Research On The Evaluation Model Of Air Combat Capability Of Intelligent Agents Based On OODA

With the rapid development of artificial intelligence,the application of new information technology in the battlefield has become more and more widespread.Through the current practice of some wars,it can also be proven that new advanced technology is gradually becoming the key to influencing the success or failure of wars,and wars are moving towards intelligence.At the same time,intelligent agent air combat has achieved quite good results in technological innovation,and the evaluation of intelligent agent air combat capabilities is also receiving widespread attention.This article first provides a brief introduction to relevant technical and theoretical knowledge,and then constructs an indicator system for evaluating the air combat capability of intelligent agents.Then,it determines the weight calculation algorithm between indicators,establishes a mathematical model for quantitative evaluation of qualitative indicators,and conducts experiments and validations on the evaluation model constructed in this article.Finally,a system for evaluating the air combat capability of intelligent agents is designed and developed.The main content is as follows:Firstly,the relevant theoretical knowledge of this article was introduced,and the multi indicator evaluation algorithm was studied,laying the foundation for subsequent evaluation research work.Secondly,the air combat capability of intelligent agents was decomposed based on OODA theory,and six aspects were proposed to reflect the air combat capability of intelligent agents,namely perception ability,analysis and judgment ability,decisionmaking ability,action ability,human-machine interaction ability,and learning feedback ability.The literature analysis method and Delphi method method are used to determine the bottom indicators of these six capabilities.The indicator system is divided into target layer,capability layer,primary indicator layer(fine layer),and secondary indicator layer(bottom layer),of which 29 are bottom indicators.A scientific,reasonable,and applicable indicator system for evaluating the air combat capability of intelligent agents has been obtained.Once again,through the previous comparative study of multi indicator evaluation algorithms,two multi indicator weighting algorithms were determined: based on the improved Analytic Hierarchy Process(AHP)and CRITIC weighting method,and the two weighting algorithms were explained in detail.Considering the advantages and disadvantages of the two weighting methods,the least squares method was used as the principle to combine the two weighting models,resulting in a combination weighting algorithm that can be used in this article.Then,combining the weighting algorithm with the comprehensive index evaluation model,a mathematical evaluation model corresponding to the capability level indicators and the underlying indicators was constructed,which can convert qualitative indicators into numerical form,making the evaluation results more intuitive and convenient for comparison.Finally,the model was applied and its reliability was verified by comparing the evaluation results of other models.Finally,an intelligent agent air combat capability evaluation software system was designed and developed using QT software,which improved the evaluation efficiency.