高温堆燃料贮罐结构-性能-成本一体化多目标优化设计

Structure-Performance-Cost Integration Multi-Objective Optimization Design for HTR Fuel Storage Canister

燃料贮罐是高温堆新燃料供应系统关键设备。为探索最佳设计方案,提出燃料贮罐结构-性能-成本一体化多目标优化设计方法:选取燃料贮罐结构板厚作为设计变量,采用拉丁超立方采样(LHS)生成均匀采样点,通过数值计算获取跌落响应,通过混合径向基函数神经网络(RBFNN)-前馈神经网络(FFNN)构造代理模型;以最大塑性变形最小、成本最低、质量最小作为优化设计目标,同时约束球床作用下的径向位移膨胀,利用强度Pareto进化算法(SPEA-Ⅱ)求解优化问题。结果表明:燃料贮罐安全性明显提高,最大塑性变形可降低20.17%;经济性与轻量化效果较好,单罐成本可降低2128元,质量可降低12.54%。本文一体化优化方法能够为燃料贮罐设计提供参考。

Fuel storage canister is a key equipment in the fuel supply system for high temperature reactor(HTR).In order to explore the optimal design scheme,a structure-performancecost integration multi-objective optimization design method for fuel storage canister is proposed as follows:select the structural plate thickness of fuel storage canister as the design variable,use the Latin hypercube sampling(LHS)method to generate uniform sampling points,obtain the drop response through numerical calculation,and employ the hybrid radial basis function neural network(RBFNN)-feedforward neural network(FFNN)to construct a surrogate model.With the optimization design objectives of minimizing the maximum plastic deformation,the cost and the mass,constrain the radial displacement expansion under the action of pebble bed,and solve the optimization problem by using the strength Pareto evolutionary algorithm(SPEA-Ⅱ).The results show that the safety of the fuel storage canister is significantly improved,and the maximum plastic deformation can be reduced by 20.17%;it has good economical efficiency and lightweight effect,the cost of a single canister can be reduced by 2,128 yuan,and the mass can be reduced by 12.54%.The integration optimization method proposed in this paper can provide reference for the fuel storage canister design.