An electronic converter is required to convert DC to AC energy . VSC is used to interconnect generation system with AC network. Now VSCis one of the best converter because it has modern power semiconductor advantages as Turn-off (GTO) and (IGBT) [1-2].
ABSTRACT- Now a day’s FACTS devices are used to control the flow of power, to increase the transmission capacity and to improve the stability of the power system. One of the most commonly used FACTS devices is Unified Power Flow Controller (UPFC)In this paper, a modulation and control method for the new transformer less unified power flow controller (UPFC) is presented. To overcome the problems with transformers, a completely transformer less UPFC based on an innovative configuration of two cascade multilevel inverters has been proposed. The new UPFC offers several advantages over the traditional technology, such as transformer less, light weight, high efficiency, low cost and fast dynamic response. This paper focuses on the modulation and control for this new transformer less UPFC, including artificial neural networks(ANN) controlling for low total harmonic distortion and high efficiency, independent active and reactive power control over the transmission line, dc-link voltage balance control, etc. Both the steady-state and dynamic response results will be shown in this paper. UPFC on controlling the flow of power and the effectiveness of controllers on the performance of UPFC is used to simulate UPFC model and to create the ANN.
14 Read more
performance of the model. Our aim is to generate a control 3 phase current that can be fed to a voltage source inverter for the control of the induction motor. The reference speed is given as input and actual speed is taken as feedback from the motor. The speed controller converts it to a control torque which in turn generates the quadrature axis current control component Iqs. Similarly the flux controller provides the direct axis current control component Ids. Also the rotor angle theta is obtained using these values. Now using inverse Clark’s transformation a control current Iabc(ref) is generated and fed to a voltage source inverter which in turn generates a control voltage Vabc for the IM. The control schematics is given in Fig.1.
Abstract: An Artificial Neural Network is a well-known AI technique for replicating human brain and offering suitable solution for any unpredictable complicated problem. Taking the advantage of it, this research will analyse the applicability of Neural Network Controller for ship manoeuvring, such as course changing. To train the controller, optimized teaching data are used to keep the consistency in the data as it could enhance the learning ability of the controller while training. A double layered feed-forward neural network and back propagation method are found suitable for this purpose. Later-on, simulations are done to justify the effectiveness of the trained controller for unknown situations.
This paper describes the neural controller as analogous to PI controller for voltage regulation, performance of the DVR under different voltage disturbances. PI controller is difficult to tune for non linear systems; neural will be best choice for non linear systems. Dynamic Voltage Restorer (DVR) is series controller which has capability to mitigate the voltage disturbance by injecting missing voltage in series to the load. DVR comprises of inverter, DC energy storage and series transformer. The paper also presents modeling of DVR, and controller. Back-propagation concept is used for tuning an Artificial Neural Network controller as analogous to the PID controller to optimize the voltage regulation and load current disturbance. The performance of the neural based DVR is analyzed through different case studies. The circuit is simulated in matlab/simulink and results are presented to validate the proposed controller. Keywords: sag, swell, DVR, Neural controller
ABSTRACT : - Low Frequency Oscillations (LFO) are a frequent adverse phenomenon which increase the risk of instability for the power system and thus reduce the total and availability transfer capability (TTC and ATC).LFO occur in power systems because of lack of the damping torque in order to dominance to power system disturbances as change in mechanical input power. In the recent past Power System Stabilizer (PSS) was used to damp LFO. FACTs devices, such as Unified Power Flow Controller (UPFC), can control power flow and increase transient stability. So UPFC may be used to damp LFO instead of PSS. UPFC damps LFO through direct control of voltage and power. In this research the linearized model of synchronous machine (Heffron-Philips) connected to infinite bus (Single Machine-Infinite Bus: SMIB) with UPFC is used and also in order to damp LFO, adaptive ANN damping controller for UPFC is designed and simulated. Simulation is performed for various types of loads and for different disturbances. Simulation results demonstrate that the developed ANN damping controller would be more effective in damping electromechanical oscillations in comparison with the conventional lead-lag controller.
The purpose of this paper is to report preliminary research in developing a neural-network-based optimal control strategy for vector control of a grid-connected rectifier/inverter in renewable and electric power system applications. First, the transient and steady-state models of a GCC system in a d-q reference frame are presented in Section II. Section III discusses the limitations associated with the conventional standard GCC vector control method and a newer direct- current vector control mechanism. Section IV proposes a neural network based vector control structure. Section V explains how to employ dynamic programming to achieve optimal neural vector control for the GCC system. The performance of the proposed DP-based neural vector control scheme is evaluated in Section VI. Finally, the paper concludes with a summary of the main points.
