System-of-Systems Effectiveness Evaluation Based On Improved FDNA Method

System-of-Systems(So S)effectiveness evaluation plays a critical role in the requirement analysis,concept design,and architecture design for So S development.Thus,how to evaluate the effectiveness of an So S scientifically and accurately is an important issue in the So S engineering and national defense domains.Quantitative analysis of complex interdependencies among component systems is a key challenge for So S effectiveness evaluation.Functional dependency network analysis(FDNA)is a new method for So S effectiveness evaluation that can capture the interdependency relationship and its cascading effect within an So S.Due to its capability of describing the So S interdependency and some of the emergent behaviors by means of dependent parameters and piecewise linear function,the FDNA method gradually receives attention in the So S domain.However,current FDNA method suffers from quite a few shortfalls that could hinder the usefulness of the method.For instance,the effectiveness function is inadequate and the configurations of dependency relationship are incomplete.Meanwhile,there exist problems of unclear model specifications and unclear rules of determining the value of dependency parameters.Therefore,this thesis carries out two aspects of research to enhance the accuracy and usefulness of the FDNA method.The first section focuses on improving the mathematical models of the current FDNA method from multiple perspectives.After developing rules and specifications for building FDNA models of combat So Ss,this thesis brings in self-effectiveness and interdependency health index to improve the ability of analyzing node and dependency failures.Then a more sophisticated function of criticality of dependency is added to the FDNA formulation to capture the situation with complete failure of same-function nodes.This thesis further proposes effectiveness aggregation method based on the redundant configuration of backup systems in the So S,so as to avoid the underestimation of the efficiency caused by the criticality constraint.Regarding the determination of the values of dependency parameters,the second portion of this thesis presents a method to determine the dependency parameters based on multi-attribute group decision-making.The proposed method examines the definition of dependency parameters and identifies the attributes based on the underlying implications of the definition.The attribute weights are then determined through a combination of subjective and objective evaluation methods,in which the ordered weighted aggregation operator is used to gather the decision data from many experts so as to reduce the individual bias.Based on the above studies,this thesis eventually develops an improved FDNA based effectiveness evaluation framework for So S.An application to an air and missile defense(AMD)So S goes through the process of FDNA model construction,dependency parameter determination and effectiveness calculation.As a result,the improved FDNA method demonstrates its feasibility and effectiveness by comparing itself with the original FDNA method.