摘要
体育馆室内空气污染物种类繁多,难以准确提取关键因子,且空气污染物浓度随时间变化具有非线性、动态性等特点,导致无法充分捕捉数据关系并动态调整对不同时间步数据的关注程度,使得体育馆室内空气污染物分布特征提取结果存在偏差。因此,研究体育馆室内空气污染物分布特征提取方法仿真。分析体育馆室内空气质量影响因素,并根据灰色关联分析确定污染物因子。为适应复杂、非线性的污染物浓度变化,输入各污染物因子数据,利用长短期记忆网络(Long Short-Term Memory networks,LSTM)预测空气污染物浓度,以充分捕捉数据长期依赖关系。并引入注意力机制,动态调整对不同时间步数据的关注程度,获得精准的体育馆室内空气污染物浓度分布,完成特征提取。结果表明,所提方法预测结果与实际结果一致,具有较高的预测准确性,且可有效提取体育馆室内空气污染物分布特征。
There are many kinds of indoor air pollutants in gymnasiums,and it is difficult to accurately extract key factors.In addition,the concentration of air pollutants changes with time with nonlinear and dynamic characteristics,which makes it impossible to fully capture the data relationship and dynamically adjust the attention level of data at different time steps,resulting in deviations in the extraction results of indoor air pollutant distribution characteristics in gymnasiums.Therefore,the simulation of the indoor air pollutant distribution feature extraction method for gymnasiums is studied.The influencing factors of indoor air quality in gymnasiums are analyzed,and the pollutant factors are determined based on grey correlation analysis.In order to adapt to the complex and nonlinear changes in pollutant concentrations,the data of each pollutant factor is input,and the long short-term memory network(LSTM)is used to predict the concentration of air pollutants to fully capture the long-term dependency of data.The attention mechanism is introduced to dynamically adjust the attention level of data at different time steps to obtain accurate indoor air pollutant concentration distribution in gymnasiums and complete feature extraction.The results show that the prediction results of the proposed method are consistent with the actual results,with high prediction accuracy,and can effectively extract the distribution characteristics of indoor air pollutants in gymnasiums.
作者
李晓兰
张红玲
LI Xiao-lan;ZHANG Hong-ling(Yan'an University,Yan'an Shaanxi 716000,China)
出处
《计算机仿真》
2025年第11期286-290,共5页
Computer Simulation
基金
陕西省教育厅青年创新团队科研计划项目(22JP102)。
关键词
体育馆室内
空气污染
分布特征
长短期记忆网络
Gymnasium interior
Air pollution
Distribution characteristics
Long short-term memory network
