在城区或化工厂有毒气体突发性泄露时,有关部门需要快速对泄漏源进行定位和识别,并科学预测气体的蔓延及影响范围.由于有毒气体扩散事件一般具有隐蔽性和突发性,泄露源的空间位置、泄露强度等信息往往无法预先获知,通过传感器获得气体浓度,结合大气情况对有毒气体扩散进行反演,以推测泄露源信息的方法得到了广泛的应用.本综述围绕有毒气体泄漏的反演方法展开讨论,首先阐述了泄漏源反演研究的意义和国内外研究情况,随后着重回顾了近年来的主要研究方法和成果,并对各种方法的优劣给出了评述.%The ability to determine the source of contaminant plumes in urban or chemical plant environments is crucial for emergency-response applications. Due to the sudden and accidental nature of gas leak, the location and strength of the source are usually unknown. Therefore we need sensor network to measure the values of concentrations at the desired locations and to acquire the weather conditions. Under such circumstances, effective and efficient source characterization can help emergency agencies evacuate people from affected areas. Once the corresponding result is determined in terms of modeling parameters, forward prediction could be performed to quantify the extent of coverage to the plume. It is generally well accepted that there does not exist a single best procedure to solve dispersion source inversion problems. The source inversion methods can be categorized into direct, optimized and probabilistic approaches in general. The direct inverse approach solves atmospheric transportation equations reversely to obtain the analytic or numerical solution. The optimized inverse approach has been used to reduce the misfit between the predicted and observed data so as to obtain the best-fitted source parameters. The probabilistic method takes the measurement error and simulation error into account and obtains the probability distributions of the source parameters. This review presents the previous source characterization methods, whose advantages and disadvantages are also discussed.
展开▼
