基于遗传-模式搜索算法的微尺度管控区域大气污染物PM2.5溯源

Back-calculation of outdoor PM2.5 pollutant source around microscale controlled area by genetic-pattern search algorithm

针对微尺度管控区域可能发生的大气污染提出有效的靶向诊断方法-结合高斯烟羽模型和遗传-模式搜索算法的大气污染物分布式溯源方法.将污染源反算模型得到的污染物理论质量浓度与传感器网络观测值的数据对应关系作为目标函数,使用模式搜索算法嵌入遗传算法加快反算模型的搜索过程,反算得到污染源强度和位置.依托杭州市亚运板球场馆大气感知器网络进行实验验证,监测2021年10月PM2.5质量浓度、气象数据,对所提出的混合式大气污染溯源方法进行实验验证.实验结果表明:改进遗传-模式搜索算法对于多维变量的搜索效果较好,能快速精准地反算污染源的位置和强度,可以为微尺度管控区域突发性气体污染防治提供应急决策参考.

An effective targeted diagnosis method,distributed traceability method for atmospheric pollutants combining Gaussian plume model and genetic-pattern search algorithm,was proposed,aiming at air pollutants that may occur in the micro-scale control area.The corresponding relationship between the calculated pollutant concentration obtained from pollution backcalculation model and the observation value of the monitoring sensor was used as the objective function.Pattern search algorithm was embedded in the genetic algorithm to speed up the search process of the inverse calculation model,then to inversely calculate the intensity and location of the pollution source.A validation experiment was conducted by monitoring the PM2.5 mass concentration,meteorology and other data based on the atmospheric sensor data of Hangzhou Asian Games cricket stadium in October 2021.Results showed,compared with other methods,the effect of the improved genetic-pattern search algorithm for multidimensional variables was better,and the location and intensity of pollution sources could be calculated more quickly and accurately.This research can provide suggested solution for environmental emergencies of air pollution in micro-scale control regions.