Model predictive control of nonlinear affine systems based on the general projection neural network and its application to a continuous stirred tank reactor

2011 ◽  
Author(s):  
Zheng Yan ◽  
Jun Wang
Keyword(s):  
Neural Network ◽  
Model Predictive Control ◽  
Predictive Control ◽  
Stirred Tank ◽  
Continuous Stirred Tank Reactor ◽  
Stirred Tank Reactor ◽  
Tank Reactor ◽  
General Projection ◽  
Continuous Stirred Tank ◽  
Projection Neural Network

Improving the Performance of a Continuous Stirred Tank Reactor Using Moving Horizon State Estimation and Model Predictive Control

2011 ◽  
Vol 403-408 ◽  
pp. 3454-3460
Author(s):  
Fazlollah Armoon ◽  
Hooshang Jazayeri-Rad
Keyword(s):  
Model Predictive Control ◽  
State Estimation ◽  
Predictive Control ◽  
Stirred Tank ◽  
Continuous Stirred Tank Reactor ◽  
Inlet Concentration ◽  
Product Concentration ◽  
Stirred Tank Reactor ◽  
Tank Reactor ◽  
Continuous Stirred Tank

Since chemical reactors are utilized to produce specific and valuable products, concentration of products should be regulated at a specified level. As a disturbance input, a change in the inlet concentrations can vary the product concentration. So, in order to regulate the product concentration, the inlet concentrations and the product concentration should be measured. However, measurement of concentration encounters some problems such as high cost and time delay. For compensation of these failures, estimation of concentration has been proposed. In this work, the inlet concentration and the product concentration of a continuous stirred-tank reactor (CSTR) are estimated based on the moving horizon state estimation (MHSE), and the product concentration is regulated based on the model predictive control (MPC). Simulation results indicate that the proposed strategy improves the performance of the CSTR compared with the method in which the inlet concentration is not estimated.

Model Predictive Control of Continuous Stirred-Tank Reactor

2013 ◽  
Vol 760-762 ◽  
pp. 1000-1003
Author(s):  
Shi Li ◽  
Xi Ju Zong ◽  
Yan Hu
Keyword(s):  
Model Predictive Control ◽  
Predictive Control ◽  
State Space Model ◽  
Stirred Tank ◽  
Continuous Stirred Tank Reactor ◽  
Stirred Tank Reactor ◽  
Tank Reactor ◽  
Nonlinear Mechanical ◽  
State Point ◽  
Continuous Stirred Tank

This paper is concerns with the study of model predictive control of continuous stirred-tank reactor (CSTR). Firstly, a nonlinear mechanical model is established for a typical CSTR. Based on mass balance and energy balance, a set of ordinary differential equations are derived. Then, the model is linearized at steady-state point to obtain state space model, which is used in model predictive control (MPC). Finally, MPC strategy is applied, controller parameters and process constrains are discussed.

Sign in / Sign up

Export Citation Format

Share Document