Predictive control based on neural networks: an application to a fluid catalytic cracking industrial unit

Santos, Vera; Carvalho, Florival Rodrigues de; Souza, Maurício B. de

doi:10.1590/s0104-66322000000400054

Cited by 8 publications

(10 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The first FLN was used to predict substrate concentration in the fermentor one step ahead with the following structure (Zhan and Ishida, 1997;Santos et al, 2000):…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Non-linear multivariable predictive control of an alcoholic fermentation process using functional link networks

Meleiro

Costa

Filho

2005

Braz. arch. biol. technol.

View full text Add to dashboard Cite

In this work a MIMO non-linear predictive controller was developed for an extractive alcoholic fermentation process. The internal model of the controller was represented by two MISO Functional Link Networks (FLNs), identified using simulated data generated from a deterministic mathematical model whose kinetic parameters were determined experimentally. The FLN structure presents as advantages fast training and guaranteed convergence, since the estimation of the weights is a linear optimization problem. Besides, the elimination of non-significant weights generates parsimonious models, which allows for fast execution in an MPC-based algorithm. The proposed algorithm showed good potential in identification and control of non-linear processes.

show abstract

“…The first FLN was used to predict substrate concentration in the fermentor one step ahead with the following structure (Zhan and Ishida, 1997;Santos et al, 2000):…”

Section: Resultsmentioning

confidence: 99%

“…) ( 1 ) Many authors suggested that the neural networks used as internal models of MPC schemes should be validated by testing their capacity to predict steady states of the process (Zhan and Ishida, 1997;Santos et al, 2000). Figs.…”

Section: Resultsmentioning

confidence: 99%

Non-linear multivariable predictive control of an alcoholic fermentation process using functional link networks

Meleiro

Costa

Filho

2005

Braz. arch. biol. technol.

View full text Add to dashboard Cite

show abstract

“…Many authors suggest that the capacity of the network to predict the steady states of the process determines its performance as internal model of non-linear predictive controllers (Zhan and Ishida, 1997;Fonseca, 1998;Santos et al, 2000). For the process studied, the neural network was only capable of predicting the steady states when the interval between the changes made in the manipulated variable to generate the training data set was large enough for new steady states to be attained between two changes.…”

Section: Discussionmentioning

confidence: 99%

“…manipulated and controlled variables at the present sampling time and, as output, the controlled variable one step ahead (Zhan and Ishida, 1997;Santos et al, 2000). The training data is obtained by randomly changing the manipulated variable.…”

Section: Neural Network Modelmentioning

confidence: 99%

See 1 more Smart Citation

Non‐Linear Predictive Control of a Three‐Phase Catalytic Reactor

Costa

Filho

2003

Can J Chem Eng

View full text Add to dashboard Cite

A t the centre of a Model Predictive Control (MPC) algorithm is a dynamic model. Until recently, most industrial applications have used linear dynamic models. These models were preferred to the non-linear ones due to the difficulty in developing generic non-linear models from plant data and also because of the computational expense involved when non-linear models are used in the MPC formulation (Piché et al., 2000). The rapid development in computer capacity in the last few years, however, has led to the development of a great number of predictive control algorithms using non-linear models.One of the main problems with linear controllers is that they are not robust, i.e., they do not present good performance for different operating regions and with changing process conditions (Dutta and Rhinehart, 1999). According to Piché et al. (2000), MPC based on linear models is acceptable when the process operates at a single set point and the primary use of the controller is for the rejection of disturbances. However, many chemical processes do not operate at a single set point, as they are operated in different conditions depending on market requirements. The operation in different set points also occurs when an optimization layer is introduced in the control structure. The main characteristic of a multilayer optimization structure is that the optimization level defines the operational conditions of the process. Thus, the control system has to be designed to maintain the process operating adequately in different operational conditions. In this case, to properly control the process, a non-linear model is needed.Although the use of non-linear models may improve the control algorithm performance, the development of such models is not always an easy task. It is desirable to develop techniques to obtain reliable models in a simple and rapid way. In recent years, there has been a strong interest in the use of neural networks to describe chemical processes, due to their ability to approximate highly non-linear systems. Different structures of neural networks have been used as non-linear models in advanced control algorithms (Zhan and Ishida, 1997;Rohani et al., 1999;Kambhampati et al., 2000;Meleiro et al., 2001). Hussain (1999) presents a review of neural network applications in process control. They conclude that neural networks are capable of describing the system in a great number of cases. Besides, they are versatile, as they can be incorporated in various well-known non-linear control methods. Seborg (1999) provides a perspective on the current status of advanced process control and concludes that many industrial applications of nonlinear predictive control employ neural networks. In this work, the predictive control of a three-phase catalytic reactor is considered. A predictive control algorithm, which has a non-linear internal model represented by functional link networks, is proposed. This network structure has been shown to have a good non-linear approximation capability, with the advantage that the estimation of its we...

show abstract

Multivariable predictive control of a pressurized tank using neural networks

Duarte‐Mermoud

Suárez-García

Bassi

2005

Neural Comput & Applic

View full text Add to dashboard Cite

The behavior of a multivariable predictive control scheme based on neural networks applied to a model of a nonlinear multivariable real process, consisting of a pressurized tank is investigated in this paper. The neural scheme consists of three neural networks; the first is meant for the identification of plant parameters (identifier), the second one is for the prediction of future control errors (predictor) and the third one, based on the two previous, compute the control input to be applied to the plant (controller). The weights of the neural networks are updated on-line, using standard and dynamic backpropagation. The model of the nonlinear process is driven to an operation point and it is then controlled with the proposed neural control scheme, analyzing the maximum range over the neural control works properly.

show abstract

Predictive control based on neural networks: an application to a fluid catalytic cracking industrial unit

Cited by 8 publications

References 4 publications

Non-linear multivariable predictive control of an alcoholic fermentation process using functional link networks

Non-linear multivariable predictive control of an alcoholic fermentation process using functional link networks

Non‐Linear Predictive Control of a Three‐Phase Catalytic Reactor

Multivariable predictive control of a pressurized tank using neural networks

Contact Info

Product

Resources

About