基于电压信号的机器学习模型在极性溶剂识别中的应用  

Application of Machine Learning Model Based on Voltage Signal in Polarity Solvent Identification

作　　者：高文丽 周亮[1] GAO Wenli;ZHOU Liang(School of Forest&Landscape Agriculture,Anhui Agricultural University,Hefei 230036,China)

机构地区：[1]安徽农业大学林学与园林学院,安徽合肥230036

出　　处：《防化研究》2023年第4期66-71,共6页CBRN DEFENSE

摘　　要：不同的溶剂由于外观特征相似,难以便捷准确地区分识别,因此研究一种便捷、精确且安全的溶剂识别方法具有重要意义。本研究以实验室常见的极性溶剂(水以及不同体积分数的水/乙醇的混合溶液)为研究对象,分别采集其滴到发电器件上产生的电压信号作为预测变量,提出一种基于机器学习算法(随机森林算法、k邻近算法和支持向量机算法)构建溶剂识别模型的方法,为安全、准确、便捷地识别极性溶剂提供技术支持。实验结果表明,该方法能够实现对水、乙醇以及体积分数为25%、50%、75%、90%的水/乙醇混合溶液的识别,其中随机森林算法构建的识别模型准确度高达87.5%。Different solvents are difficult to distinguish and identify accurately due to their similar appearance.Therefore,it is important to develop a convenient,accurate and safe method for solvent identification.In this study,common polar solvents in the laboratory were selected as research objects.Voltage signals generated when these solvents were dropped onto a power generating device were collected as predictors,and a machine learning model based on the random forest algorithm,k-nearest neighbor algorithm and support vector machine algorithm was proposed to build the solvent recognition model.This approach provides technical support for the safe,accurate and convenient identification of polar solvents.The experimental results show that the proposed method can accurately recognize the polarity solvent(water,ethanol,25%ethanol,50%ethanol,75%ethanol and 90%ethanol)with a maximum accuracy rate of 87.5%on the test set based on random forest algorithm.

关 键 词：极性溶剂 乙醇 水 机器学习 随机森林 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]