The Research And Implementation Of Situation Assessment In The Target Intention Recognition

Situation assessment is a means of processing the information and forecast, it can provideaided information for makers, and improve the accuracy rate of the decision-making.Situation understanding is the second level of situation assessment function model, its mainfunction is the recognition of target intention, and provides evidence for situation forecast.This paper puts forward the intention recognition system based on the blackboard model,it achieves information storage, establishes the knowledge base and the matching betweenevidence information and knowledge base. First of all, the system divided the process oftarget intention recognition into four stages: events feature extraction, tasks reasoning, actionplan identify and intention recognition to overcome the existing problem that battlefieldinformation prediction level is not clear. Secondly, this paper put forward the knowledgeexpression patterns "rules frame + rules body" which based on the frame, and through theAccess software to establish the knowledge base of intention recognition system, which canovercome the weaknesses that the weak calculate ability of the producing type with traditionalknowledge base and the weaknesses speed with plenty rules; Finally, through the applicationof Bayesian network message transform algorithm in matching and reasoning which solve theexisting problem that information propagation direction of a single with the intentionrecognition method based on d-s evidence theory, and aim at the defect that the Updateconfidence which produced by the traditional Bayesian network message transform algorithmhave credibility gap with reality, this paper provides a weighted Bayesian network messagetransform algorithm.This paper simulated the intention recognition system, through the operation interfaceinput the Evidence parameters, matching the evidence and database according to the weightedBayesian network message transform algorithm, the system can realize the information ofevidence spread to different direction, and improve the accuracy of the updated confidence.