| 引用本文: | 孙叶,吴宝利,马军,尤世界.基于主成分分析的神经网络用于新污染物环境影响预测[J].哈尔滨工业大学学报,2026,58(6):31.DOI:10.11918/202511031 |
| SUN Ye,WU Baoli,MA Jun,YOU Shijie.Neural network based on principal component analysis for predicting environmental impacts of emerging contaminants[J].Journal of Harbin Institute of Technology,2026,58(6):31.DOI:10.11918/202511031 |
|
| |
|
|
| 本文已被：浏览 1026次 下载 103次 |
码上扫一扫！ |
|
|
| 基于主成分分析的神经网络用于新污染物环境影响预测 |
|
孙叶1,2,吴宝利1,3,马军1,2,尤世界1,2
|
|
(1.哈尔滨工业大学 环境学院,哈尔滨 150090; 2.城乡水资源与水环境全国重点实验室(哈尔滨工业大学), 哈尔滨 150090; 3.中国市政工程华北设计研究总院有限公司,天津 300381)
|
|
| 摘要: |
| 新污染物具有结构多样性、环境持久性和生物毒性,对生态系统和人类健康构成潜在风险。生命周期评价（Life Cycle Assessment,LCA）是评估其环境影响的重要方法,但存在基础数据缺失、难以准确量化等问题。为此,提出基于主成分分析（Principal Component Analysis,PCA）优化输入特征的人工神经网络（Artificial Neural Network,ANN）模型,通过对大量分子描述符进行PCA降维,提取主要特征以减少数据冗余和维度负担,用于实现新污染物的环境影响预测。构建了4种不同特征保留比例的模型,即NOPCA（全部描述符）、PCA95（保留累计方差贡献率95%的主成分）、PCA85（保留累计方差贡献率85%的主成分）和PCA75（保留累计方差贡献率75%的主成分）,评价并比较预测性能差异。结果表明:适度的特征提取可提升模型泛化能力,在全球变暖潜势（GWP）和人类毒性潜势（HTP）预测任务中,采用PCA85预处理描述符所构建的模型表现最佳,测试集R2分别为0.64和0.73；在化石能源消耗潜势（FDP）和陆地酸化潜势（TAP）预测任务中,模型则在PCA95时达到最佳,测试集R2分别为0.77和0.76。基于PCA特征提取的ANN模型可有效缓解LCA数据不全问题,为新污染物环境影响预测提供了高效、可靠的方法。 |
| 关键词: 主成分分析 新污染物 人工神经网络 生命周期评价 分子描述符 |
| DOI：10.11918/202511031 |
| 分类号:X502 |
| 文献标识码:A |
| 基金项目:国家自然科学基金(U25A5,6,52325003) |
|
| Neural network based on principal component analysis for predicting environmental impacts of emerging contaminants |
|
SUN Ye1,2,WU Baoli1,3,MA Jun1,2,YOU Shijie1,2
|
|
(1.School of Environment, Harbin Institute of Technology, Harbin 150090, China; 2.State Key Laboratory of Urban-rural Water Resource and Environment (Harbin Institute of Technology), Harbin 150090, China;3.North China Municipal Engineering Design & Research Institute Co., Ltd., Tianjin 300381, China)
|
| Abstract: |
| Emerging contaminants have structural diversity, environmental persistence, and biological toxicity, which pose potential risks to ecosystems and human health. Life Cycle Assessment (LCA) is an important method for evaluating their environmental impacts, but it has problems such as a lack of basic data and difficulty in accurate quantification. To address these issues, this paper proposed an Artificial Neural Network (ANN) model based on Principal Component Analysis (PCA) to optimize input features. By performing PCA dimensionality reduction on a large number of molecular descriptors, main features were extracted to reduce data redundancy and dimensional burden for achieving the prediction of environmental impacts of emerging contaminants. Four models with different feature retention ratios were constructed, namely NOPCA (all descriptors), PCA95 (retaining principal components with 95% cumulative variance contribution rate), PCA85 (retaining principal components with 85% cumulative variance contribution rate), and PCA75 (retaining principal components with 75% cumulative variance contribution rate), and the differences in prediction performance were evaluated and compared. The results show that moderate feature extraction can improve the generalization ability of the model; in the prediction tasks of global warming potential (GWP) and human toxicity (HTP), the model constructed using PCA85-preprocessed descriptors performs the best, with test-set R2 of 0.64 and 0.73, respectively; in the prediction tasks of fossil energy depletion (FDP) and terrestrial acidification (TAP), the model achieves optimal performance under PCA95, with test-set R2 of 0.77 and 0.76, respectively. The ANN model based on PCA feature extraction can effectively mitigate the problem of incomplete LCA data and provides an efficient and reliable method for predicting the environmental impacts of emerging contaminants. |
| Key words: Principal Component Analysis emerging contaminant Artificial Neural Network life cycle assessment molecular descriptor |
|
|
|
|
