摘要
以喷染车间挥发性有机物为研究对象,对喷染车间挥发性有机物(Volatile Organic Compounds, VOCs)质量浓度预测方法进行研究。首先,使用随机森林(Random Forest, RF)算法对影响喷染车间挥发性有机物质量浓度的特征变量进行权重分析。同时,构建基于长短期记忆神经网络(Long-Term and Short-Term Memory Neural Network, LSTM)的挥发性有机物质量浓度预测模型,并在此基础上引入麻雀搜索算法(Sparrow Search Algorithm, SSA)进行参数优化选择。最后,以浙江省杭州市某汽车喷染车间7月29日—10月28日的数据为样本,将温度、相对湿度、室内大气压、室外大气压作为模型输入变量,并与LSTM模型、随机森林-长短期记忆神经网络(Random Forest-Long Short-Term Memory neural network, RF-LSTM)模型、随机森林-反向传播神经网络(Random Forest-BP neural network, RF-BP)模型进行对比试验。结果显示,基于随机森林-麻雀搜索算法-长短期记忆神经网络(Random Forest-Sparrow Search Algorithm-Long Short-Term Memory neural network, RF-SSA-LSTM)模型的预测效果最佳,平均绝对误差、均方根误差和决定系数分别为2.812 2、3.457 4、0.988。同时,为验证RF-SSA-LSTM模型性能,通过不同时间步长实现对喷染车间VOCs质量浓度预测,结果显示预测误差较小,在可接受范围内。RF-SSA-LSTM预测模型提高了挥发性有机物质量浓度的预测精度,为减少挥发性有机物排放提供科学依据。
This paper made Volatile Organic Compounds(VOCs)in a spray dyeing workshop be the subject investigated,to study the prediction method of volatile organic compounds in a spray dyeing workshop.Firstly,the Random Forest(RF)algorithm was used to analyze the weight of the characteristic variable that affects the mass concentration of volatile organic compounds in the spray dyeing workshop.A prediction medal of volatile organic compounds mass concentration based on Long Short-Term Memory neural network(LSTM)was constructed and the Sparrow Search Algorithm(SSA)was introduced to optimize the parameters.Finally,with the data from July 29 to October 28 of an automobile spray dyeing workshop in Hangzhou,Zhejiang Province as the sample,temperature,humidity,indoor atmospheric pressure,and outdoor atmospheric pressure were selected as the model input variables and were compared with LSTM model,RF-LSTM model,RF-BP model,RF-SSA-LSTM model.The prediction effect of the SSA-LSTM model is the best,with IMAE,IRSE,and R°being 2.8122,3.4574,and 0.988,respectively.Besides,to validate the performance of the model,this paper also realizes the prediction of the mass concentration of volatile organic compounds in the spray dyeing workshop through different time steps.The average absolute errors(MAE)of volatile organic compounds in the spray dyeing workshop prediction in advance of 48 h,72 h,96 h,and 120 h were 5.9842,8.6243,12.7084,and 19.3338,respectively.The results show that the prediction error is small,within an acceptable range.The prediction model proposed in this paper improves the prediction accuracy of the mass concentration of volatile organic compounds,can more accurately guide the terminal management load and can provide a scientific basis for reducing the emission of volatile organic compounds,and the also has an ideal prediction effect on the frequency under the prediction of volatile organic compounds mass concentration of different time lengths,which confirms the feasibility of this research scheme.
作者
彭来湖
张权
李建强
李杨
PENG Laihu;ZHANG Quan;LI Jianqiang;LI Yang(School of Mechanical Engineering,Zhejiang Sci-Tech University,Hangzhou 310000,China;ZhejianggSci-tech University Longgang Research Institute,Wenzhou 325000,Zhejiang,China;School of Mechanical Engineering,Zhejiang University,Hangzhou 310000,China)
出处
《安全与环境学报》
CAS
CSCD
北大核心
2024年第1期186-195,共10页
Journal of Safety and Environment
基金
浙江省博士后科研项目(ZJ2020004)。
关键词
安全卫生工程技术
挥发性有机物
随机森林
麻雀搜索算法
LSTM神经网络
safety and hygiene engineering technology
volatile organic compounds
Random Forest(RF)
SSA
Long Short-Term Memory neural network(LSTM)