Game Theory Optimization via Diverse Genetic Crossover Intelligence  

作　　者：David Webb Eric Sandgren David Webb;Eric Sandgren(AdvanSix Inc., Parsippany, New Jersey, USA;Systems Engineering Department, Donaghey College of Engineering & Information Technology, University of Arkansas at Little Rock, Little Rock, Arkansas, USA)

机构地区：[1]AdvanSix Inc., Parsippany, New Jersey, USA [2]Systems Engineering Department, Donaghey College of Engineering & Information Technology, University of Arkansas at Little Rock, Little Rock, Arkansas, USA

出　　处：《Journal of Applied Mathematics and Physics》2024年第10期3315-3327,共13页应用数学与应用物理（英文）

摘　　要：Game theory is explored via a maze application where combinatorial optimization occurs with the objective of traversing through a defined maze with an aim to enhance decision support and locate the optimal travel sequence while minimizing computation time. This combinatorial optimization approach is initially demonstrated by utilizing a traditional genetic algorithm (GA), followed by the incorporation of artificial intelligence utilizing embedded rules based on domain-specific knowledge. The aim of this initiative is to compare the results of the traditional and rule-based optimization approaches with results acquired through an intelligent crossover methodology. The intelligent crossover approach encompasses a two-dimensional GA encoding where a second chromosome string is introduced within the GA, offering a sophisticated means for chromosome crossover amongst selected parents. Additionally, parent selection intelligence is incorporated where the best-traversed paths or population members are retained and utilized as potential parents to mate with parents selected within a traditional GA methodology. A further enhancement regarding the utilization of saved optimal population members as potential parents is mathematically explored within this literature.Game theory is explored via a maze application where combinatorial optimization occurs with the objective of traversing through a defined maze with an aim to enhance decision support and locate the optimal travel sequence while minimizing computation time. This combinatorial optimization approach is initially demonstrated by utilizing a traditional genetic algorithm (GA), followed by the incorporation of artificial intelligence utilizing embedded rules based on domain-specific knowledge. The aim of this initiative is to compare the results of the traditional and rule-based optimization approaches with results acquired through an intelligent crossover methodology. The intelligent crossover approach encompasses a two-dimensional GA encoding where a second chromosome string is introduced within the GA, offering a sophisticated means for chromosome crossover amongst selected parents. Additionally, parent selection intelligence is incorporated where the best-traversed paths or population members are retained and utilized as potential parents to mate with parents selected within a traditional GA methodology. A further enhancement regarding the utilization of saved optimal population members as potential parents is mathematically explored within this literature.

关 键 词：Crossover Intelligence Game Theory Maze Navigation Genetic Optimization 

分 类 号：TP1[自动化与计算机技术—控制理论与控制工程]