Semi-batch reactor predictive control using artificial neural network

Samek, David; Macků, Lubomír

doi:10.1109/med.2008.4602140

Cited by 4 publications

(7 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here we assume that (24) represents the underlying nonlinear system (5), so that (12) becomeŝ Y (t|θ (k)) = J(Z N , ϕ(k),θ (k)). (27) By interpreting the regression vector ϕ(k) as a vector defining the state of the system, the local linearization of (27) around the current state ϕ(k ′ ) at time k ′ = 1, 2, . .…”

Section: Generalized Predictive Control (Gpc)mentioning

confidence: 99%

“…The most widely used NN architecture for dynamic system modeling is the dynamic feedforward NN (DFNN) [23,[26][27][28]. The use of recurrent NN (RNN) for modelling nonlinear dynamic systems has also been reported [24,[29][30][31].…”

Section: Introductionmentioning

confidence: 99%

“…NN Model-based process identification for use in MPC applications has been reported using dynamic feedforward NN (FNN) [23,[26][27][28] and recurrent NN (RNN) [16,24,29,31]. The MPC methods reported in these papers all use the linearized form of the identified nonlinear NN model and employ the generalized predictive control (GPC) [41] techniques, except in [24] where the GPC control actions are updated on-line in real-time for a relative slow process with long time sample using a recursive least squares adaptation rule that is obtained from a network trained with a basic BP algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Nonlinear model identification and adaptive model predictive control using neural networks

2011

View full text Add to dashboard Cite

Section: Generalized Predictive Control (Gpc)mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Nonlinear model identification and adaptive model predictive control using neural networks

2011

View full text Add to dashboard Cite

“…However, if the NN controller is cascaded in series with the controlled plant, eu(t) is not known since the desired control input Ud(t) is unknown. So, the immediate problem in designing such an NN-controller is how to train the NN [12].…”

Section: B Artificial Neural Networkmentioning

confidence: 99%

“…The neuron sum the weighed inputs and the threshold, and passes the result through its characteristic transfer function. The transfer function is same for all neurons [12]. Weights are commonly labelled w number of layers.…”

Section: B Artificial Neural Networkmentioning

confidence: 99%

Automation of biodiesel plant with bio-sensing technologies

Dhar

Mazumdar

2012

2012 1st International Symposium on Physics and Technology of Sensors (ISPTS-1)

View full text Add to dashboard Cite

Among the alternative fuels, biodiesel and its blends are considered suitable and the most promising fuel for diesel engine. The properties of biodiesel are found similar to that of diesel. Many researchers have experimentally evaluated the performance characteristics of conventional diesel engines fuelled by biodiesel and its blends. The biodiesel production process requires simultaneous real time monitoring of certain properties like viscosity, acid number, iodine value, moisture content etc, such that the values are within ASTM standards. Therefore, an automated Biodiesel plant fitted with new innovative biosensors have been proposed which could make real time decision taking ability and monitor the overall production mechanism. The control system of the biosensor and production process was coupled with ARTIFICIAL NEURAL NETWORK (ANN). The proposed system have capability to detect 8 different starting oil samples and optimizes itself the best production yield of biodiesel. Therefore, the optimizing problems of the quality of initial feed of vegetable oil used for production of biodiesel are overcome. This paper provides experimental knowledge on the enormous studies for the development of advanced reactor system engineering for large scale production of biodiesel for commercial purpose.

show abstract

Predictive control of processes with fast dynamics

Barot

Kubalcik

2015

Proceedings of the 2015 16th International Carpathian Control Conference (ICCC)

View full text Add to dashboard Cite

The predictive control is a modern approach, which can be used for many kinds of processes in the engineering practice. Also processes with fast dynamics can be controlled by the principle of the predictive control. However, there is a problem to realize a control algorithm in a short sampling period in case of the higher horizons and many constraints. Then the optimization task is not feasible using the traditional approach.In this paper, the modification of the controller and optimizer is presented. The proposed approach is based on using of the information from the control and optimization, which is not generally used. The aim is to reduce the computational complexity for this specific case of predictive control. The modified approach is discussed and verified by simulation.

show abstract

Semi-batch reactor predictive control using artificial neural network

Cited by 4 publications

References 7 publications

Nonlinear model identification and adaptive model predictive control using neural networks

Nonlinear model identification and adaptive model predictive control using neural networks

Automation of biodiesel plant with bio-sensing technologies

Predictive control of processes with fast dynamics

Contact Info

Product

Resources

About