Neural network algorithm for parameter identification of dynamical systems involving time delays

Yadaiah, N.; Deekshatulu, B. L.; Sivakumar, L.; Rao, V. Sree Hari

doi:10.1016/j.asoc.2006.06.007

Cited by 20 publications

(7 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…,  (20) where NM m denotes the B-spline network which is used to calculate k p and k i ; w ρ is the corresponding weighting factor; m = 1,2,3 number of PI controllers. Fig.…”

Section: Resultsmentioning

confidence: 99%

“…[10] adaptive PID gains for each controller to achieve satisfactory performance is proposed. The use of neural networks to model chaotic systems and nonlinear identification problems has attracted considerable attention in recent years [20]. Neural networks are originally inspired by the functionality of biological neural networks, which can learn complex functional relations based on a limited number of training data.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Adaptive PI Controllers for Doubly Fed Induction Generator using B-spline Artificial Neural Networks

Tapia-o.¹,

Aguilar²,

Abel³

et al. 2013

IJCA

View full text Add to dashboard Cite

This paper presents the design and simulation of adaptive PI controllers for doubly fed induction generators using b-spline neural networks. The control structure is based on a back-toback arrangement where the interest variables are regulated by PI linear controllers. Also, to deal with the nonlinear and uncertain system conditions, we proposed that the control parameters are updated online. The main task is that the power converters operation adapt by itself during the grid changes. Then, the basic problem consists of tuning the PI controllers simultaneously when the system and load are subjected to disturbances. The applicability of the proposal is demonstrated by simulation in a three-node grid, where one end is an infinite bus and the other connects the wind system, between them there are two transmission lines. The results show that the proposed controllers' tuning is comparable to that obtained by a conventional design, without requiring of a detailed model, which enable optimal speed tracking for maximum energy capture from the wind.

show abstract

“…,  (20) where NM m denotes the B-spline network which is used to calculate k p and k i ; w ρ is the corresponding weighting factor; m = 1,2,3 number of PI controllers. Fig.…”

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Adaptive PI Controllers for Doubly Fed Induction Generator using B-spline Artificial Neural Networks

Tapia-o.¹,

Aguilar²,

Abel³

et al. 2013

IJCA

View full text Add to dashboard Cite

show abstract

“…For the conventional linear rule, each component of the potential simply is a quadratic polynomial 1 2 e 2 i . However, for the nonlinear rule, the corresponding potential has a relatively flatter well due to the boundedness and nonlinearity of each function g i (e i ) in (6). Besides, the adaptive rules for the coupling terms, which could be regarded as the friction force acting on the particle, are invariant in design of the response system (3).…”

Section: Next It Is To Prove That Any Trajectory [ Y(t) B(t) (T) mentioning

confidence: 99%

“…Investigations of the second question have generated enormous literature on the technique of parameters identification in various dynamical models. Representative examples include: least-squares fitting algorithm was developed for parameters identification in the Hodgkin-Huxley model [1]; evolution strategy with a fitness function was utilized to estimate those parameters in the Purkinje Cells model [2,5]; neural networks learning technique was implemented to identify the accurate form of models with or without time delays [6]; extended Kalman filter and the unscented Kalman filter were applied to estimate the reaction rate in biochemical networks [7,8]. Apart from techniques used in parameters identification for models with known vector forms, rational or polynomial L 2 approximation and successive derivatives as embedding coordinates are managed to reconstruct a standard system from scalar observable time series in some typical systems [9,10] and some experimental cases [11] where the vector forms are presumably unknown.…”

Section: Introductionmentioning

confidence: 99%

Nonlinear adaptive synchronization rule for identification of a large amount of parameters in dynamical models

Lin

2009

Physics Letters A

View full text Add to dashboard Cite

“…The application of ANN makes possible to reduce considerably the laboratory experiment time while networks learn how to predict properties of insulation for duration longer than those of the tests thus constituting a tool making more economic the tests of high voltage in general. ANN method is more accurately applied to Dissolved Gas Analysis since the hidden relationships between fault types and dissolved gases can be recognized by ANN through training process [12][13][14][15].…”

Section: Artificial Neural Network Application To Dgamentioning

confidence: 99%

Review of Artificial Intelligence Techniques Application to Dissolved Gas Analysis on Power Transformer

Sharma¹,

Tiwari²,

Sood³

2011

IJCEE

View full text Add to dashboard Cite

Abstract-Transformers are a critical part of an electrical utility's asset base. On-line monitoring and diagnostics is a useful tool to help operators to manage their assets and make decisions on continuing operation, maintenance or replacement. Dissolved Gas Analysis (DGA) is the heart of on-line monitoring as it is a well-established method of transformer diagnosis. DGA techniques are simple, inexpensive, and widely used to interpret gases dissolved due to the deterioration of the insulating oil of power transformers and hence to diagnosis, possibility of various type of faults in power transformer. Various diagnostic criteria based on gas analysis have been developed. In this paper, the application of many AI techniques have been presented such as Artificial Neural Network (AAN), Fuzzy Interface System (FIS), Genetic Algorithm (GA), Extended Relation Function (ERF), Bayesian Network (BN), Self Organizing Map (SOM) and Discrete Wavelet Network (WNs) Transforms, which can be used to increase the efficient and accurate diagnosis for off line and on line monitoring of power transformers.Index Terms-Power Transformer, Dissolve Gas Analysis, Artificial Intelligence Techniques. I. INTRODUCTIONPower transformers play an important role in both the transmission and distribution of electrical power system. It is very essential oil-insulated components in power systems and its operational state determines the safety of the whole power system. Failure of a transformer may cause long interruptions in power supply and require expensive repairs. An incipient fault in a transformer should be detected as early as possible, preventing the transformer from further deterioration. Diagnostic approaches can be divided into two groups: on-line and off-line. Transformer ageing process of insulating oil and cellulose materials has been monitored by many different techniques such as Dissipation factor, capacitance, Breakdown voltage of oil and paper, Degree of polymerization (DP), Total combustible Gases (TCG), furan analysis, Interfacial Tension (IFT) analysis etc. Dissolved Gas Analysis is most widely method to detect incipient faults in oil filled transformer and Electrical equipment. Dissolved gas analysis is made on the basis of the standard IEC60599 standards [1]. Several diagnosis methods, such as the Rogers and Duval Triangle methods, are available to identify the different types of fault occurring in service such as arcing, partial discharges, or hot spots. DGA includes detection, quantification and characterization of the Gases. These gases, called characteristic gases, include, Hydrogen (H2), Methane (CH4), Ethane (C2H6), Ethylene (C2H4) and Acetylene (C2H2), and Carbon Oxides such as CO and CO2. The nature and the amount of the individual component gases extracted from the oil may be indicative of the type and degree of abnormality. The DGA data provides information about the condition of the transformer and advance warning of developing faults, monitoring the rate of fault development, confirm the presence of faults, conve...

show abstract

Neural network algorithm for parameter identification of dynamical systems involving time delays

Cited by 20 publications

References 12 publications

Adaptive PI Controllers for Doubly Fed Induction Generator using B-spline Artificial Neural Networks

Adaptive PI Controllers for Doubly Fed Induction Generator using B-spline Artificial Neural Networks

Nonlinear adaptive synchronization rule for identification of a large amount of parameters in dynamical models

Review of Artificial Intelligence Techniques Application to Dissolved Gas Analysis on Power Transformer

Contact Info

Product

Resources

About