Artificial Neural Network Controller Research Papers

Speed Control of Induction Motor Using Genetic Algorithm-based PI Controller

2025, Acta Polytechnica Hungarica

In this paper, we present the design of a Proportional Integral (PI) controller using Genetic Algorithm (GA) to control the speed of an induction motor (IM) using indirect field-oriented control method (IFOC). The main advantage of this... more

descriptionView Paper arrow_downwardDownload

Hybrid sliding neural network controller of a direct driven vertical axis wind turbine

by beei iaes

2024, Bulletin of Electrical Engineering and Informatics

This study aims to propose a robust hybrid sliding mode artificial neural network control (SM-ANN) scheme for controlling the stator power (active/reactive) of a doubly fed induction generator (DFIG)-based direct drive vertical axis wind... more

descriptionView Paper arrow_downwardDownload

A new MIMO ANFIS-PSO based NARMA-L2 controller for nonlinear dynamic systems

by Ali Jizany

2023, Engineering Applications of Artificial Intelligence

The proposal of this study is a new nonlinear autoregressive moving average, NARMA-L2 controller, which is based on an adaptive neuro-fuzzy inference system, ANFIS architecture. The new control configuration employs Sugeno-type fuzzy... more

descriptionView Paper arrow_downwardDownload

Hybrid sliding neural network controller of a direct driven vertical axis wind turbine

by Youcef Bakou

2023, Bulletin of Electrical Engineering and Informatics

This study aims to propose a robust hybrid sliding mode artificial neural network control (SM-ANN) scheme for controlling the stator power (active/reactive) of a doubly fed induction generator (DFIG)-based direct drive vertical axis wind... more

descriptionView Paper arrow_downwardDownload

Hybrid sliding neural network controller of a direct driven vertical axis wind turbine

by beei iaes and

2022, Bulletin of Electrical Engineering and Informatics

This study aims to propose a robust hybrid sliding mode artificial neural network control (SM-ANN) scheme for controlling the stator power (active/reactive) of a doubly fed induction generator (DFIG)-based direct drive vertical axis wind... more

Based on (7) and (8), the stator power can be given as (9): Starting from (4), the strong linkage between the stator and rotor components is appearing, which makes DFIG difficult to control, in order to simplify DFIG controlling. The field orientation technique was applied in literature [12], [30]. By orienting the stator flux ®, with the direct d axis and when taken into account the stator resistance, R, is ignored, we have (7) and (8) [32]: In (9) prove that the DFIG stator power can be regulated via the control of the currents rotor. The relationship between the rotor voltages and its current is given as (10):

Figure 2. ANN structure of stator power controller and its progression (a) active power and (b) reactive power a a a, a In our case, the training process used is that of the Levenberge-Marquardt algorithm, in order to determine the optimal synaptic weights. It’s an excellent optimization method due to its quick convergence properties and robustness. the structure of the ANN controller designed to replace the switching section of the SM regulator, for controlling the stator active power, consists of one linear input, 17 nodes in the hidden layer and, one neuron in the output layer as depicted in Figure 2(a), besides that, the stator reactive power loop controlled by an ANN controller with one linear input, 8 nodes in the hidden layer, and one neuron in the output layer showing in Figure 2(b), the curve of training, test and validation of the controller ANN of stator active power and stator reactive power are depicted in Figure 3(a) and Figure 3(b) respectively.

Figure 3. Performance curve of training (a) stator active power and (b) stator reactive power

Figure 5. Average daily wind speed values (April 2020) Figure 4. NEAL station in URER/MS Adrar The data used in this work are data from April 2020 taken from measurements of the weather stations called new energy Algeria (NEAL) as shown in Figure 4, placed in research unit in renewable energies in saharan medium (URER/MS), in Adrar region, which is a section of renewable energy development center (CDER), in Algeria capital. Knowing that the Adrar area in south Algeria has even greater potential for wind power. The yearly average wind speed is over 6 m/s [36]. The curve of average daily wind speed for April 2020 at Adrar zone. are displayed in Figure 5.

Table 1. Characteristics of the wind turbine system Figure 6. Schematic diagram of direct-drive vertical axis wind system based SM-ANN

Figure 7. DFIG stator power with PI, SMC and SM-ANN controllers (a) active power and (b) reactive power

Figure 8. DFIG one-phase stator current (a) I,a of SMC and (b) THD of SMC Furthermore, using the fast fourier transform (FFT) approach, the THD of stator currents are presented in Figure 8 and 9, as shown, the SM-ANN controller reduces THD down to 2.05% (Figure 9(b)), compared to SMC, where THD is 4.40% (Figure 8(b)), which helps to reduce the chattering phenomenon effect. The stator current signal quality of the SM-ANN controller has a good sinusoidal shape and, is less corrugated (Figure 9(a)) compared to SM controllers (Figure 8(a)).

