Reinforcement Learning-Driven Plant-Wide Refinery Planning Using Model Decomposition
Zhouchang Li, Runze Lin, Hongye Su, Lei Xie
TL;DR
The paper addresses plant-wide refinery planning under price volatility by reformulating the global MINLP into interactable sub-models via model decomposition, coordinated by a DRL-based pricing strategy. A PPO-Clip-based pricing agent learns period-specific parameters $\lambda_{i,t}$ to steer each sub-model toward global optimality, while sub-models are solved sequentially to assemble the full solution. Three industrial case studies demonstrate substantial computational speedups (up to >95%) and profitability gains (e.g., ~23.6% over deterministic pricing; ~4.7% improvement over multi-period MINLP baselines), and show robustness to price fluctuations. The approach offers a scalable, adaptive framework for large-scale refinery planning that can exploit distributed computing and better respond to market dynamics.
Abstract
In the era of smart manufacturing and Industry 4.0, the refining industry is evolving towards large-scale integration and flexible production systems. In response to these new demands, this paper presents a novel optimization framework for plant-wide refinery planning, integrating model decomposition with deep reinforcement learning. The approach decomposes the complex large scale refinery optimization problem into manageable submodels, improving computational efficiency while preserving accuracy. A reinforcement learning-based pricing mechanism is introduced to generate pricing strategies for intermediate products, facilitating better coordination between submodels and enabling rapid responses to market changes. Three industrial case studies, covering both single-period and multi-period planning, demonstrate significant improvements in computational efficiency while ensuring refinery profitability.