确定建筑物内空气传播物污染源的反问题研究

项目名称： 确定建筑物内空气传播物污染源的反问题研究

项目编号： No.50808028

项目类型： 青年科学基金项目

立项/批准年度： 2009

项目学科： 武器工业

项目作者： 张腾飞

作者单位： 大连理工大学

项目金额： 21万元

中文摘要： 当封闭空间内发生有害物的突发释放及空气传播时，如何准确辨识出污染源所在的位置与释放强度，对于采取措施来有效清除污染源并给人员提供应急防护显得尤为重要。本课题首先开发了一种具有污染源位置识别功能的空气样品智能采集仪，采集仪采用"二叉树"形式的管路拓扑结构，当封闭空间内发生有害物释放时，可对污染源所在的大致位置进行快速定位。然后开发了一组CFD（计算流体力学）反问题数学模型，运用传感器提供的有效信息，对污染源具体位置及逐时释放信息进行精确辨识。由于反问题属于病态问题，本研究提出可用于复原之前发生的颗粒物传播过程的准可逆反问题模型，模型中考虑颗粒的重力沉降作用，将时间流逝逆向，把目前只能用于辨识室内气体污染物传播源的反问题模型推广至颗粒物。此外，还开发了基于流场取反而复原颗粒物传播轨迹的逆拉格朗日反问题模型，并提出基于颗粒物最大浓度边线布置包围颗粒而界定颗粒物传播源的逆拉格朗日包围线法。研究还成功引入由源强来确定浓度的传递关系矩阵，在业内首次将Tikhonov规整化理论与传递矩阵法相结合，并将其运用于室内污染源释放强度的辨识，提出了可高效准确地辨识室内气体污染源非稳态逐时释放过程的逆向模型。

中文关键词： 污染源辨识；反问题；Tikhonov 规整化；传递矩阵；空气环境安全

英文摘要： In case an accidental release of harmful agents occurs, it is very crucial to know where and how the agents are released, so that it can guide to eliminate the agent sources and provide emergency protection for occupants. This project developed an intelligent multi-point sampler that can roughly tell the location of an indoor contaminant source released in a building with many compartments. The tube network within the sampler is manufactured into the structure like a binary tree, so the contaminant source can be immediately determined based on the binary search theory. Then a series of inverse computational fluid dynamics (CFD) models were proposed to accurately pinpoint a contaminant source in a room and determine its unsteady release pattern and strength, once the alarming information is provided by some detection sensors. Since inverse problems are ill-posed, a quasi-reversibility model was developed to recovering the transport history of particulate polluants by reversing the time marching direction and including the particle deposition effect. In addition, the Lagrangian reversibility model was also proposed to backward trace the particle trajectories by deploying the track particles on the zone boundary of the largest concentration contour at a time. The particulate source can then be circumscribed by the track particles after reversing the flow field. To quantitatively determine the unsteady release rate of a contaminant source, the Tikhonov regularization technique was adopted after casting the concentration from a polluant source into the product of a transfer matrix with the release strength vector. After that, the Tikhonov regularization can be highly efficient to accurately quantify the unsteady release pattern and strength of a contaminant source.

英文关键词： Contaminant source identification; Inverse modeling; Tikhonov regularization; Transfer matrix; Air environment security