The IJPEAI Journal
|
|
| Home Volume 1 Volume 2 Volume 3 ISSN: 1985-6431 | |
| Volume > Volume 2, Number 1, March 2009 | |
|
The IJPEAI Journal |
|
|
Board of International Reviewers:
|
|||
|
| |
|
Performance Analysis of Two-variate ANN Models for Predicting the Output Power from Grid- connected Photovoltaic System
S. I. Sulaiman, T. K. A. Rahman, I. Musirin, S. Shaari | |
|
Abstract:
This paper presents the prediction of total AC power output from a grid-connected photovoltaic (PV) system using two-variate artificial neural network (ANN) models. In this study, multi-layer feedforward ANN models for the prediction of total AC power output from a grid-connected PV system has been considered. Three ANN models were developed based on different sets of two-variate ANN inputs. The first ANN model utilizes solar radiation and ambient temperature as its inputs while the second model uses solar radiation and wind speed as its inputs. On the other hand, the third model uses solar radiation and PV module temperature as its inputs. However, all the three models utilize similar type of output which is the total AC power produced from the grid-connected PV system. Data filtering process was introduced to choose quality data patterns to be processed during training. Thus, only informative features were available for the prediction. In addition, the performance of each ANN model was characterized by the correlation coefficient (R) and root mean square error (RMSE) of the prediction. After training process was completed, testing process was performed to decide whether the training process should be repeated or stopped. Besides selecting the best prediction model, this study also exhibits some of the experimental results which illustrate the effectiveness of the data filtering in predicting the total AC power output from a grid-PV system. Fully trained ANN models are expected to be able to predict the AC power output from a set of un-seen data patterns in the future.
Index Terms: Artificial neural network (ANN), photovoltaic (PV), correlation coefficient (R), root mean square error (RMSE), prediction.
|
|
Optimal Transformer Tap Changer Setting for Voltage Stability Improvement | |
|
Abstract: This paper presents Ant Colony Optimization (ACO) technique for optimal transformer tap changer setting (OTTCS) in order to improve voltage stability condition along with transmission loss and voltage profile monitoring. ACO is a new cooperative agent’s approach, which is inspired by the observation of the behaviours of real ant colonies on the topics of ant trial formation and foraging method. The set of cooperating agents called “ant” cooperate to find the optimal point of OTTCS. Comparative studies presented with respect to Evolutionary Programming (EP) and Artificial Immune System (AIS) had indicated the merit of the proposed technique. All of the algorithms are programmed on MATLAB applied to the IEEE 30-bus Reliability Test System (RTS).
Index Terms: Ant colony optimization, evolutionary programming, artificial immune system, optimal transformer tap changer setting, voltage stability improvement |
|
The Application of Box-Jenkins Models in Short-Term Load Forecasting M. M. Othman, K. A. Abd Rahman, I. Musirin, A. Mohamed, A. Hussain |
|
Abstract: Short-term load forecasting (STLF) plays an important role in obtaining secure and economic operations of electric utilities in a deregulated power system. This paper presents the autoregressive (AR) Box-Jenkins model that used to perform the STLF of the Malaysian hourly peak loads. The AR Box-Jenkins model was selected based on the behaviors of the sample autocorrelation (SAC) and sample partial autocorrelation (SPAC) functions of the time series. Comparison in terms of accuracy in estimating the STLF has been made between the AR and autoregressive integrated moving average (ARIMA) Box-Jenkins models. The results have shown that the AR Box-Jenkins model is robust in forecasting the Malaysian hourly peak load for the next 24 hours with less error. Index Terms: Short-term load forecasting, autoregressive Box-Jenkins model, sample autocorrelation function and sample partial autocorrelation function. |
Fuzzy Logic Unit Commitment based on Load Forecasting using ANN and Hybrid Method U.BasaranFilik , M.Kurban |
|
Abstract : In this study, unit commitment (UC) problem for four-unit Tuncbilek thermal plant which is in Kutahya region in Turkey, is solved for an optimum schedule of generating units based on the load data forecasted by using the conventional Artificial Neural Network (ANN) and an improved hybrid method, ANN model with Weighted Frequency Bin Blocks (WFBB). Fuzzy Logic (FL) method is used for solving the UC problem. Since under forecasting results in the requirement of purchasing power from spot market or over forecasting brings about an unnecessary commitment of generating units, an accurate load forecasting is the first step to solve the UC problem. Total costs calculated for both actual and forecasting load data are compared. The data used in the analysis was taken from Turkish Electric Power Company and Electricity Generation Company. All the analyses are implemented using MATLAB. Index Terms : Fuzzy logic, unit commitment, load forecasting, artificial neural network, weighted frequency bin blocks. |
Wavelet Transform-Signal Processing for Fault Detection in 500 kV EHV Transmission Lines of Power System Using Genetic Algorithm C. Jaipradidtham |
|
|
