Neural Network Based Power System Research Articles

Convolutional neural network-based power system transient stability assessment and instability mode prediction

Online transient stability assessment (TSA) is vital for power system control as it provides the basis for operators to decide emergency control actions. But none of previous TSA research has taken into consideration the difference between two instability modes (aperiodic instability and oscillatory instability), which may threaten secure operation of power system. To address this problem, a TSA and instability mode prediction method based on convolutional neural network is proposed. The method takes the bus voltage phasor sampled by phasor measurement units (PMUs) during a short observation window after disturbance as input, and outputs the prediction result promptly: stable, aperiodic unstable or oscillatory unstable. The end-to-end model automatically extracts needed features from the raw measurement data, thus freeing itself from reliance on expertise. At the offline training stage, stochastic gradient descent with warm restart (SGDR) optimization algorithm is employed so that the model tends to converge to 'flat' and 'wide' minima with better generalization ability. Case studies conducted on New England 39-bus system and Western Electricity Coordinating Council (WECC) 179-bus system demonstrate superior accuracy, adaptability and scalability of the proposed method compared with conventional machine learning methods. Furthermore, the proposed model is empirically proven to be robust to PMU noise and loss.

Neural network-based power system dynamic state estimation using hybrid data from SCADA and phasor measurement units

Summary This paper points out the application of artificial neural network for short-term load forecasting where the projected loads are utilized to define a discrete-time state transition model (i.e., process model). The model is applied to estimate states dynamically and to generate pseudo measurements. Weights of neural network are not treated static and would be carried out under reevaluation alongside the estimation of state vector dynamically. The unscented Kalman filter estimation approach, which requires less approximation of power system, is used in the proposed method. The unscented Kalman filter is implemented through a dual structure due to the interactions of the state vector and the dynamic model of power system. The performance of the proposed method from accuracy prospective is compared with a couple of widely used methods. An optimum solution for wide-area monitoring system would be realized through implementation of more realistic process model along the simplicity of the proposed method and its capability to handle hybrid measurement data from SCADA and phasor measurement units.

Artificial Neural Network Based Power System Stabilizer on a Single Machine Infinite Bus System Modelled in Digsilent Powerfactory and Matlab

Artificial Neural Network Based Power System Stabilizer on a Single Machine Infinite Bus System Modelled in Digsilent Powerfactory and Matlab

Layer Recurrent Neural Network based Power System Load Forecasting

This paper presents a straight forward application of Layer Recurrent Neural Network (LRNN) to predict the load of a large distribution network. Short term load forecasting provides important information about the system’s load pattern, which is a premier requirement in planning periodical operations and facility expansion. Approximation of data patterns for forecasting is not an easy task to perform. In past, various approaches have been applied for forecasting. In this work application of LRNN is explored. The results of proposed architecture are compared with other conventional topologies of neural networks on the basis of Root Mean Square of Error (RMSE), Mean Absolute Percentage Error (MAPE) and Mean Absolute Error (MAE). It is observed that the results obtained from LRNN are comparatively more significant.

Simulation of Artificial Neural Network Based Power System Stabilizer using Digsilent Power Factory and Matlab Interface

Simulation of Artificial Neural Network Based Power System Stabilizer using Digsilent Power Factory and Matlab Interface

Simulation Analysis of a Power System Protection using Artificial Neural Network

There has been significant development in the area of neural network based power system protection in the previous decade. Neural network technology has been applied for various protective relaying functions including distance protection. The aim of this Paper is to develop a software module acting as a protective relay using neural network techniques. The Artificial Neural Network (ANN) software developed module employs the back-propagation method to recognize the waveform patterns of impedance in a transmission line. The input waveforms are generated using PSCAD. The generated waveforms then are used as training and testing data for the ANN software. The ANN software is simulated using the Neural Network Toolbox. The design has been tested for different fault conditions including different fault resistances and fault inception angles. The test results show that the relay is able to detect faults in lesser time as compared to conventional relay algorithms. DOI: http://dx.doi.org/10.11591/ijece.v3i1.1938

Design of an adaptive neural network based power system stabilizer

Power system stabilizers (PSS) are used to generate supplementary control signals for the excitation system in order to damp the low frequency power system oscillations. To overcome the drawbacks of conventional PSS (CPSS), numerous techniques have been proposed in the literature. Based on the analysis of existing techniques, this paper presents an indirect adaptive neural network based power system stabilizer (IDNC) design. The proposed IDNC consists of a neuro-controller, which is used to generate a supplementary control signal to the excitation system, and a neuro-identifier, which is used to model the dynamics of the power system and to adapt the neuro-controller parameters. The proposed method has the features of a simple structure, adaptivity and fast response. The proposed IDNC is evaluated on a single machine infinite bus power system under different operating conditions and disturbances to demonstrate its effectiveness and robustness.

