A physics informed machine learning framework for optimal sensor placement and parameter estimation

doi:10.1016/j.compchemeng.2026.109652

A physics informed machine learning framework for optimal sensor placement and parameter estimation

2026-04-04

0

PRE

AI

G

Georgios Venianakis

C

Constantinos Theodoropoulos *

M

Michail E. Kavousanakis *

DOI：10.1016/j.compchemeng.2026.109652

Original

Original request for help

Save

Abstract

En 中文

• A PINN framework for joint sensor placement and parameter estimation is presented. • Sensor locations are optimized using sensitivity functions and D-optimal design. • The framework is demonstrated for reaction–diffusion–advection systems. • Optimal placement yields improved parameter estimates over heuristic choices.

Keywords:

Physics-Informed Neural Networks

Sensor Placement Optimization

Parameter Estimation

D-Optimal Design

Reaction-Diffusion-Advection Systems

A physics informed machine learning framework for optimal sensor placement and parameter estimation

Abstract

Journal

Organization