List of Figures List of Tables. Chapter 1: Introduction 1

Transcription

1 Contents List of Figures List of Tables iii viii Chapter 1: Introduction 1 Chapter 2: Study of Pedestrian Behaviors in Urban Space Effects of Space Configuration and Attraction on Spatial Behavior Simulation Pedestrian Dynamics and Local Movement Behavior at the Individual Level 20 Chapter 3: Implementing Simulations as Individual-Based Models Introduction to an Individual-based Model Example Models Continuous Individual-based Simulation Discrete Individual-based Simulation Cellular Automata Object-oriented Programming System Architecture 32 Chapter 4: Representing Environment 38 Chapter 5: Representing Pedestrian Behaviors Structure of an Agent "Mouse" External Perception Internal States Behavior Production System Goal Structure 49 i

2 5.3 Modeling Individual Behavior Reflex Behavior Reactive Behavior Motivated Behavior Modeling Social Behavior Imitation of Behavior Induction of Behavior 68 Chapter 6: Experiments About the Simulation Exploring Dynamic Individual Behavior Experiment 1: Changes of Behavior Experiment 2: Changes of Environment Exploring Complex Social Interaction Experiment 3: Pedestrian Flow Sensitive Experiment 4: Flow of Information 88 Chapter 7: Discussion 89 References 92 Appendix A: A Guide to the Accompanying CD ROM 100 Pocket Material: CD ROM ii

3 List of Figures Number 1.1 Westlake Plaza Crist Langton s Emergence diagram illustrating the concept of complexity (Langton, 1995) Emergent Property: A circular mill of army ants (Langton, 1995) Example of city simulation based on the idea of Fractal City (Batty, 1994) Simulation scenes from Mouse.class project Pedestrian activities from the book The Social Life of Small Urban Spaces (Whyte, 1980) Mouse Palace, a Thai-Chinese traditional toy for a child to learn about behavior and environment relationship while playing with mice, food, and wooden blocks Space syntax analysis: Axial map global integration: London (Hillier, 1993) Moment of two pedestrian groups trying to pass a narrow door. If one person has been able to pass the door, others with the same desired walking direction can follow easily (Helbing, 1994) At crowded intersections there develops roundabout traffic for certain time periods (Helbing 1994) An urban environment representing significant places (Jiang, 1999) A snapshot of simulation of pedestrian moving purposely (Jiang, 1999) Sugarscape simulation scene: yellow areas represent sugar, red and blue dots are agents (http: Hierarchy of individual behavior (Based on: Baerends, 1976; McFarland, 1981; Gerhenson, 2001) A hierarchy of motion behaviors (based on Blumberg, 1995; Reynolds, 1998; Gerhenson, 2001) Simulated boid flock avoiding obstacles (Reynolds, 1986) Steering Behavior (Reynolds, 1999) 26 iii

4 3.3 Demonstrations of steering behavior (Reynolds, The aggregation process of 100 square cells The Village of Seripe, after Mumtaz (Hillier, 1984) Model of termites piling woodchips developed with the use of StarLogo. (Resnick, 1994) Regular uniform of Cellular Automata The model of multiple lanes and separated flows (Blue, 1999) System Architecture Construction of the system Three (separate) basic components in an urban scene (Thiel, 1997) Combination of the three basic components (Thiel, 1997) Diagram shows blocks representing different geometry in urban space. (photographs of urban spaces from the book Public Spaces Public Life by Jan Gehl, 1996) Diagram shows cheese representing attractive elements in urban space. (photographs of urban spaces from the book Public Spaces Public Life by Jan Gehl, 1996) Gates defined at the edge of space (shown in the orange lines) Field of vision based on isovist concept A scale of motivation degree of Hunger Left: a state of very hungry, Right: a neutral state A scale of motivation degree of hurriedness Left: high degree indicates rush, Right: low degree indicates relax) Behavior production system showing connections with external perception and internal system. The feedback loops (red arrows) of internal drive and external perception are important in determining an appropriate behavior Final goal and motivated goal from the simulation scene Smooth walking line (Kovacs and Galle, 1993) Cutting off the corners of lawn to make a short cut (Gehl, 1996) Sketch Diagram of pink s default behavior, Wander 52 iv

