CorrAI: A Self-driving Lab (SDL) for the closed-loop optimization of corrosion protection technologies

The data, codebase and figure-generation scripts used by the paper CorrAI: An AI-Driven Automated Platform to Bayesian Optimize corrosion inhibitors in electrochemical systems are stored in this record.The contents of the package are: (i) the complete processed electrochemical data that was employed in the study, (ii) folders of detailed measurements of OCP, EIS, LPR, and chronoamperometry measurements associated with each data item, (iii) code used to accomplish Gaussian-process-based Bayesian Optimization of inhibitor compositions, (iv) code used to compute replicate statistics and regression measures, and inhibition-power reproducible to recreate Figures 1-9 of the study, and (v) all the code needed to generate figures necessary to reproduce The Data/ directory has the master dataset that was cleaned (Dataset.csv) and organized subfolders with both raw and processed files of electrochemical measurements. The Figures/ directory has the figure results utilized in the manuscript (heatmaps, correlation matrices, sampling maps, GP parity plots, and optimization trajectories). All the analyses, model predictions, and visualizations reported in the Results and Discussion sections are reproducible using the Python scripts that are included (optimization.py, Analysis.py, corrobotdataanalysis.py and db connection.py). This archive offers complete computational re-producibility of analysis and optimization bits of the CorRobot platform. Higher level code Hardware-control and automation code are excluded on purpose, because they depend on the system in ways that are not very general.本存档收录了论文《CorrAI：电化学体系中腐蚀抑制剂贝叶斯优化的AI驱动自动化平台》（CorrAI: An AI-Driven Automated Platform to Bayesian Optimize corrosion inhibitors in electrochemical systems）所使用的数据、代码库与绘图脚本。该存档包含以下内容：(i) 本研究中使用的全部预处理电化学数据；(ii) 与各数据条目对应的开路电位（OCP, Open Circuit Potential）、电化学阻抗谱（EIS, Electrochemical Impedance Spectroscopy）、线性极化电阻（LPR, Linear Polarization Resistance）及计时电流法测试的详细测量数据文件夹；(iii) 用于实现基于高斯过程（Gaussian Process, GP）的抑制剂配方贝叶斯优化的代码；(iv) 用于计算重复统计量与回归指标、实现缓蚀效率可复现性并复现本研究图1至图9的代码；(v) 所有用于生成必要图表以完成复现工作的代码。 Data/目录下包含经清理后的主数据集（Dataset.csv），以及包含电化学测试原始与处理后文件的结构化子文件夹。Figures/目录下包含论文手稿中使用的图表结果，包括热图、相关矩阵、采样分布图、高斯过程（GP）奇偶图及优化轨迹图。本研究结果与讨论部分中提及的所有分析、模型预测及可视化内容，均可通过附带的Python脚本（optimization.py、Analysis.py、corrobotdataanalysis.py及db connection.py）实现复现。 本存档可完整复现CorRobot平台的分析与优化模块的计算流程。本存档刻意未包含更上层的硬件控制与自动化代码，因为这些代码依赖于特定系统，通用性较差。