Abstract:In the context of industrial automation and smart manufacturing, the necessity to optimize the performance of high-power induction furnaces (IFs) has increased considerably. Operational status precise monitoring emerges as a foundational element, crucial for realizing targeted optimization goals. The challenge is twofold. IFs are highly complex thermoelectric systems involving multiple disciplines, which makes them impossible to model accurately. In addition, the electrical characteristic parameters of IF undergo continuous variations during the melting process and cannot be directly measured. In this paper, we propose a novel equivalent circuit model thus facilitating the analysis of the IF operating status. Second, we establish a physics-informed neural network (PINN) to identify the electrical dynamics throughout the melting process. Based on the system constructed by the electrical parameters and the circuit model, we design a novel physical constraint to overcome the limitations of standard PINNs when dealing with complex multidimensional systems. Third, we leverage meta-learning to improve the generalization and fast adaptation of learning models for the PINN-based IF digital twin (DT). Finally, the feasibility of the method is evaluated using actual operational data covering an entire working cycle of IF.

Multi-variable PID Neural Network Decoupling Algorithm in Scrap Copper Smelting Process Control

PIDNN Based Intelligent Control of Ignition Oven

Decoupling Control Design and Simulation of Aluminum Alloy Pulsed MIG Welding Based on Dynamic PLS Framework

A Meta-Pinn Framework for Online Operational Monitoring of High-Power Induction Furnace

A multivariable nonlinear self-tuning PID controller

A Pid Controller With Neuron Tuning Parameters For Multi-Model Plants

Decoupling Control and Simulation for Variable Flow Heating Systems

Combining predictive functional control and PID for liquid level of coking furnace

PID Parameters Auto-Tuning Method for Industrial Temperature Adjustment

The Neural Network Decoupling Control Based on the Internal Model Control

Study and Realization of Automatic Control for Variable Flow Heating Systems

Neuron Model-Free Pid Control

Research on PID Neural Network Control Strategy in Converter Gas Recovery System

STUDY ON MULTIPURPOSE NEURAL PID CONTROLLER AND ITS APPLICATION IN SUPERHEATED STEAM TEMPERATURE CONTROL

Design of Subspace Predictive Controller in Scrap Copper Smelting

The research on regenerative furnace combustion control based on multi-scale wavelet network

Control and Optimization of Negative Pressure in Scrap Copper Smelting

Research on Loop Decoupling Control Based on Fuzzy RBF Neural Network

The Analysis and Design of Paper Multivariable Control System

Self-Tuning Fully-Connected PID Neural Network System for Distributed Temperature Sensing and Control of Instrument with Multi-Modules

Decentralized PID Controllers Tuning Based on Desired Dynamic Equations