Artificial intelligence-based fluid chemical reaction process prediction and optimization method and system

By constructing a hierarchical hybrid modeling system that combines mechanism delimitation, data optimization, and constraint guarantees, the problems of difficult model parameter identification and poor real-time performance in chemical processes are solved. This enables high-precision and robust prediction and optimization of fluid chemical reaction processes, and is applicable to multiphase flow, transient heat and mass transfer, and heterogeneous catalytic reactions.

CN122224378APending Publication Date: 2026-06-16SHANGHAI INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHANGHAI INST OF TECH
Filing Date
2026-04-22
Publication Date
2026-06-16

AI Technical Summary

Technical Problem

Existing technologies in the chemical, petrochemical, and fine chemical industries suffer from difficulties in identifying model parameters for fluid chemical reaction processes, poor real-time performance, weak generalization ability, lack of complete solutions for collaborative work with DCS/PLC, and insufficient accuracy and stability in modeling multiphase flow, transient heat and mass transfer, and heterogeneous catalytic reactions.

Method used

An AI-based method for predicting and optimizing fluid chemical reaction processes is adopted. By constructing a hierarchical hybrid modeling system that combines mechanism delimitation, data supplementation, constraint guarantee, and real-time optimization, and combining multimodal data acquisition, preprocessing, online identification, and dynamic updating, integrated multi-objective and strongly constrained optimization control is executed to achieve linkage with the field control system.

Benefits of technology

It improves prediction accuracy and generalization ability, enhances robustness and online executability, ensures the safety of process operation and industrial adaptability, and is suitable for complex chemical reaction systems with multiphase coexistence and multi-scale coupling.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122224378A_ABST
    Figure CN122224378A_ABST
Patent Text Reader

Abstract

The application provides a kind of based on artificial intelligence's fluid chemical reaction process prediction and optimization method and system, the method includes: step 1, the acquisition of fluid chemical reaction process multi-modal original data;Step 2, original data preprocessing;Step 3, construct coarse-grained mechanism model+data-driven residual model+physical constraint neural network's hierarchical hybrid modeling system;Step 4, online identification and dynamic updating;Step 5, based on hybrid model executes integrated multi-objective and strong constraint optimization control strategy;Step 6, the prediction result and optimization strategy output to field control system.The method of the application can accurately predict key process variables through hierarchical hybrid model;Online identification and updating realize process real-time optimization;Through constraint control, optimization under complex conditions is realized.The method of the application can significantly improve production efficiency, reduce energy consumption, improve target product yield, and ensure safe and stable operation of the reaction process.
Need to check novelty before this filing date? Find Prior Art