Docstoc

Agent Base Models for Integrated Urban Water Management

Document Sample
Agent Base Models for Integrated Urban Water Management Powered By Docstoc
					Agent-Based Models for Integrated
   Urban Water Management


                Arlex Sanchez Torres
                       Promoter
           Prof. Dimitri Solomatine , PhD
                       Mentors
                  Z. Vojinovic, PhD
                 Prof. R.K. Price, PhD
The Challenge
                Urbanization growth, on one
                hand is considered as a symbol of
                economical development, brings
                out on the other hand more
                complexity and stress to the
                natural systems and resources.


                There is a call for innovative
                thinking and redesign of the
                urban water cycle to achieve
                environmental sustainability.
                SWITCH, 2006.
To achieve a holistic analysis and assessment of effects of measures,
an integrated modeling approach is required. In this context,
computer based simulation models are accepted planning tools



                                       Huge Amount of Data
                                       Data Management
                                       Summarized and Presented
                                       Clearly
                                       Essential Information




                                       Decision Support System
           Objectives (Arlex)
• Explore the use of agent-based technology in
  urban water management
• How to apply the concepts and principles of
  emergence in the development of urban areas
• How to consider random influences on urban
  developments. Risks and opportunities.
• Application of optimization procedures for the
  calibration of agent-based models.
      Agent based simulation (ABS)
• The Agent-based simulation (ABS) is a simulation technique that
  provides a way for a direct representation of phenomena
• Agent can represent human individuals, organizations, social institutions
• ABS begin with the creation of agents and environment and finish with
  the analysis of the observed collective behavior
• Agent based models consist of dynamically interacting rule based agents
• The systems within which they interact can therefore create complexity
  like that which we see in the real world.
• The fundamental feature of an agent is the capability of the component
  to make independent decisions
ABMS tools are not developed to forecast the exact state of the modeled system,
but to explore how the system will evolve in view of a possible future.

Nostrum-DSS - 2006
            FIRMABAR simulator
• The FIRMABAR is an agent based simulator for integrated
  freshwater assessment
• Has been developed and successfully apply to the
  metropolitan region of Barcelona and Valladolia
• Provide policy makers with additional tool to evaluate
  alternative water policies in different scenarios
• Base on water cycle and the role that agents play in it
        ABSS approach to the system
• The simulator has been design as 2 coupled model
   – Territorial model
       • Represent physical and urban space
       • Base on cellular automata
   – Social model
       • Represent the elemental agent (family)
    The world cellular water game




Valkering et al 2007 University of Maastricht
Net Logo Examples
 People’s behavior during flood




Hossein, New Casttle 2007
GIS - Functionality
Comparing with 1D-2D physically
base model
Models for Urbanization - Tijuana Bordertowns model
Models for Urbanization - Cities
Procedural modeling of cities




                                Lechner et all, 2004
Initial Conditions          At t = 5 years (ie)           At t = 20 years (end
At t = 0                    New Developments              simulation)
Layout of the initial       are build, new                Performance of the
water distribution          layout for the                projected system?
network and                 network (new                  Impacts?
performance                 developments and
information of the          densification of old)
system (Projected)          assess
                            performance.
                        Intermediate step needs a
                        loop to optimize the system
                        (capacity of pipes, tanks, etc)
Initial Conditions   Layout of the initial   Simulation at t= n     Optimization to find
Land use, socio-     drainage network        New developments       the best possible
economics,           and performance         and densification of   set of solutions for
regulations          information of the      existing. Several      the projected city.
At t = 0             system (flooding)       options for the
                                             drainage system
                                             depending on the
                                             new environmental
                                             impacts.
•The outcome of the simulation should produce some sensible
results of the projected urban future and scenarios.

•The agents can decide where to lay-down new pipes and
extend the water distribution and drainage network and add
special features like tanks, pumps, ponds, WWTP, etc.

•That will allow the construction of scenarios to evaluate the
environmental impact and/or sustainability of the “new city”
and the evolution of the “water related” infrastructure in the
city.

•Incorporate environmental rules that affect urban growth,
water availability, water quality, etc.
The Great Dublin Area – EU Project
                       Input : Urban land use map
  The Framework        and numerical models of urban
                       water systems

Suitability maps

Considering,
elevation,
existing
infrastructure,
water bodies,
water quality,
water
availability.


                                      Feedback sub-model
                                                           New
                   Accessibility                           developments
                   map                                     (transitional
  Regional model                                           potential)
                   Road network,
  Socio-           Pipe networks,
  economics,       State natural
  regulations,     streams.
  master plans,
  legislation.
The Feedback Model – Dynamic Suitability




                         New Developments




                     New Suitability
Neighborhood and Rules
1978
Water Pressure
Map overlay
JRC – EC, 2005
            Next Steps
•Develop and test with small case
study (continue)
•Integration with Water Models
•Start Belo Horizonte Case,
Birmingham

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:10
posted:5/15/2012
language:English
pages:30