• Andy Turner's PhD Research Outline Version 2.1

  • [An image of Andy Turner]

  • Development of Agent Based Models of Daily Activity

    • Introduction

      This PhD sets out to solve a number of current issues in large-scale agent-based modelling. These will be addressed by the development of a sound and rigorous framework for agent-based modelling that tackles these problems. The framework will be validated by using it to construct a regional development model.

      The chief problems examined by this thesis will be:

      1. How may we best create an individual-scale model of the UK and link this with useful non-agent-based and agent-based models?
      2. What techniques might we use for validating models that are both spatial and involve data/commodity exchange on networks and via global processes?
      3. How do we best visualize, quantify and control the propagation of errors through large scale, dynamic, and complex systems?

      The agent based model (ABM) will be similar to that of the SimCity computer game (EA games, 2006). The model agents will be individual people agents, most of these will be grouped into household agents and family agents. Additionally agents will be:

      • villages, towns and cities as functional entities
      • businesses (including public or private institutions - service bodies).

      Connections between these agents will provide for modelling of social and business networks.

    • Case Study

      The contemporary UK will be studied. This is appropriate because of good data availablitiy, and because many issues of organisation and policy operate between the individual and national level. Key to the agent-based models will be calculations of accessibility and mobility, and the push and pull factors associated with locational change in agents. For individuals and businesses these are the relationships between, and influences on, migration and commuting.

    • Description of Work

      The ABM will be run to produce dynamic simulations that can be visualised as animated maps with states that can be preserved along with provenance. Mechanisms will be built into the framework for testing the sensitivities of simulations to random factors and changes in model configuration and parameters. This is important for reasons of repeatability and testing the verisimilitude of any given model. In addition, part of the reliability of such models is an understanding of the mechanism by which errors are enlarged or dampened during model runs. Such errors are usually ignored in agent-based models under the erroneous assumption that behavioural models do not suffer from the same kinds of numerical errors that more mathematical model suffer. The largest problem in this area is the location and visualization of such errors, which needs to be addressed before they can be investigated in detail.

      The final model used during the testing of the system will be decided after the first six months of work, as the social system chosen will be dependent on the model capabilities and also on the models chosen for associated work external to the PhD. However, viable candidate systems are available, including the modelling of regional development; industrial manufacturing changes or the housing market.

      A conceptual model of daily activity

      • Start off with everybody at a home or residence.
      • Each person is an agent and they choose by random an available transport mode to work.
      • Move people from home to work and back again on a 15 minute time step.
      • Each day a new transport mode is chosen based on the weather and experience of previous journeys.
      • Try to settle things down to get an average journey mode profile for aggregate census areas.
      • Compare this with 1991 SWS mode of transport data.
    • First attempt at a method

      • Using CAS, determine the numbers of people in various occuptations and industries at a fine scale (e.g. OAs).
      • Use ST Theme Table TT010 from 1991 Census, which has daytme and workplace populations at ward level.
      • Create a model with known transport infrastructure.
      • Stage 1: agent model the people in different occupations.
        • Move the agents to work, such that you end up with the right number of people in different occupations at the workplaces.
        • What patterns does this produce?
      • Stage 2: Attempt to improve the model using the SWS, which tells you where (in aggregate) the people living in a given location go to work.
        • Are the agents in Stage 1 going to the 'right' locations? If not: are they going to an economically/environmentally/rational 'better' location? Or is the model simply wrong!
      • Stage 3: Attempt further improvement on the basis that SWS3 provides high resolution flows by transport mode.
        • Are the agents using the 'right' transport? etc...
      • If there is reasonable success in getting agents to go to the right locations using the right transport: a) hurray! b) try changing the transport infrastructure.
      • Somewhere in all of this, compare results to a standard stocks based spatial interaction model.

      References

      • EA games (2006) SimCity4.http://simcity.ea.com/ accessed 10-01-2006.

      Acknowledgments

      • Thanks to my supervisors Andy Evans and Mark Birkin for reviewing a draft of Andy Turner's PhD Research Outline Version 2.0 on which this is based.
      • Thanks to Oliver Duke-Williams for our valuable discussions regarding the development of a method for testing and validating the agent based models.
  • Validation and Metadata

    • [Validate RDF] [Validate CSS]
    • This is Version 0.0.2 of this page published on 2006-11-03. Page based on Andy Turner's PhD Research Outline Version 2.0
    • Page hosted on the School of Geography webserver at the University of Leeds.
    • Copyright: Andy Turner, University of Leeds