Neural network based fraud detection methods are most popular. An interconnected group of artificial neurons is contained in artificial neural network. The functions of the brain especially associative memory and pattern recognition are responsible to motivate the principle of neural network. In neural network similar patterns are identified, future values or events are predicted which are based upon the associative memory of the patterns. Neural network has widely useful in classification and clustering. It has main advantage over other techniques: the neural network model is capable of learning from the past and thus, results can be improved as time passes. Also, the rules can be extracted in this model. Moreover the future activity can be predicted on the basis of the present situation. By utilizing neural networks, efficiently, it will become easy for the banks to detect fraudulent use of a card in faster and more efficient way. Amongst the study of credit card fraud that has been reported, it was observed that most have paid attention on using neural networks. Nonlinear statistical data modeling tools are neural networks. Through these networks, complicated input-output relations can be easily depicted. A neuro-fuzzy system uses a learning algorithm that is derived from neural network theory. This system determine parameters : Fuzzy set and fuzzy rules by processing data samples. This is a fuzzy based model developed in the theory of neural networks using a learning algorithm. The heuristic learning process is based on local data and only creates local changes in the underlying fuzzy structure. This can be regarded as a neural feedback network of three layers. The first layer is input variable, the center layer is hidden and it represents fuzzy rules
Abstract— Now a day’s electrical system is at intervals the tactic to convert into smart power system with interconnected national and regional grids. wattage system is growing, and quality in all sectors like generation, transmission, distribution and lading systems. Faults like tangency condition lands up in severe economic losses & reduces responsibility of the electrical system. These faults cause interruption to electrical flows, instrumentality damages & even cause death of humans, birds & animals. For avoiding these styles of things, we've got to clear or nullify the fault. Fault clearing is also an important task in facility network. It is often done by exploitation protecting devices like switch gears. Protection plays a very important role in fashionable facility network, right from generation through transmission to distribution end. Multi useful relay additionally put in in fashionable power grid network for cover of conductor, generator protection, motor protection, real time fault location, protection of bus bar and totally different necessary equipment’s. A reliable, continuous give of electricity is very important for functioning of today’s modern sophisticated advanced society. it's usually obtained by providing protection against to the faults in grid. This paper is devoted to abnormal system behaviour below conditions of faults in power transmission lines exploitation MATLAB Simulation and planned for fault detection, classification & location by exploitation ANN (Artificial Neural Networks).
Experimentally, the neural networks are used to model the human cardiovascular system. By building a model of the cardiovascular system of an individual and then comparing with the real time physiological measurements, such as : Heart rate, Blood pressure, Blood sugar, Cholesterol etc.,taking from the patients, we can make an early prediction of the disease . If the model is found to be adapted to an individual, then it becomes a model of that individual. Diagnosis of heart disease is a difficult and tedious task in medical field. In general, all the doctors are predicting heart disease by learning and experience.
The proposed structured of the survey is a three layer architecture. First is the input layer which receives input from the segmented character images of standard size. Second layer is a hidden layer, this layer is use to train the neural network for specific font styles in case of character recognition process. The final layer is the output layer, this layer is used to generate Unicode values for different characters which itself worked as a matching criteria.
Hewit and Burdess firstly presented the AFC which established on Newton’s second law of movement . In this control strategy, the actuated force/torque is measured by sensor or other techniques. Also, the desired torque/force calculated from acceleration of output by multiplying to estimated mass/inertia. The difference between these two torques/forces is delivered to the actuator to compensate. Thus, enough torque/force is generated by actuator which adjusted by controller. In fact, this difference is estimated disturbance which it should be compensated and can be mentioned as bellow:
12 Read more
ABSTRACT : In this design, we purpose to design a single inductor dual output implementation using flay back converter for power factor correction. By which the regulated output will obtain through each output, here we study between the PID controller and the artificial neural network which is better to improve the power factor. Compared with two stages multiple outputs, SIDO PFC using fly back converter with the ANN controller has some benefits that provide the reduced undershooting, overshooting, harmonics and oscillation (i.e. total harmonic distortion (THD). Main purpose to design this circuit improving the power factor from 0.933 to 0.994 is can be assume in our study because ANN controller provides less settling time, that also small size, light weight, cost saving. This project is shows that at the same value the ANN controller is more accurate. Here also AC input providing and dual DC output, there is no need of converter. We can use motor instead of resistive load to conclude that our purposed converter is also suitable to DC drive or DC motor. The MATLAB Simulation is using for designing purpose.
