个基于代理模型与多目标优化算法的氦氙-空气混合式PCHE优化

Optimization of helium-xenon and air hybrid PCHE based on surrogate model and multi-objective optimization algorithm

  • 摘要: 混合式印刷电路板式换热器(PCHE)适用于小型移动式反应堆二回路预冷器中两侧工质体积热容差异较大的换热场景。针对压降、热侧出口温度和总质量流量约束下的体积与经济性协同优化问题,提出了一种基于代理模型与多目标优化算法的优化方法。该方法采用优化拉丁超立方采样、计算流体力学、高斯过程回归构建代理模型,针对该模型开展参数敏感性分析;在此基础上,利用非支配排序遗传-III算法获得帕累托前沿,再采用逼近理想解排序法进行综合决策。结果表明,支出目标中热侧通道雷诺数占主导地位,体积目标中热侧通道直径占主导地位,仅在热侧通道直径的部分设计范围内,体积与支出目标不存在显著冲突;冷侧压损约束对帕累托前沿形态具有显著影响。与初始方案相比,最终方案的体积增加23.8%,支出降低60.3%,表明该方法能够在有限体积增量下显著改善换热器经济性。

     

    Abstract:
    Background Small mobile reactors require compact and reliable precoolers in their secondary loops. A hybrid printed circuit heat exchanger (PCHE) is a promising candidate when the volumetric heat capacities of the hot and cold working fluids differ substantially, but its design must balance volume, economic performance, and hydraulic and thermal constraints.
    Purpose This study aims to develop a surrogate-model-based multi-objective optimization framework for a hybrid PCHE precooler to minimize volume and expenditure under constraints on pressure drop, hot-side outlet temperature, and total mass flow rate, and to clarify parameter sensitivity and constraint effects for engineering selection.
    Methods Optimized Latin Hypercube Sampling was used to generate design samples, and Computational Fluid Dynamics simulations were performed to evaluate thermal-hydraulic and economic responses. Gaussian Process Regression models were then constructed as surrogates, and parameter sensitivity analysis was conducted. On this basis, the Non-dominated Sorting Genetic Algorithm-III (NSGA-III) was employed to generate Pareto-optimal designs, and the technique for order preference by Similarity to Ideal Solution (TOPSIS) was applied for comprehensive decision-making. The effects of key geometric and flow parameters, particularly the hot-side channel Reynolds number and diameter, were analyzed, and the influence of the cold-side pressure-drop constraint on the Pareto front was examined.
    Results Sensitivity analysis showed that the hot-side channel Reynolds number dominates the expenditure objective, whereas the hot-side channel diameter dominates the volume objective. The two objectives are non-conflicting only within part of the design range of the hot-side diameter. The cold-side pressure-drop constraint significantly affects the Pareto front and alters its trend. Compared with the initial design, the final selected design increases volume by 23.8% but reduces expenditure by 60.3%, indicating that volume is traded for economic benefit.
    Conclusions The proposed surrogate-based multi-objective optimization method can effectively support the design of hybrid PCHE precoolers under multiple thermal-hydraulic and economic constraints. The results reveal the dominant parameters and constraint-dependent trade-offs and provide a rational basis for selecting compact, cost-effective precooler designs for small mobile reactors. The approach is expected to be useful for similar compact heat exchanger optimization problems in which strong property mismatch and coupled constraints exist.

     

/

返回文章
返回