along a connection block this way for other data. Nevertheless, nobody wants to wait to get data.
  • Goods traverse complex production processes. Production costs are often influenced by the pass-through time.
All these time-oriented processes benefit from good structure and control. Planning them requires an appropriate model.

There is a variety of very different mathematical approaches to model dynamic behaviour. In combinatorial optimization it can be modelled as flows in networks. The underlying structure, for example streets and junctions in traffic systems, is described as a network. Moving objects like cars are interpreted as flow in this network.

Research in flow theory from the last forty years has been focussed on static flows. Such flows do not reflect time. Flows over time or dynamic flows as a generalization of static flows provide an appropriate way to include time aspects. For both static and dynamic flows, underlying networks consists of nodes and arcs with arc capacities. In static flows, to every arc a flow value is assigned limited by the given arc capacity. Flows over time, however, are described by a flow rate which enters an arc at a specified time. Arc capacities limit the amount of flow entering an arc at the same time and transit times describe how long it takes to travel over an arc.

Our work is strongly related to applications in traffic networks. We investigate various theoretic questions in the context of dynamic flows.



Our work

2003:
2002:
2001:
top top
source last modified: Thu Aug 26 2004, last built: Thu Aug 26 2004
Ines Spenke <spenke@math.tu-berlin.de>
Validate HTML