声明
摘要
ABSTRACT
Contents
Notation
Chapter 1 Introduction
1.1 Research Content and Significance
1.2 Related Work on Solving the TSP
1.2.1 Tour Construction Algorithms
1.2.2 Local Search Approaches
1.2.3 Evolutionary Computation Method
1.2.4 Hybrid Algorithms
1.2.5 Summary of Algorithms for the TSP
1.3 Experimental Framework
1.3.1 TSP Suite Framework
1.3.2 TSPLIB Benchmark
1.4 Contributions
1.5 Structure of this Thesis
Chapter 2 Modified Ejection Chain Method forSolving the TSP
2.1 Efficient Implementation of the Ejection Chain Method
2.1.1 Data Structure
2.1.2 Search Strategy
2.1.3 Experimental Setups
2.1.4 Experimental Results
2.2 Fine Tuning Maximum Level and Size of Root Node Set
2.2.1 Experimental Setups
2.2.2 Experimental Results
2.3 Comparing ECM algorithms with LK heuristics
2.3.1 Experimental Setups
2.3.2 Experimental Results
2.4 Conclusions
Chapter 3 Hybrid Algorithms for Solving the TSP
3.1 Hybrid Two Local Searches with Crossover Operator Algorithms
3.1.1 Comparison of LS approaches and Hybrid LS-LS algorithms
3.1.2 Combining Different LS approaches with Crossover Operator
3.1.3 Experimental Setups
3.1.4 Experimental Results
3.1.5 Conclusions
3.2 Hybrid Global Search-Local Search Algorithms
3.2.1 Comparative Study of LS Approaches and EC Methods
3.2.2 Hybrids with Evolutionary Algorithms
3.2.4 Experimental Setups
3.2.5 Experimental Results
3.2.6 Conclusions
Chapter 4 Conclusions
4.1 Research Summary
4.2 Future Tasks
Bibliography
Acknowledgements