3) Fully-connected layer: fully-connected surface shows that every neuron in the preceding surface is associated to all the neurons in the current surface. The quantity of classes is resolved by the total figure of fully-connected neuron in last surface, Figure.7 impart a diagrammatic depiction of fully-connected surface. The neurons are all associated and all links has a definite weight. This surface substantiates all outputs of the preceding surface to discover a certain target output. A leaky rectifier linear unit  is practiced as an activation function following the convolution surface. The motive is to chart the output to the input group and instigate non-linearity along with sparsity to the network. The CNN instructed with backpropagation  and the hyper- parameters may be adjusted for most favorable instruction accomplishment. Apart from CNN there are some more deep learning architectures, for example deep generative models   and recurrent Neural Network (RNN) etc. are monitoring the physiological signals equally. A deep generative model has two general structure deep belief network  and restricted Boltzmann machine , in short it is written DBN and RBM. The RBM is built up of a two-surface neural net accompanied by one perceptible and one concealed surface. In opposite to the feed-forward network. The RNN appoints a recurrent approach broadly recurrent network by which the network accomplishes a schedule job accompanied by the output existence determined by the preceding computation. The most prevalent sort of RNN is the long short-time memory network . The LSTM algorithm subsumes a memory block accompanied by three gates: the input, output, and forget gate. These gates influence the cell condition decision is taken to add or remove data-info from the network. For each input the process replicate itself. The architecture of deep learning verifies their potential with exceptional interpretation of conventional machine leaning procedure . Further deep learning algorithms reduce the desire for feature engineering.
ABSTRACT: In this paper, we proposed a methods of implementation of intelligent controller for speed control of an induction motor using indirect vector control method has been analyzed in detail. Induction motor is used in many industrial applications of the total used electrical energy. This paper proposes a new control scheme based on artificial neural networks to obtain certain torque and speed operating point. The combine performances of PI speed controller and ANN is used with indirect VOC. Due to its simplicity of designing and construction this method is most effective
Back-propagation (BP) algorithms are the most popular training algorithms that are widely used due to their simplicity and the application for training FF-NN (Kulluk, 2013). In FF-BP networks, which are considered in this study, output error is reported back, and in this way, a more desirable output is acquired through updating the weighting coefficients matrix. This action is carried out until the error between the target data and output data derived from the weighting matrix is insignificant and consequently the value of the objective function is minimized (Fig.4). For further details on FF-NNs, the reader is referred to the bibliography (ASCE Task Committee on Application of Artificial Neural Networks in Hydrology, 2000).
13 Read more
Mohammad Sarchami and Mehdi Eftekhari  forecasting per share earnings in investments is very im- portant because it is a significant factor in methods of stock evaluation; and in most of these cases, it is a fun- damental factor in investing in the stock market. In order to forecast per share earnings using an “artificial neural network with an error backward propagation algorithm” and an “artificial neural network with a genetic algo- rithm”, 61 firms in 7 financial years, from the beginning of 1381 until the end of 1387, with 9 variables (8 input variables and 1 output variable) were chosen; from which 3843 (61 × 7 × 9) data points were extracted. The hy- potheses are based on the idea that: 1) an artificial neural network with an error backward propagation algorithm is able to forecast the earnings of per share; 2) a neural net- work with a genetic algorithm is able to forecast the earnings per share; and 3) the neural network with the error backward propagation algorithm has less error in forecasting the earnings per share than the neural net- work with the genetic algorithm. To test the hypotheses, the results confirm all hypotheses.
The user has a few other options as far as what type of network to simulate. The first option is the ideal neural network, which is a neural network on a PC using standard IEEE 754 floating point precision. This allows the user to compare the quality of the trained network to that of the training patterns before any error from the microcontroller is introduced. The user can simulate the error produced by the microcontroller by selecting the simulation button. This then compares the simulated network to the ideal network. The simulator engine discussed previously uses a configurable number of patters for testing. At this point any possible overflows or other errors should be caught before hardware is introduced.
ABSTRACT: Chaos and chaos control are new theories and new fields of nonlinear dynamics. Chaotic motion is a complex irregular behavior produced by nonlinear dynamical systems, which is prevalent in various fields of nature. This paper introduces the development of chaos, the development of chaos and the development of chaos control, and summarizes the different strategies of chaos control. This paper introduces the idea, principle and characteristics of OGY method, introduces the adaptive control method, continuous feedback control method and neural network method, and puts forward some opinions on the possible difficulties in the future, and points out the application prospect and research direction of chaos control.
sists of seven neurones, and output layer of one neurons as shown in Figure 3. The activation function used in this work is “logsig” for hidden layer, and “purelin” for output layer. The NN is trained using a back propagation with Levenberg-Marquardt algorithm. The Back propa- gation is a form of supervised learning for multi-layer nets. Error data at the output layer is back propagated to earlier ones, allowing incoming weights to these layers to be updated. It is most often used as training algorithm in current neural network applications. Figure 4 presents the mean square error between the network output and the target. The network response analysis is depicted in Figure 5. As shown in the figure the regression “R” equal one which mean the output track the target in a correct way.