Research reactor core loading optimization method and system based on hierarchical cross-evolution
By decomposing the core loading problem into fuel and non-fuel component optimization layers through a hierarchical cross-evolutionary approach, and using an improved evolutionary algorithm for collaborative optimization, the inefficiency of asymmetric core design under HFETR multi-mission irradiation scenarios is solved, and efficient and reliable core loading scheme generation is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NUCLEAR POWER INSTITUTE OF CHINA
- Filing Date
- 2026-04-09
- Publication Date
- 2026-06-26
AI Technical Summary
Existing optimization methods are difficult to coordinately optimize multiple objectives in the asymmetric core design of the High Flux Engineering Test Reactor (HFETR) under multi-mission irradiation scenarios within a limited computation time, and the search efficiency is low, which cannot meet the core design efficiency requirements of high-frequency refueling scenarios.
A research reactor core loading optimization method based on hierarchical cross-evolution is adopted. By constructing a hierarchical optimization model, the core loading problem is decomposed into a fuel assembly optimization layer and a non-fuel assembly optimization layer. An improved evolutionary algorithm is used for collaborative optimization to generate the arrangement scheme of fuel and non-fuel assemblies. Combined with multi-objective fitness evaluation and iterative optimization, a core loading scheme that meets the preset conditions is generated.
It effectively reduces the solution dimension of asymmetric core optimization, improves the convergence speed and computational efficiency of the optimization process, generates a more reliable core loading scheme in terms of physical feasibility and engineering practicality, and solves the optimization problem of asymmetric core layout in multi-task irradiation scenarios.
Smart Images

Figure CN122287124A_ABST