Abstract
This study proposes an integrated Kriging-gene expression programming (GEP) framework for simulating nonlinear processes in granular materials.This method uses the Kriging method to enhance the data of uncertainty perception, and realizes interpretable symbolic regression through GEP applied to two case studies (a) the electrochemical degradation of bromophenol blue on boron-doped diamond (BDD) electrodes and (b) the abrasive water jet (AWJ) trimming of diamond grinding wheels. The proposed framework is superior to the traditional response surface method (RSM), increasing R2 from 0.72 to 0.84 to 0.81–0.92. Symbolic equations reveal terms of physical significance, and sensitivity/SHAP analysis identifies the dominant process parameters. This work provides a transparent and robust modeling tool for complex material processes with limited data.
| Original language | English |
|---|---|
| Article number | 122283 |
| Journal | Powder Technology |
| Volume | 476 |
| DOIs | |
| Publication status | Published - Jun 2026 |
Keywords
- Abrasive waterjet (AWJ) machining
- Boron-doped diamond (BDD)
- Diamond
- Gene expression programming (GEP)
- Graphene
- Particulate materials
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver