摘要
反应动力学是物理化学课程的重要组成部分,其中反应速率常数的提取与动力学方程的建模是教学中的核心难点。以蔗糖在盐酸催化下的水解反应为案例,设计了基于Python的3种建模路径:传统线性拟合、神经网络和符号回归,并结合具体教学环节进行了比较与应用。3种方法分别体现了建模的直观性、拟合能力与可解释性,帮助学生从多个角度理解建模过程,逐步拓展其分析与思维深度。本案例已融入河北大学本科生通识选修课“趣味机器学习”的实际教学中,教学反馈显示学生能够掌握建模基本流程,理解不同算法的特点与适用场景,显著提升了跨学科综合应用能力与数据素养。通过本教学设计,构建了理论知识、实验数据与建模方法的有机融合路径,为“化学+人工智能”背景下的课程改革提供了可迁移的实践案例。
Chemical kinetics is a fundamental component of physical chemistry courses,in which the extraction of rate constants and the modeling of kinetic equations represent core challenges in teaching.Taking the acid-catalyzed hydrolysis of sucrose as a case study,this work designs a three-tier modeling framework based on Python:traditional linear fitting,neural network,and symbolic regression.These methods respectively highlight intuitiveness,fitting accuracy,and explainability,enabling students to understand the modeling process from multiple perspectives and progressively deepen their analytical thinking.This teaching case has been integrated into the undergraduate general elective course Fun Machine Learning at Hebei University.Feedback from implementation indicates that students have effectively mastered the basic modeling workflow,gained insights into the characteristics and applicability of different algorithms,and significantly improved their interdisciplinary competence and data literacy.The proposed design establishes an organic integration of theoretical knowledge,experimental data,and modeling techniques,offering a transferable and practical case study for curriculum reform under the framework of“Chemistry+Artificial Intelligence”.
作者
张照胜
霍树营
张红
孙素芳
马海云
马京
ZHANG Zhao-Sheng;HUO Shu-Ying;ZHANG Hong;SUN Su-Fang;MA Hai-Yun;MA Jing(College of Chemistry and Materials Science,National Experimental Teaching Demonstration Center of Chemistry,Hebei University,Baoding 071002,China)
出处
《化学教育(中英文)》
2026年第2期114-122,共9页
Chinese Journal of Chemical Education
基金
国家自然科学基金青年项目(22103021)
河北省自然科学基金优秀青年项目(B2024201051)
河北省教育厅青年拔尖人才项目(BJK2024094)
材料化学国家级一流本科专业(YS23-YLZY-021)
河北大学人工智能赋能教改专项课题(XRGZN15)。
关键词
反应动力学
蔗糖水解
机器学习
神经网络
符号回归
chemical kinetics
sucrose hydrolysis
machine learning
neural network
symbolic regression