Figure 9. DFIG one-phase stator current (a) Isa of SM-ANN and (b) THD of SM-ANN

Figure 10. Impact of resistances variation R;, R,: (a) active stator power and (b) reactive stator power

Figure 11. Impact of inductances variation L;, L,and Lyn: (a) active power and (b) reactive power

descriptionView Paper arrow_downwardDownload

Maximum Power Point Tracker for Photovoltaic Systems using On-line Learning Neural Networks

by Yassine ABDESSEMED

2022, International Journal of Computer Applications

In this paper a new maximum-power-point-tracking (MPPT) controller for a photovoltaic (PV) energy conversion system is proposed. Nowadays, PV generation is more and more used as a renewable energy source. However, its main drawback is... more

descriptionView Paper arrow_downwardDownload

A new MIMO ANFIS-PSO based NARMA-L2 controller for nonlinear dynamic systems

by Maysam Abbod

2022, Engineering Applications of Artificial Intelligence

The proposal of this study is a new nonlinear autoregressive moving average, NARMA-L2 controller, which is based on an adaptive neuro-fuzzy inference system, ANFIS architecture. The new control configuration employs Sugeno-type fuzzy... more

descriptionView Paper arrow_downwardDownload

A new MIMO ANFIS-PSO based NARMA-L2 controller for nonlinear dynamic systems

by Yousif Al-Dunainawi

2021, Engineering Applications of Artificial Intelligence

The proposal of this study is a new nonlinear autoregressive moving average, NARMA-L2 controller, which is based on an adaptive neuro-fuzzy inference system, ANFIS architecture. The new control configuration employs Sugeno-type fuzzy... more

Figure 1 shows the schematic diagram of a binary distillation column. The feed mixture is separated into two products; one is a distillate or overhead, and the other is the bottom product. Heat is supplied to the column via a reboiler, in order to vaporise the liquid in the base of the column. The vapour goes up through trays inside the column to reach the top. The vapour then liquefies in the condenser. Liquid from the condenser drops into the reflux drum. Finally, the some of the distillate product is removed from this drum as a pure product. The rest of the liquid is fed back near the top of the column as reflux, while another product is produced at the bottom.

Figure 2 Schematic diagram of ith stage of a binary distillation column The abbreviations, operation conditions and steady states of the column are in the appendix, 3 Process Identification

Therefore, many innovative solutions have been proposed to make an automatic selection of these parameters. One of the most efficient approaches is ‘adaptive neuro-fuzzy’ or ‘adaptive network-based fuzzy inference systems’ (ANFIS) [33] that uses, commonly, the hybrid backpropagation-least square error method to tune the parameters of FIS.

Figure 4 An ANFIS architecture of two inputs, four rules, and first order Sugeno model The main motivational aims of using ANFIS are the features that present the learning capabilities of both ANN and fuzzy inference systems. The learning algorithm adjusts the membership functions of a Sugeno fuzzy model, using input-output dataset. Figure 4 shows an example of the simplified ANFIS architecture, which contains two membership functions (A, B) for both inputs (x1, x2) and four rules as well as four membership functions of output (y). B) for both inputs (x1, x2) and four rules as well as four membership functions of output (y).

Figure 5 Schematic diagram of the proposed MIMO NARMA-L2 model

Figure 6 Figure 6 Simulation model of a binary distillation column

Figure 7 Inputs/outputs of the simulation of the distillation

It is clear that, as shown in Table 1, the FISs submodels are better optimised using PSO compared to the hybrid method with less RMSE. The visual comparison between the actual and predicted outputs for the both methods is shown in Figure 8 for training and testing cases. ANFIS tuned by PSO model shows adequately accurate predictions for distillate and bottoms compositions, while the ANFIS trained by the hybrid method lack of adequate prediction especially for the second output in both training and testing subsets.

Figure 8 Prediction model behaviour vs. actual outputs (distillate and bottoms compositions), training and testing subsets,

where Y;; and Y;2 are the set points or desired compositions of the column products. Figure 9 is a schematic diagram of the proposed controller.

