Surrogate Model-Based Rapid optimization for the Deployment of Combat System of Systems

The deployment of a combat system of systems (CSoS) is a key step to win a confrontation. However, traditional optimization methods usually consume fairly time, resulting in passivity in military operations. Therefore, a rapid optimization model is required to meet the rapidity feature of the information war. Firstly, considering connections and interactions among system entities, a network-based method is adopted to model the CSoS. Secondly, based on the operation loop (OODA) theory, and simultaneously, given that the connection quality is affected by the distance of two entities, the connection probability between two entities is calculated as the assessment of the edge quality, based on which, the connection probability of an OODA loop and the assessment result of the whole CSoS can be obtained. Thirdly, with the evaluation model constructed, the Kriging surrogate model with two steps of response face fitting and guided optimizing is used to provide a rapid optimization strategy on the deployment of CSoS. Finally, a case with two confront parts is studied. By comparing with a mature optimization method DE, the surrogate model is proved to have an approximate optimal solution with a shorter time.