17 September 2018 in IEEE Access
System of systems (SoS) architecting is a process to bring systems together as a network structure for specific goals with unique capabilities and the cumulative capabilities achieved from their interoperability. How the capability is delivered across the network is crucial for capability-based planning mission. To meet diverse capability needs by architecting different types of systems into a network, we formulate and analyze an SoS architecting problem with network theory as a multi-objective optimization model. More specifically, we construct a three-step capability delivery model from a local and globe perspective to describe the positive and negative consequences caused by system interactions. A three-level response model with a feasibility concept is proposed as the conversion of the capability value to meet different scale combat needs and balance the deployment of new and old weapons. And then, certain intelligent optimization algorithms incorporated with the proposed model is designed to approximate Pareto fronts of three response levels, respectively. Results from a set of numerical studies demonstrate that an efficient structure with fewer nodes performs better than a full set structure, and the deployment of a system should be a comprehensive consideration of its overall structural influence rather than single system performance. In addition, four SoS-level attributes are also defined and calculated as supplements to make the value of solution architecture more comprehensive.
4 July 2018 in IEEE Access
As system dynamic behaviors are difficult to capture and analyze in model-based systems engineering, executable architectures, and simulation methods are widely studied to understand the behaviors that result from interactions amongst system components. However, most current studies often strive for analyzing independently specific aspects of architecture rather than capturing simultaneously the structural, behavioral, and performance related features, leading to failures of understanding the whole system from a global perspective. Accordingly, an interactive model-driven simulation that includes meta-model mapping and collaborative engine capabilities specific to DoDAF, Rhapsody, and STK simulations is first proposed to integrate the full advantages of different analytical tools. Then, a synergic engine is developed to provide synchronized control between time-driven and event-driven execution through the dynamic analysis process. Two executable techniques communicate with each other throughout the entire execution cycle to provide complementary necessary process information. The events occurred in the STK simulation scenario are used as the trigger of the executable state machine model that describes only the logical procedure of system behaviors without timeand space-related constraints. Last, an illustrative example of the Russian-Turkish Plane Incident is carried out to demonstrate the feasibility of the foregoing approach.
1 August 2025 in IEEE Transactions on Evolutionary Computation
Dynamic multi-objective optimization problems (DMOPs) are widely applied in scheduling, resource allocation, and complex task assignments, requiring algorithms to respond quickly to environmental changes. Existing prediction methods do not id=R2fully id=R3utilizemake use of historical information and rely on simplistic linear or nonlinear assumptions, which fail to effectively handle complex environmental dynamics in practice, especially leading to increased prediction errors over long periods of change. This paper id=R3proposesintroduces a novel prediction framework based on incremental LSTM and inverse models (IncLSTM-IM). For nonlinear changes, the IncLSTM is employed to learn the evolutionary patterns of the pareto optimal set (POS), fully exploiting the capacity for handling sequential data and learning long-term dependencies to guide offspring solutions id=R3toward thebeing closer to next POS. For linear changes, an inverse model is introduced to predict using local historical information. Furthermore, a population selection mechanism based on clustering is implemented to combine diverse offspring populations. IncLSTM-IM overcomes the limitations of existing prediction methods by effectively capturing complex environmental changes and enhancing population diversity through cluster-based id=R3population selection. Experimental results, compared with existing algorithms, demonstrate the robust performance and adaptability of the proposed id=R3approachalgorithm.
1 January 2026 in IEEE Transactions on Evolutionary Computation