Table 2 Comparison of performance index of the proposed controller vs. other controllers

descriptionView Paper arrow_downwardDownload

NERC’S APPROACH FOR MULTI-AREA DEREGULATED POWER SYSTEM UNDER DIFFERENT CONTRACT SCENARIOS

by IAEME Publication

2021, IAEME PUBLICATION

This paper deals with performance of fuzzy logic and artificial neural network on an electrical power system. A comparison of fuzzy controller and ANN controller based approaches shows the superiority of proposed ANN based approach over... more

This controller requires the least computation of the three architectures described in this chapter. The controller is simply a rearrangement of the neural network plant model, which is rained offline, in batch form. The only online computation is a forward pass through the neural network controller. The drawback of this method is that the plant must either be in companion form, or be capable of approximation by a companion from model [8]—[13]. The controller, however, requires a significant amount of on-line computation, since an optimization algorithm is performed at each sample time to compute the optimal control input. For the simulation of he two area interconnected power system this NARMA-L2 controller has been introduced for better dynamic responses. 4. MODEL REFERENCE CONTROL

constrain the outputs of a network (such as between 0 and 1), then the output layer should us¢ a sigmoid transfer function (such as logsig) [11]. In Neuron Model and Network Architectures for multiple-layer networks the number of the layers is used to determine the superscript on the weight matrices. This network can be used as a general function approximation as shown ir Fig. 3. It can approximate any function with a finite number of discontinuities, arbitrarily well given sufficient neurons in the hidden layer.

Figure 4 Diagram of Multi Input Neuron Model

Figure 6 Height membership function for (a) a crisp set A

Figure 7 Height membership function for (b) a Fuzzy set H

NERC’s Approach for Multi-area Deregulated Power System under Different Contract Scenarios

Piyush Kumar Yadav, Jaya Shukla, Rajnish Bhasker, Satyam Kumar Upadhyay 10. TYPES OF MEMBERSHIP FUNCTION 10.1. Triangular Membership Function

Since all information contained in a fuzzy set is described by its membership function, it useful to develop a lexicon of terms to describe various special features of this function. Fc purposes of simplicity, the functions shown in the following figures will all be continuous, bt the terms apply equally for both discrete and continuous fuzzy sets.

Figure 12 Max membership defuzzification method

Where { denotes an algebraic integration.

13.4. Mean — Max Membership (Middle of Maxima) 13.3. Weighted Average Method

13.5. Centre of Sums NERC’s Approach for Multi-area Deregulated Power System under Different Contract Scenarios

14. FUZZY ASSOCIATIVE MEMORIES (FAMS) Piyush Kumar Yadav, Jaya Shukla, Rajnish Bhasker, Satyam Kumar Upadhyay

NFIS architecture is a hybrid combination of neuro and fuzzy approaches and it is deemed adaptive network having zero synaptic weight. An Adaptive network having no synapt sight is also called as adaptive and non-adaptive nodes. An adaptive network can easi msform to neural network architecture with classical feed forward topology. The ANF ntroller works like an adaptive network simulator of Takagi—Sugeno type fuzzy controll NFIS controller is functionally equivalent to a fuzzy inference system (FIS) and its operatio > performed under Takagi— Sugeno type fuzzy controller. According to the input/output dé ts, ANFIS adjust the parameters using back propagation gradient descent and least squa oes of method for non-linear and linear parameters respectively. In this article, the ANF ntroller uses the back propagation through time algorithm, which is an advanced adapti ntrol configuration of ANN. The back-propagation technique is an iterative meth yploying the gradient decent algorithm for minimizing the minimum square error between t tual output and the objective for each pattern in the training. The ANFIS system consists > components of a conventional fuzzy system except that computation at every phase tried out by a layer of hidden neurons and the neural network’s learning capacity is offer boost the system knowledge.

Figure 18 Dynamic response of the test system under unilateral contract scenario. (a) Frequency deviations in area-1. (b) Frequency deviations in area-2. (c) Frequency deviations in area-3. (d) Frequency deviations in area-4 NERC’s Approach for Multi-area Deregulated Power System under Different Contract Scenarios

Figure 21 Dynamic response of the test system under unilateral contract scenario. (a) Tie-line power deviations between area-1&2. (b) Tie-line power deviations between area-1&3. (c) Tie-line power deviations between area-2&4. (d) Tie-line power deviations between area-2&3. (e) Tie-line power deviations between area-3&4. Piyush Kumar Yadav, Jaya Shukla, Rajnish Bhasker, Satyam Kumar Upadhyay

Figure 22. Excess raise/lower signals of Genco-2 with three different controllers. (a) Speed changer raise/lower signals of GENCO-2 with three different controllers. (b) Excess raise/lower signals of GENCO-?2 of PID controller over NERC standards-based ANFIS controller. (c) Excess raise/lower signals of GENCO-2 of ANFIS controller over NERC standards-based ANFIS controller.

descriptionView Paper arrow_downwardDownload