correction-ea-vhr 0.2

Neural network correction tool for activation energy in VHR thermoluminescence

Correction of Activation Energy Affected by Variable Heating Rate (EaVHRCorrector)

===================================================================================

This project presents the development of a tool based on artificial intelligence to assist researchers in the correction of kinetic parameters obtained through the deconvolution of thermoluminescence TL glow curves. The correction is particularly useful when the results are affected by anomalous behavior associated with the heating rate, mainly due to Thermal Quenching.

Overview

The model takes as input the area under each TL emission peak, which is a magnitude easily obtained from the analysis of experimental data. The input to the model is a vector of differences in area:

X ≡ [∆A_medium, ∆A_median, ∆A_maximum, ∆A_minimum]

Features

• AI-based correction of activation energy (Ea) affected by variable heating rate
• Designed for thermoluminescence glow curve analysis
• Accepts Excel files as input
• Provides corrected Ea, uncertainty, and model prediction

Usage

Here is a minimal example of how to use the tool:

from correction_ea_vhr.corrector import EaVHRCorrector
import pandas as pd

# Load your data
df = pd.read_excel("TLGC.xlsx")
X = df.iloc[:, 0:4].to_numpy()
Y = df.iloc[:, 4:5].to_numpy()

# Use the corrector
modelo = EaVHRCorrector()
media, incertidumbre, pred = modelo.corregir(X, Y)

Input Format

The input Excel file should contain 5 columns:

1. ∆A_medium
2. ∆A_median
3. ∆A_maximum
4. ∆A_minimum
5. Experimental activation energy (Ea)

Output

• media: Corrected activation energy (mean)
• incertidumbre: Associated uncertainty
• pred: Model prediction array