A novel analysis and design of a neural network assisted nonlinear controller for a bioreactor

Efe, Mehmet Önder; Abadoglu, Ender; Kaynak, O.

doi:10.1002/(sici)1099-1239(199909)9:11<799::aid-rnc441>3.3.co;2-l

Cited by 6 publications

(12 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The process is valuable as its dynamics consists of coupled and nonlinear differential equations, which display a rich set of behaviors containing limit cycles, attractors and repellers [9,10]. The bioreactor is a tank containing a mixture of water, biological cells and nutrients.…”

Section: Data Sets 1-2: Bioreactor Benchmark Problemmentioning

confidence: 99%

“…The state of the process is the amount of cells denoted by c k 1 ∈[0,1] and the amount of nutrients denoted by c k 2 ∈ [0, 1]. The two difference equations obtained after the discretization of the continuous time process are given in (4)-(5), where γ = 0.48 and β = 0.02 are the nutrient inhibition parameter the growth rate parameter, respectively [9][10][11].…”

Section: Data Sets 1-2: Bioreactor Benchmark Problemmentioning

confidence: 99%

See 1 more Smart Citation

Novel Neuronal Activation Functions for Feedforward Neural Networks

Efe

2008

Neural Process Lett

View full text Add to dashboard Cite

Feedforward neural network structures have extensively been considered in the literature. In a significant volume of research and development studies hyperbolic tangent type of a neuronal nonlinearity has been utilized. This paper dwells on the widely used neuronal activation functions as well as two new ones composed of sines and cosines, and a sinc function characterizing the firing of a neuron. The viewpoint here is to consider the hidden layer(s) as transforming blocks composed of nonlinear basis functions, which may assume different forms. This paper considers 8 different activation functions which are differentiable and utilizes Levenberg-Marquardt algorithm for parameter tuning purposes. The studies carried out have a guiding quality based on empirical results on several training data sets.

show abstract

Section: Data Sets 1-2: Bioreactor Benchmark Problemmentioning

confidence: 99%

Section: Data Sets 1-2: Bioreactor Benchmark Problemmentioning

confidence: 99%

Novel Neuronal Activation Functions for Feedforward Neural Networks

Efe

2008

Neural Process Lett

View full text Add to dashboard Cite

show abstract

“…where c 1 ðtÞ is the cell concentration that is considered as the output of the process ðyðtÞ ¼ c 1 ðtÞÞ; c 2 ðtÞ is the amount of nutrients per unit volume and uðtÞ is the flow rate as the control signal [36]. In our simulations, gðtÞ is considered as the time-varying parameter of the process, where its nominal value is g nom ¼ 0:48: The aim of the control task is to keep the cell concentration c 1 ðtÞ as close as possible to the reference trajectory by manipulating the flow rate, magnitude of which is allowed to be between u min ¼ 0 and u max ¼ 2; and duration of which is kept constant at t min ¼ t max ¼ 0:5 s: The GPC parameters appeared in the performance index (2) are set to N 2 ¼ 10; N u ¼ 2 and l j ¼ 0:0001 for j ¼ 1; .…”

Section: Bioreactormentioning

confidence: 99%

Online trained support vector machines-based generalized predictive control of non-linear systems

İplikçi

2006

Int. J. Adapt. Control Signal Process.

View full text Add to dashboard Cite

In this work, an online support vector machines (SVM) training method (Neural Comput. 2003; 15: 2683-2703, referred to as the accurate online support vector regression (AOSVR) algorithm, is embedded in the previously proposed support vector machines-based generalized predictive control (SVM-Based GPC) architecture (Support vector machines based generalized predictive control, under review), thereby obtaining a powerful scheme for controlling non-linear systems adaptively. Starting with an initially empty SVM model of the unknown plant, the proposed online SVM-based GPC method performs the modelling and control tasks simultaneously. At each iteration, if the SVM model is not accurate enough to represent the plant dynamics at the current operating point, it is updated with the training data formed by persistently exciting random input signal applied to the plant, otherwise, if the model is accepted as accurate, a generalized predictive control signal based on the obtained SVM model is applied to the plant. After a short transient time, the model can satisfactorily reflect the behaviour of the plant in the whole phase space or operation region. The incremental algorithm of AOSVR enables the SVM model to learn the new training data pair, while the decremental algorithm allows the SVM model to forget the oldest training point. Thus, the SVM model can adapt the changes in the plant and also in the operating conditions. The simulation results on non-linear systems have revealed that the proposed method provides an excellent control quality. Furthermore, it maintains its performance when a measurement noise is added to the output of the underlying system.

show abstract

“…Ungar [31] defines a Bioreactor Benchmark Problem that excellently fits in the context. The state of the process is described by two dimensionless variables named the cell mass denoted by c 1 (t) and the amount of nutrients denoted by c 2 (t) (see Ungar [31]; Puskorius and Feldkamp [23] and Efe et al [8]). The units for both variables are the normalized concentrations, i.e.…”

Section: Introductionmentioning

confidence: 99%

“…Efe et al [8] consider this problem for developing a nonlinear control law forcing the process states to those of a first order linear one. A MLP structure is used to build the nonlinear function in the control law and the plant is forced to follow a reference model.…”

Section: Introductionmentioning

confidence: 99%

A comparison of networked approximators in parallel mode identification of a bioreactor

Efe

2010

Advances in Engineering Software

View full text Add to dashboard Cite

A novel analysis and design of a neural network assisted nonlinear controller for a bioreactor

Cited by 6 publications

References 0 publications

Novel Neuronal Activation Functions for Feedforward Neural Networks

Novel Neuronal Activation Functions for Feedforward Neural Networks

Online trained support vector machines-based generalized predictive control of non-linear systems

A comparison of networked approximators in parallel mode identification of a bioreactor

Contact Info

Product

Resources

About