Research On Combat Effectiveness Evaluation Method Of UAV Formation Based On Bayesian Network

In the context of modern warfare,it is difficult for a single unmanned aerial vehicle(UAV)to complete complex tasks.UAV formation operations can not only interfere or disperse enemy anti-aircraft fire,but also cooperate with other UAVs to conduct operations with high flexibility.As a result,UAV formation operations have been widely adopted.In order to select a reasonable and effective combat plan and maximize the effectiveness of combat,the evaluation of the combat effectiveness of UAV formations has received widespread attention.Accurate assessment results can provide a basis for the selection of tactical solutions and reduce unnecessary economic losses.Inaccurate evaluation results may lead to the failure of combat missions.UAV formation combat effectiveness evaluation is a hot issue in the military field.At present,domestic research in this area started relatively late,and most of the research is directed to a single UAV.There are two main difficulties in the evaluation process.(1)The combat effectiveness of UAV formation is affected by factors such as battlefield environment,personnel command level,and aircraft performance,resulting in many evaluation indicators and redundancy.At the same time,the experts' comments on the importance of indicators are ambiguous,which may make the calculated indicator weights inaccurate and affect the evaluation results.(2)The battlefield environment changes rapidly and there are many uncertain factors in the process of mission execution.It is difficult to construct the evaluation model.The evaluation model constructed must have strong reasoning ability.The main work of this paper is as follows:(1)For the first difficulty,the triangular fuzzy analytic hierarchy process is used to optimize the indicators system.In this process,through the study of the method of establishing the index system and the analysis of the factors affecting the operational effectiveness,the operational effectiveness evaluation indicators system is established.A fuzzy evaluation matrix is created according to the index system,and the index system is optimized through steps such as consistency check of the matrix,calculation of the initial weight of the index,and defuzzification the initial weight of the index.(2)For the second difficulty,the model of UAV formation combat effectiveness evaluation is built by Bayesian network.The construction of the model mainly includes two steps of structure learning and parameter learning.Above all,k2 algorithm is used for structure learning,and the structure learned by this algorithm is relatively accurate.Then the maximum likelihood estimation method is used for parameter learning,whichgreatly reduces the influence of the subjectivity of the experts on the conditional probability table.The evaluation model based on Bayesian network has strong inference ability,which can be well applied to the evaluation of UAV formation combat effectiveness in uncertain environment.(3)After establishing the combat effectiveness evaluation indicators system of the UAV formation and constructing the evaluation model,the inference method of the Bayesian network is used to reason about the combat effectiveness of the individual formations and the combat effectiveness of the entire formations to verify the rationality of the model.The experimental results show that the proposed method based on Bayesian network has good applicability.The method has characteristics of strong inference ability and multiple inference methods.This method has the characteristics of strong reasoning ability and multiple reasoning methods.It reasonably evaluate the formation combat plan and tactical decision-making.At the same time,it provide the decision-maker with a clear assessment process to better guide the UAV formation combat.