Scaling up machine learning-based chemical plant simulation: A method for fine-tuning a model to induce stable fixed points

TL;DR

This paper presents a method to fine-tune machine learning models for chemical plant simulation to ensure stable initialization and convergence, especially in large, complex plants with nested cycles.

Contribution

It introduces a fine-tuning approach that enhances the stability of ML-based chemical plant models during initialization, addressing issues caused by complex plant dynamics.

Findings

Fine-tuning improves model stability during initialization

The approach enables reliable simulation of larger, more complex plants

Addresses solver convergence issues in ML-based plant models

Abstract

Idealized first-principles models of chemical plants can be inaccurate. An alternative is to fit a Machine Learning (ML) model directly to plant sensor data. We use a structured approach: Each unit within the plant gets represented by one ML model. After fitting the models to the data, the models are connected into a flowsheet-like directed graph. We find that for smaller plants, this approach works well, but for larger plants, the complex dynamics arising from large and nested cycles in the flowsheet lead to instabilities in the solver during model initialization. We show that a high accuracy of the single-unit models is not enough: The gradient can point in unexpected directions, which prevents the solver from converging to the correct stationary state. To address this problem, we present a way to fine-tune ML models such that initialization, even with very simple solvers, becomes robust.