| 引用本文: | 张潞,叶彧维,艾青,刘梦,帅永.多热源系统全域温度场的实时重构策略[J].哈尔滨工业大学学报,2025,57(2):25.DOI:10.11918/202409011 |
| ZHANG Lu,YE Yuwei,AI Qing,LIU Meng,SHUAI Yong.Real-time reconstruction strategy for full-domain temperature field of multi-heat source systems[J].Journal of Harbin Institute of Technology,2025,57(2):25.DOI:10.11918/202409011 |
|
| 摘要: |
| 在多热源系统的动态热管理中,全域热场分析与实时评估是核心要素,而传统的离散测量和重构技术难以实时获取温度场动态变化。为此,提出一种基于离散测点的全域温度场重构策略,通过引入奇异值分解提取温度场的特征基函数,并引入灰狼优化算法对重构离散测点的布局进行优化,通过与离散测点值结合,建立全域温度场的关联系数矩阵。采用数值实验,基于4种类型热源系统验证了该方法的可靠性。结果表明:测点布局优化后,4种温度模型的理论重构误差水平显著降低了至少3个数量级。采用该方法对某多芯片PCB的温度场进行重构,与数值实验结果对比显示温度场平均误差为0.12 ℃,均方根百分比误差低于1%,验证了该策略在实际应用中的可靠性,为电子设备热控热分析提供了参考依据。 |
| 关键词: 模型降阶 最小二乘 温度场重构 布局优化 |
| DOI：10.11918/202409011 |
| 分类号:TK311 |
| 文献标识码:A |
| 基金项目:国家自然科学基金(52476066);电磁信息控制与效应全国重点实验室开放基金(20230201) |
|
| Real-time reconstruction strategy for full-domain temperature field of multi-heat source systems |
|
ZHANG Lu1,YE Yuwei2,AI Qing2,LIU Meng2,SHUAI Yong2
|
|
(1.School of Computer Science, Shenyang Aerospace University, Shenyang 110136, China; 2.School of Energy Science and Engineering, Harbin Institute of Technology, Harbin 150001, China)
|
| Abstract: |
| In dynamic thermal management of multi-heat source systems, full-domain thermal analysis and real-time assessment are core elements. However, conventional discrete measurements and reconstruction techniques struggle to capture dynamic evolution of the temperature field in real time. Therefore, a full-domain temperature field reconstruction strategy based on discrete measurements is described. The eigenbasis functions of the temperature field are extracted by singular value decomposition (SVD), and the Gray Wolf Optimization (GWO) is introduced to optimize the layout of discrete sensors. A coefficient matrix for the full-domain temperature field reconstruction is computed by combining the eigenbasis with the discrete measurements. The reliability of this approach is validated based on numerical reconstruction experiments of four heat source systems. The results indicate that after sensor placement optimization, the level of theoretical reconstruction error for the four heat source systems is significantly reduced by at least three orders of magnitude. Moreover, in the temperature field reconstruction of a typical PCB multi-heat source system, comparison with numerical experiment results shows an average error of 0.12 ℃ and a root-mean-square percentage error below 1%. The reliability of the strategy in practical applications is validated, providing a reference for thermal analysis and control of electronic equipment. |
| Key words: order-reduced model least square temperature field reconstruction placement optimization |
