Font Size: a A A

Recognition And Response Of The Hostile Plan

Posted on:2008-07-04Degree:MasterType:Thesis
Country:ChinaCandidate:T J ShaoFull Text:PDF
GTID:2178360215979737Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
As a research field of science artificial intelligence has a history of more than half a century, and it has affected the computer science in every aspects. AI planning and plan recognition as branches of artificial intelligence have achieved great research results, meanwhile with good performance in many application fields such as aviation and industry manufacture.Most planning research is focus on the relationship of cooperation and independent between agents, but fewer attentions on hostile relationship. However with the using of AI planning in many real fields that with hostile actions, such as network security, military antagonizing. So the research on hostile plans is imperative under the situation, which would has great research value and widely application.Some research works have been done in this paper on the recognizing and countering of hostile plans based on the AI planning and planning recognition, brings forward a counter planner, and makes some new definitions on hostile environment, hostile plan, counter plan, hostile agent, and counter planner and so on. The classifying of hostile plans is also discussed in this paper.The counter planner in this paper is used to detect, recognize and make counter plans to the hostile plans, and its logical functions are recognizing hostile plans and making counter plans. The main modules are hostile actions catcher, hostile plan library, hostile attack detector, fuzzy controller and counter plan producer. All of these modules work together to detect and recognize the hostile plans, and carry out relevant counter plans to maintain the system's normal performance.This paper firstly brings the similarity and fuzzy control theory into the counter planner. The similarity is used to describe the similitude of two random variables, and in this paper the similarity is used to explain the similitude of two action sequences in order to judge whether the detected actions are hostile plans; the fuzzy control method is based on the fuzzy mathematics, is a kind of control's strategy used the fuzzy reasoning, and could carry out the control of the non-linear, complex system. The hostile plans are fuzzy, so the using of fuzzy control method could make the counter planner with good performance.The counter planner in this paper is a universal model; it could be applied in many environments as well as with well activity and expansibility.
Keywords/Search Tags:AI Planning, Plan Recognition, Hostile Plan, Counter Plan, Counter Planner
PDF Full Text Request
Related items