A knowledge-based intelligent decision support framework integrating symbolic reasoning and swarm intelligence for redundancy allocation

Decision-making in complex engineering systems increasingly requires the integration of domain knowledge, operational data, and intelligent optimization techniques. Although redundancy allocation has been widely studied in reliability engineering, existing approaches rarely integrate formal knowledge representation with AI-driven optimization within a unified decision-support framework. This paper proposes a knowledge-driven intelligent decision support framework that integrates structured system knowledge, readiness inference, and AI-based optimization for redundancy allocation in complex multi-component systems. The proposed framework combines symbolic knowledge representation of system architectures with data-driven readiness inference and a hybrid AI optimization module based on binary nonlinear programming and swarm intelligence. System configurations are encoded as a knowledge base describing logical dependencies among components, enabling reasoning about readiness. The inference layer evaluates system readiness using operational data, while the optimization layer employs a binary nonlinear programming (BNLP) model as a symbolic reasoning engine, complemented by a swarm-based computational intelligence module based on Binary Artificial Bee Colony (B-ABC) for adaptive search in nonlinear decision spaces. The proposed IDSS generates redundancy recommendations that maximize readiness-related profit under budget constraints. A real-world fleet operation scenario is used to demonstrate the system’s decision-support capabilities and robustness under budget and cost variations. Experimental results based on operational fleet data demonstrate that the framework identifies budget-feasible redundancy strategies that increase modelled system readiness while providing explainable recommendations. The proposed framework is designed to be domain-independent and can support decision-making in various engineering and infrastructure systems, including energy networks, telecommunication systems, and production facilities, where redundancy planning under uncertainty is critical.