The Dynamic Geometry of Geographical Vector Agents

Yasser Hammam, Antoni Moore, Peter Whigham and Claire

Freeman*Spatial Information Research Centre,

Department of Information Science,

*Department of Geography

University of Otago, New Zealand

Overview

RationaleThe Vector AgentModel Implementation and

Experimental ResultsDiscussionConclusions

IntroductionLike Geographic Automata Systems,

Vector Agents aim to introduce a bit of geographic realism to agent modellingBut adds a systematic framework to the

geometric element (“georeferencing convention”)

With emphasis on irregular and dynamic aspects

Aim to use boundary manipulation through simple controls to generate vector objects of a wide variety of shapes and complexitiesPertinence to real world object characteristics

are key to effectiveness

The Vector AgentUses irregular fractal-like process to

generate vector objectsAlso more direct boundary manipulation

AimsTo represent any discrete geographic

phenomena through an irregular (or regular) data structure

May move “bodily”, either based on a real world object, or is “born” with a non-deterministic shape boundary

Abstracted so that it is able to define its own location in space

Regular – irregularStatic – dynamic

Geometric manipulationConventional midpoint displacement

Pnew = 0.5 (P1 + P2) + µơ02-lh

Where P1 and P2 are the start and end points of the line segment

µ is a random number from a Gaussianơ0 is the S.D. of the Gaussianl is the level of recursivityh is the Hurst exponent governing roughness

(= 2 – F.D.)

Point displacement (not nec.midpoint)Pnew = (1 – r)P1 + rP2 + µơ02

-lh Where r is proportion along line segment

Edge / vertex displacementP = P + µơ0

Sequence of growth

b c d e fa

g h i j

Results show evidence of both irregular and regular

Graphs

Shape control schematic

SHAPE GROWTH RATEhigh low

Control of

shape complexity

high

low

MIDPOINT DISPLACEMENT

VERTEX DISPLACEMENT

EDGE DISPLACEMENT

Cities with similar shapes

ConclusionVector agents are geometry-led

agentsInterplay of midpoint, edge and vertex

displacementEvolved polygons have shape and

complexity characteristics of real-world objects

Able to be controlled by alteration of simple parameters

ConclusionLike GAS, each object has its own

identity, an improvement on a group of contiguous CA cells

But complexity and processing speed ramped up

NextTest the system on a specific urban area

Having tested on classical urban models 1st

Build in other elements of GASStates, transition rules, neighbourhood,

neighbourhood rules