ANN Based Power System Stabilizers for Large Synchronous Generators

This paper presents an artificial neural network based power system stabilizer (ANNPSS) for excitation control for a large synchronous generator. The generator operates over a wide range of operating conditions and is subjected to different types of disturbances and emergency states. A 300 MW turbogenerator is used to generate appropriate training data for the ANN controller. Off-line simulations using a suitable conventional PSS to control the generator for different working conditions are used to generate training input-output pairs. A multi-layered back propagation (BP) ANN is utilised to design the ANN based controller where speed error deviation and the incremental change in machine terminal voltage are fed to the ANNPSS controller. The proposed ANNPSS and conventional PSS are compared and test results indicate that the ANN based controller is more adaptive and flexible than conventional stabilizers and show a good performance over a wide range of operating conditions and disturbances.

Discussion of "A neural network based power system stabilizer suitable for on-line training-a practical case study for EGAT system" [and reply

S.N. Singh et al. comment on the paper by B. Changaroon et al. (see ibid., vol.15, no.1. p.103-9, 2000). They ask for the original authors' opinions on five points. The original authors reply to the comments.

A neural network based power system stabilizer suitable for on-line training-a practical case study for EGAT system

This paper presents the development of a neural network based power system stabilizer (PSS) designed to enhance the damping characteristics of a practical power system network representing a part of Electricity Generating Authority of Thailand (EGAT) system. The proposed PSS consists of a neuro-identifier and a neuro-controller which have been developed based on functional link network (FLN) model. A recursive on-line training algorithm has been utilized to train the two neural networks. Simulation results have been obtained under various operating conditions and severe disturbance cases which show that the proposed neuro-PSS can provide a better damping to the local as well as interarea modes of oscillations as compared to a conventional PSS.

Neural network based power system damping controller for SVC

The development of a neural network based power system damping controller (PSDC) for a static Var compensator (SVC), designed to enhance the damping characteristics of a power system network representing a part of the Electricity Generating Authority of Thailand (EGAT) system is presented. The proposed stabilising controller scheme of the SVC consists of a neuro-identifier and a neuro-controller which have been developed based on a functional link network (FLN) model. A recursive online training algorithm has been utilised to train the two networks. The simulation results have been obtained under various operating conditions and disturbance cases to show that the proposed stabilising controller can provide a better damping to the low frequency oscillations, as compared to the conventional controllers. The effectiveness of the proposed stabilising controller has also been compared with a conventional power system stabiliser provided in the generator excitation system.

97/02890 Speed deviation driven adaptive neural network based power system stabilizer

97/02890 Speed deviation driven adaptive neural network based power system stabilizer

Speed deviation driven adaptive neural network based power system stabilizer

The paper presents an online adaptive artificial neural network (ANN) based power system stabilizer (PSS). The proposed controller is first trained offline using a pole placement based state feedback gain technique at different operating points. The trained ANN parameters (weights and biases) are updated and tuned online using the speed deviation as the reinforcement signal. The proposed PSS is tested at different operating conditions and a variety of regulator gains. The digital results validate the effectiveness and reliability of the new PSS in terms of fast system response under different loading conditions compared with the conventional PI controller and the modern control theory approach of pole placement.

A neural network-based power system stabilizer using power flow characteristics

A neural network-based power system stabilizer (neuro-PSS) is designed for a generator connected to a multi-machine power system utilizing the nonlinear power flow dynamics. The use of power flow dynamics provides a PSS for a wide range of operation with reduced size neural networks. The neuro-PSS consists of two neural networks: neuro-identifier and neuro-controller. The low-frequency oscillation is modeled by the neuro-identifier using the power flow dynamics, then a generalized backpropagation-through-time (GBTT) algorithm is developed to train the neuro-controller. The simulation results show that the neuro-PSS designed in this paper performs well with good damping in a wide operation range compared with the conventional PSS.