ScholarLens Back to search
2010•Machine Tool & HydraulicsRequires access

Application of Internal Model Control Based on Neural Network for CSTR System Simulation

Liang Ma

Open publisher page 0 citations

Abstract

The structure and property of internal model control(IMC)based on neural network were studied to overcome the shortcomings of nonlinearity and time-delay of continuous stirred tank reactor(CSTR)system.After analyzing the characteristic of CSTR system,the IMC controller was constructed based on neural network.Both the CSTR system model and the controller were constructed via BP neural network.LMBP algorithm was used to train the weights of the neural network online.In MPCE-1000 multifunction process control experimental platform,the simulation analysis of the CSTR system was carried on to confirm the validity of the designed scheme.The simulation results demonstrate the performance of this algorithm is better than that of PID algorithm.It has good dynamic and static properties,and the impacts of nonlinearity and time-delay are solved.

About this research paper

What this paper is about

The structure and property of internal model control(IMC)based on neural network were studied to overcome the shortcomings of nonlinearity and time-delay of continuous stirred tank reactor(CSTR)system.After analyzing the characteristic of CSTR system,the IMC controller was constructed based on neural network.Both the CSTR system model and the controller were constructed via BP neural network.LMBP algorithm was used to train the weights of the neural network online.In MPCE-1000 multifunction process control experimental platform,the simulation analysis of the CSTR system was carried on to confirm the validity of the designed scheme.The simulation results demonstrate the performance of this algorithm is better than that of PID algorithm.It has good dynamic and static properties,and the impacts of nonlinearity and time-delay are solved.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The structure and property of internal model control(IMC)based on neural network were studied to overcome the shortcomings of nonlinearity and time-delay of continuous stirred tank reactor(CSTR)system.After analyzing the characteristic of CSTR system,the IMC controller was constructed based on neural network.Both the CSTR system model and the controller were constructed via BP neural network.LMBP algorithm was used to train the weights of the neural network online.In MPCE-1000 multifunction process control experimental platform,the simulation analysis of the CSTR system was carried on to confirm the validity of the designed scheme.The simulation results demonstrate the performance of this algorithm is better than that of PID algorithm.It has good dynamic and static properties,and the impacts of nonlinearity and time-delay are solved.

Key concepts: Continuous stirred-tank reactor, Control theory (sociology), Artificial neural network, Nonlinear system, Internal model, Controller (irrigation), PID controller, Control engineering

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
Application of Internal Model Control Based on Neural Network for CSTR System Simulation — Research Paper | ScholarLens