Abstract : This paper presents the use of wavelet transform for analyzing power system fault detection in order to determine the fault location and the fault phase selection in 500 kV double circuit transmission lines of type DL3° and DT20° using genetic algorithms. The wavelet transform (WT)has been successfully applied in many fields. The technique is based on using the absolute sum value of coefficients in multiresolution signal decomposition based on the discrete wavelet transform(DWT). By per-form the simulation of fault signal and parameter adjustments, with have effect wit ATP/EMTP program. The result of experiment shows that can indicate the specify fault location, fault classification, phase selection and fault circuit transmission line. The simulation of the 500 kV power system using EMTP program were used to test the performance of the simple genetic algorithms. Index Terms :
|
Risk-constrained Optimal Bidding Strategies for Generation Companies using Differential Evolution R. Rajathy, R. Gnanadass, K. Manivannan, Harish Kumar |
|
Abstract : The emerging electricity market behaves more like an oligopoly than a perfectly competitive market. The profit of each supplier is influenced to varying extents by differences in the degree of imperfection of knowledge of their rivals. This paper discusses the optimal bidding strategies of Generating companies (Gencos) solved using Differential Evolution (DE) algorithm for the first time. It is assumed that each Genco bids a linear supply function, and chooses the coefficient in the linear supply function to maximize their benefits, subject to expectations about how the rivals will bid. A normal probability distribution function (pdf) is used to describe the bidding behaviors of rivals and the problem of building optimal bidding strategies for Gencos is formulated as stochastic optimization problem. The proposed algorithm is tested for six generating companies with different risk coefficients and load price elasticity factor. The obtained results are compared with those obtained by Reference [14]. Index Terms :
|
|
Tuning of PI Controller for Current Source Inverter Fed Induction Motor Drive P. Kumar, V. Agarwal |
|
|
Abstract: Direct torque control (DTC) of Induction motor is preferred as compared to vector control scheme due to its quick torque response, simplicity and robustness against rotor parameters variation. PID controllers are very common since they can offer a satisfactory performance over a wide range of operation. The main problem with this controller is the correct choice of the PID gains and the fact that by using fixed gains, the controller may not provide the required control performance, when there are variations in the plant parameters and operating conditions. Therefore, a tuning process must be performed to ensure that the controller can deal with the variations in the plant. The tuning of these controllers is governed by system nonlinearities and continuous parameter variations. In this paper, a complete and rigorous study is made for tuning of PI controller used in a speed control loop in a Direct Torque Control (DTC) scheme applied in a current source inverter (CSI) fed induction motor drive system. The controller value is adjusted by Ziegler and Nichols method. A comparative study is made between P and PI controller. It has been found that with P controller the transient time to reach the steady state value is small. Index Terms: Direct torque control, current source inverter, Ziegler and Nichols |
Fuzzy Logic Application in Evaluation of DGA Interpretation Methods N.A. Muhamad, B.T. Phung and T.R. Blackburn |
|
Abstract: Dissolved gas-in-oil analysis (DGA) is one of the most useful techniques to detect the incipient faults in large oil-filled transformers. Various methods have been developed to interpret DGA results. Among them are the Key Gas, Rogers Ratio, Logarithmic Nomograph, Doernenburg, IEC Ratio and Duval Triangle. This paper used the DGA data from 69 different cases to test the accuracy and consistency of these methods in interpreting the transformer condition. The key gases considered for evaluation are hydrogen, methane, ethane, ethylene and acetylene. MATLAB programs with and without using Fuzzy logic were developed to automate the evaluation of each method. The difference on accuracy and consistency of each method using and not using Fuzzy logic is presented. Index Terms: DGA interpretation method, Fuzzy Logic, fault gases. |
|
An Adaptive Distance Relaying Strategy Based On Global Network Simulation V. Gohari Sadr, Sh. M. Kuhsari
|
|
|
Abstract: In this paper, an adaptive distance relaying strategy based on Global Network Simulation (GNS) concept is presented. The GNS concept is actually a Distributed Simulation Approach (DSA) for piecewise analysis of large-scale power grids using Diakoptics and Large Change Sensitivity (LCS) concepts. The proposed strategy employs the decentralized nature of GNS short circuit analysis for calculation of relay settings of each sub-network
independently in the federal environment. Updating the relay settings of each utility in every power system
conditions adaptively based on local database and computational resources as well as meeting individual utilities'
adopted coordination philosophies are great achievements of presented approach. No computational errors
encountered in the proposed strategy since the applied concepts eliminates the usual needs for equivalent network
or reduced models; hence presenting an accurate and secure federative approach on geographically decomposed Index Terms: Adaptive Distance Relaying, Diakoptics, Distributed Simulation Approach (DSA), Global Network Simulation (GNS), Large Change Sensitivity (LCS), Mutually Coupled Lines. |
