| Summary: | In high-tech informative warfare, the combat system-of-systems which become increasingly functional and complex can be abstracted as heterogeneous combat networks (HCNs). The operational capability index (OCI) formula is an important model to evaluate the performance of HCNs. To prevent the error of accepting wrong conclusions when using OCI model, the correctness and accuracy of the model should be substantiated before making critical decisions. Accordingly, this paper presents an integrated methodology of framework named operational capability index model validation based on characteristic topology analysis (OCVCT) for validating and testing the OCI model. In this framework, a concept named characteristic topology, which conforms to military rules and has the highest operational capability assessment, is put forward to replace the compact model as the validation candidate. To search for characteristic topology efficiently, we propose an improved genetic algorithm (GA) with the key-gene oriented crossover operator which considers the prior knowledge of combat networks and takes advantages of both binary and real encoding methods. A case study proves the effectiveness of OCVCT. Moreover, compared with two state-of-the-art and one classical GAs, the improved GA has superiority in convergence speed and reliability. The idea of OCVCT also has a potential application prospect for various evaluation model validations of combat networks.
|
|---|