Fuzzy modeling of offensive maneuver in an evader-pursuer task

Abstract

In this paper we propose a new guidance law based on fuzzy logic that can be successfully used for modeling and generating complicated offensive maneuver in an evader-pursuer task encounter between two highly responsive simplified dynamic systems called as object. Based on human expert's decision-making process, an AI based method is proposed to model the maneuvering. Fuzzy "if ... then ..." rules are used to represent the pursuer preferences in guiding his/her system. The rules are directly obtained from expert's knowledge. Each rule relates the desired moving directions of the pursuer to the task parameters. The control parameters of the object are computed through a mean square error scheme. A large amount of simulations are used to ensure the satisfactory performance of the model. The results show the similarity of the model output to human like maneuvers.