发表机构
Korea Advanced Institute of Science and Technology (KAIST); Narnia Labs（韩国科学技术院; Narnia实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出风险感知生成式修复框架，通过估计编辑结果分布并按风险水平排序来选择电动汽车电池冷却通道的局部编辑位置，将随机生成编辑转化为可控设计决策。
AI 中文摘要
电动汽车电池组的冷却通道布局必须实现温度均匀性和低压降，同时保持单一连续通道。在设计后期，局部修改是在保留既有全局特征的同时提升性能的实用方法。基于扩散的修复技术支持此类局部编辑，但其随机性导致即使对同一区域进行编辑也会产生不同结果。这引出了一个基本问题：当每次修改的结果具有随机性时，应如何选择编辑位置？我们提出了一种风险感知的生成式编辑框架，通过考虑这种变异性来选择编辑位置。该方法并非根据单一期望改进值对每个候选位置进行评分，而是利用经CFD训练的代理模型评估的离线编辑结果来估计结果分布，并根据选定的风险水平对位置进行排序。因此，一个训练好的模型在推理时可支持不同的编辑偏好，既可强调更高的期望改进，也可强调更高的一致性，而无需重新训练。在七个掩码配置的留出配对研究中，该策略与随机编辑进行了对比评估。在大多数配置中，它相较于随机编辑改善了冷却通道目标，且该优势在编辑设计的独立CFD验证下依然成立。改变风险水平揭示了平均改进与逐次运行一致性之间的一致权衡，而学习到的分布有助于对位置进行排序，但不应将其解读为校准的概率分布。这些结果共同表明，通过风险感知的位置选择，随机生成式编辑可以从变异性来源转化为可控的设计决策：有效的编辑不仅取决于设计如何被修改，还取决于修改位置以及可接受的结果变异性程度。
英文摘要
Cooling-channel layouts for electric-vehicle battery packs must deliver temperature uniformity and low pressure drop while maintaining a single continuous channel. In late-stage design, local modification is a practical way to improve performance while retaining established global features. Diffusion-based inpainting supports such local edits, but its stochastic nature produces different outcomes even for the same region. This raises a fundamental question: how should edit locations be selected when each modification's outcome is stochastic? We propose a risk-aware generative editing framework that selects edit locations by accounting for this variability. Rather than scoring each candidate location by a single expected improvement, the method estimates a distribution of outcomes from offline edit results evaluated with a CFD-trained surrogate, and ranks locations by a chosen risk level. A single trained model therefore supports different editing preferences at inference time, emphasizing either higher expected improvement or greater consistency, without retraining. The policy is evaluated against random editing in a held-out paired study across seven mask configurations. It improves the cooling-channel objective over random editing in most configurations, and the advantage holds under independent CFD verification of the edited designs. Varying the risk level reveals a consistent trade-off between mean improvement and run-to-run consistency, while the learned distribution is useful for ranking locations but should not be read as a calibrated probability distribution. Together, these results show that stochastic generative editing can be converted from a source of variability into a controllable design decision through risk-aware location selection: effective editing depends not only on how a design is modified, but also on where and how much outcome variability is acceptable.
Comments35 pages, 11 figures, 5 tables
