方法
各模型使用其规定的线性化普通最小二乘变换。
原始空间 RMSE 使用同一组正值 Ce/qe 数据点进行比较。
返回 计算工具
工程工具 · 吸附与化学品
吸附等温线初步筛选
对实验室数据进行 Freundlich、Langmuir 和 BET 拟合，并分别报告数学拟合与物理可用性状态。
PFAS 与活性炭阶段 03 — Adsorption Sizing
ATLAS WTS · Process Engineering
Activated Carbon Adsorption Isotherm Analyzer
Enter bottle-point equilibrium data and screen Freundlich, Langmuir, and BET isotherms by linearized regression. Physical parameter validity is gated separately, and model comparison uses a common raw-space error metric.
Equilibrium Test Data
Ce = residual (equilibrium) concentration (dissolved for Liquid Phase, in the gas stream for Gas Phase). qe = mass adsorbed per unit mass of carbon, typically qe = (C0−Ce)·V/M from a bottle-point test. At least 3 points are recommended per model for a meaningful regression. Liquid ppm/ppb assume a dilute aqueous solution (density ≈ 1 kg/L); gas ppmv/ppbv/Pa require the Molar Mass field (ideal gas law, 25°C/1 atm). g/kg is numerically identical to mg/g; g/100g is numerically identical to wt%.
Chart Identity (for report title — optional)
Model
Isotherm Fit
Active Model R² —
Single-Point Predictor
Predicted qe
—
Enter a Ce value to predict qe only after the active model passes the physical-parameter gate. The point is highlighted on the chart above with dashed projection lines.
Model Comparison
R² is the linearized-regression metric. Cross-model ranking uses common raw-space qe RMSE on the same positive Ce/qe points; physically unusable fits are excluded from prediction and handoff.
Predicted vs. Observed (Active Model)
Notes & Warnings
Powered by ATLAS WTS · internal tool, not for external distribution
此工程工具目前仅提供英文版本。
01
工程依据
标准 / 参考依据
假设
- 使用实验室 Ce/qe 数据筛查 Freundlich、Langmuir 和 BET 模型。
- 回归拟合成功与参数是否具有物理可用性分别报告。
范围 / 限制
- 用于吸附等温线初步筛查，并分别报告拟合结果和物理可用性。
- 负值、奇异、非有限或其他不可用的参数组不能作为有效预测参数。
- 只有物理可用的拟合才能预测或将参数传至活性炭选型；最终使用前应结合项目台架数据和参考资料核实。