The Fighter Aircraft LCS: A Case of Different LCS Goals and Techniques

作者: Robert E. Smith , B. A. Dike , B. Ravichandran , A. El-Fallah , R. K. Mehra

DOI: 10.1007/3-540-45027-0_15

关键词: Control (management)AeronauticsComputer scienceOperations researchDomain (software engineering)

摘要: A system employed by the authors to acquire novel fighter aircraft manoeuvres from combat simulation is more akin traditional LCS model than recent systems. Given difficulties often experienced in research on simple problems, one must ask how a relatively primitive has had consistent success complex domain of manoeuvring. This paper presents LCS, greater detail previous publications. Positive results are discussed. The then focuses primary reasons avoided LCS. believe system's three origins: differences credit assignment, action encoding, and (possibly most importantly) difference goals. In system, goal been simply discovery innovative, tactics, rather online control. concludes discussing salient features learning those may be profitably combined with other developments.

参考文章(20)
Damien Ernst, Arthur Louette, Introduction to Reinforcement Learning MIT Press. ,(1998)
Stewart W. Wilson, State of XCS Classifier System Research Lecture Notes in Computer Science. pp. 63- 82 ,(2000) , 10.1007/3-540-45027-0_3
S. Nolfi, D. Floreano, God Save the Red Queen! Competition in Co-Evolutionary Robotics 2nd Conference on Genetic Programming. ,(1997)
Stewart W. Wilson, Classifier fitness based on accuracy Evolutionary Computation. ,vol. 3, pp. 149- 175 ,(1995) , 10.1162/EVCO.1995.3.2.149
John J. Grefenstette, Credit assignment in rule discovery systems based on genetic algorithms Machine Learning. ,vol. 3, pp. 225- 245 ,(1988) , 10.1007/BF00113898
Stewart W. Wilson, Zcs: A zeroth level classifier system Evolutionary Computation. ,vol. 2, pp. 1- 18 ,(1994) , 10.1162/EVCO.1994.2.1.1
Howard Raiffa, R. Duncan Luce, T. Teichmann, Games and decisions ,(1957)