I=
Pakistan Research Repository Home
 

Title of Thesis

Optimization of Route Planning in Dynamic Environments using Swarm Intelligence

Author(s)

Kashif Zafar

Institute/University/Department Details
Department of Computer Science / Fast-National University of Computer and Emerging Sciences, Islamabad
Session
2010
Subject
Computer Science
Number of Pages
134
Keywords (Extracted from title, table of contents and abstract of thesis)
Efficiency, Route, Deals, Traditional, Dynamic, Intelligence, Optimization, Swarm, Static, Technique, Using, Environments, Field, Planning, Algorithms

Abstract
Environments for algorithms can be categorized as static or dynamic.A static environment remains stationary throughout the execution of the algorithm, while in a dynamic environment the environment changes during the execution of the algorithm.The algorithms for planning in static and dynamic environments can be divided into offline and online algorithms.This research implements an online algorithm for an unknown environment and combined exploration and planning in a hybrid architecture. A simulated system of agents based on swarm intelligence is presented for route optimization and exploration.Two versions of the system are implemented and compared for performance- i.e., a simulated ant agent system and a simulated niche based particle swarm optimization. A simulated ant agent system is presented to address the issues involved during route planning in dynamic and unknown environments cluttered with obstacles and objects. A simulated ant agent system (SAAS) is proposed using a modified ant colony optimization algorithm for dealing with online route planning. The SAAS generates and optimizes routes in complex and large environments with constraints.The traditional route optimization techniques focus on good solutions only and do not exploit the solution space completely. The SAAS is shown to be an efficient technique for providing safe, short, and feasible routes under dynamic constraints, and its efficiency has been tested in a mine field simulation with different environment configurations.It is capable of tracking a stationary as well as a non-stationary goal and performs equally well as compared to moving target search algorithm.
Route planning for dynamic environment is further extended by using another optimization technique for generation of multiple routes.Simulated niche based particle swarm has been used for dynamic online route planning, optimization of the routes, and it has proved to be an effective technique. It efficiently deals with route planning in dynamic and unknown environments cluttered with obstacles and objects.A simulated niche based particle swarm optimization (SN-PSO) is proposed using a modified particle swarm optimization algorithm for dealing with online route planning. The SN-PSO generates and optimizes multiple routes in complex and large environments with constraints. The SN-PSO is shown to be an efficient technique for providing safe, short,and feasible routes under dynamic constraints. The efficiency of the SN-PSO is tested in a mine field simulation with different environment configuration, and it successfully generates multiple feasible routes. Finally, the swarm based techniques are further compared with an evolutionary algorithm (genetic algorithm) for performance and scalability. Statistical results showed that evolutionary techniques perform well in less cluttered environments and their performance degrades with the increase in environment complexity. For small size maps, the evolutionary technique performs well but its efficiency decreases with an increase in map size.

Download Full Thesis
1,494 KB
S. No. Chapter Title of the Chapters Page Size (KB)
1 0 CONTENTS

 

vi
91 KB
2

1

INTRODUCTION

1.1 Problem Formulation
1.2 General Introduction to Evolutionary Algorithm
1.3 General Introduction to Ant Colony Optimization (ACO)
1.4 General Introduction to Particle Swarm Optimization (PSO
1.5 Thesis Overview
1.6 Summary

17
167 KB
3 2 ROUTE PLANNING APPROACHES

2.1 Route Planning
2.2 Search Strategies
2.3 Related Work in Offline Route Planning
2.4 Related Work for Online Route Planning
2.5 Summary

27
215 KB
4 3 NAVIGATION AND ROUTE PLANNING USING HYBRID FRONTIER BASED ARCHITECTURE

3.1 Introduction
3.2 Problem Solving to Planning
3.3 Basic Elements of Search Based Problem Solver
3.4 Mine Field Exploration and Planning
3.5 Frontier Based Exploration for Mine Field
3.6 Coordination Strategies for Mine Field Exploration using Agents
3.7 Hybrid Frontier Based Architecture
3.8 Using a Coordinating Agent
3.9 Experimentation
3.10 Comparative Analysis of the Heuristic
3.11 Summary

51
241 KB
5

4

ANT COLONY OPTIMIZATION BASED ROUTE PLANNING

4.1 Introduction
4.2 Biological Background for Ant Colony Optimization
4.3 Ant Colony System Development
4.4 Classification of Problems for Ant Colony Algorithm
4.5 Working of Ant Colony Systems
4.6 Simulated Ant Agent
4.7 Experimentation
4.8 Results
4.9 Optimization Methodology for Resultant Routes
4.10 Summary

69
536 KB
6

5

PARTICLE SWARM OPTIMIZATION BASED ROUTE PLANNING

5.1 Introduction
5.2 Basic Particle Swarm Optimization Method
5.3 Parameter Setting for Particle Swarm Optimization (PSO)
5.4 Comparison with Evolutionary Computation Techniques
5.5 Simulated Particle Swarm System
5.6 Simulated Niche Based Particle Swarm Optimization
5.7 Single-route Optimization
5.8 Multiple-route Optimization
5.9 Experimentation
5.10 Results
5.11 Optimization of Route Planning
5.12 Summary

91
601 KB
7

6

EVOLUTIONARY ALGORITHM BASED ROUTE PLANNING

6.1 Introduction
6.2 Background
6.3 Previous Work using Genetic Algorithm
6.4 Collaborative Evolutionary Planning Framework
6.5 Exploration Phase
6.6 Planning Phase
6.7 Cooperative Planner
6.8 Experimentation
6.9 Results and Evaluation
6.10 Summary

117
316 KB
8

7

CONCLUSION

7.1 Summary
7.2 Main Contributions of the Thesis
7.3 Future Work

135
101 KB
9

8

REFERENCES

 

140
125 KB