5 5.9 Flow chart of Wander process From a simulation scene: Pink mice - wander walkers - following their temporary goals (end points of each gray lines) Sketch diagram of blue s default behavior, Pass-Through Flow chart of Pass-Through process Sketch diagram of pink s avoiding obstacle case Sketch diagram of pink s avoiding obstacle case Flow chart of pink s Avoid Obstacle process The actual path: using a series of intermediate goal to arrange the path towards the destination (Alexander, 1977) Sketch diagram of blue s avoiding obstacle Flow chart diagram of blue s Avoid Obstacle process From a simulation scene: A Blue - purposive walker picking a temporary goal (end point of a gray line) to avoid an obstacle then proceeding to a final goal (end point of an orange line) Sphere influence Neighborhood Field Two situations in neighborhood field (evasive distance). Situation A: no moving object in range and situation B: moving object is in range, preparing to take evasion to avoid collision Sketch diagram of collision avoidance Flow chart diagram of collision avoidance From a simulation scene: blues and pinks perform obstacle avoidance if others are in their neighborhood fields and their directions get crossed Flow chart diagram shows local decision ' action selection' process of pinks Flow chart diagram shows local decision ' action selection' process of blues Graphics describes imitation and induction of behavior Imitation of neighbors' directions Blue-cheese story Blue-cheese story Sketch diagram of induction of behavior Applet Interface 72 v

6 6.2 The vicinity of simulation scene including the ideal space surrounded the scene Two typical types of mice Experiment 1, Scenario 1: Pinks moving in context-free environment Experiment 1, Scenario 1: Blues moving in context-free environment Experiment 1, Scenario 2: Pinks avoiding obstacles Experiment 1, Scenario 2: Blues avoiding obstacles Experiment 1, Scenario 3: No Green space Experiment 1, Scenario 3: A Green space Experiment 1, Scenario 4: Changing Behavior Simulation Experiment 1, Scenario 4: Changing Behavior simulation The trace of movement pattern of the above scene showing the distribution of movement through time Experiment 2: Simulation Scene Scene 1: Traced Pattern Experiment 2: Simulation Scene Scene 2: Traced pattern Experiment 2: Simulation Scene Scene 3: Traced Pattern Pedestrian Flow Flow sensitive model running through 547 cycles Mouse conditions: Yellow mouse carries cheese; Orange mouse is hungry Flow of information simulation 88 A 1 Westlake Plaza 100 A 2 Social Life of Small Urban Spaces 100 A 3 Track the move 101 A 4 Mouse Palace 101 A 5 Parallel worlds 101 A 6 Types of behaviors 102 A 7 Individual behavior 104 A 8 Social behavior 104 A 9 System Architecture 105 vi

7 A 10 Autonomous agent Mouse 105 A 11 Mouse ability 105 A 12 Forward moving 106 A 13 Modeling forward moving 106 A 14 Obstacle avoidance 107 A 15 Modeling obstacle avoidance 107 A 16 Collision avoidance 108 A 17 Modeling collision avoidance 108 A 18 Motivated behavior 109 A 19 Simulation scene 110 A 20 Movement pattern 111 A 21 Imitation of behavior 112 A 22 Blue Cheese Story 113 A 23 Induction of behavior 114 vii

8 List of Tables Number 2.1 When the quality of outdoor environment is good, optional activities occur with increasing frequency. Furthermore, as level of optional activity rise, the number of social activities usually increases substantially (Gehl, 1996) Behavior repertoire of an agent mouse to perform the movement in the simulation Interpersonal Distance based on the North American experience and suggested by Edward T. Hall, The Hidden Dimension The internal states, Motivation degree and possible actions The combination of two internal states and adjustment of motivation degree 69 viii