Accurate modeling of windspeedprofile is crucial as the windspeed dynamics are non-deterministic, having chaotic behavior and highly nonlinear in nature. Therefore, obtaining mathematical model of such windspeedprofile is rather difficult and vague. In this brief manuscript, the windspeed distribution in PeninsularMalaysia is modeled via the real-time wind data obtained from the Malaysian Meteorological Services (MMS). Artificialneuralnetwork (ANN) has been exploited to train the data such that the exact model of windspeed can be identified. The induced windspeedmodel worthwhile for control engineers to develop control apparatus for wind turbine systems at the selected area of studies. With the windspeed distribution profile, turbine output power can be analyzed and were discussed thoroughly.
This paper describes the application of principal component analysis (PCA) and artificialneuralnetwork (ANN) to pre- dict the air pollutant index (API) within the seven selected Malaysian air monitoring stations in the southern region of PeninsularMalaysia based on seven years database (2005-2011). Feed-forward ANN was used as a prediction method. The feed-forward ANN analysis demonstrated that the rotated principal component scores (RPCs) were the best input parameters to predict API. From the 4 RPCs, only 10 (CO, O 3 , PM 10 , NO 2 , CH 4 , NmHC, THC, wind direction, humidity
A sequential dataset is required for performing analysis and modelling processes. Missing values imputation can be accomplished using classical methods such as linear, nearest neighbour or others. The classical methods may not provide accurate imputations when the windspeed data contains nonlinearity. In this study, a hybrid AR-ANN model was proposed to reform the missing values problem by imputing missing values and jointly handling the nonlinearity problem. This method can be called hybrid AR-ANN model (Liu et al., 2012b) or ANN (Khashei and Bijari, 2010, 2011; Zhang, 2003). The missing values in windspeed data with nonlinear characteristic can be imputed more accurately using AR-ANN model.
This paper presents a novel method of skin diseases classification. The complete work is divided into four parts. First is preprocess the image then segment the image by using modified sobel edge detection technique, and extract the features of the segmented image, extracted features are sub divided in to sub space features and calssified the features by artificialNeuralNetwork(ANN). The performance of the different training algorithm has been investigated. Mean Square Error (MSE) is evaluated. Bayesian regularization backpropagation algorithm gives minimum MSE is 4.8561e- 13 and gradient is 1.6337e-08 at 190 epochs. Levenberg- Marquardt backpropagation algorithm provides MSE 1.0559e- 10 and gradient is 9.9001e-08 at 105 epochs. Resilient backpropagation algorithm 3.5354e-07 and gradient is 8.5468e-06 at 347 epochs. Scaled conjugate gradient backpropagation algorithm give MSE 0.02269 and gradient is 8.6124e-07 at 115 epochs.
The computer-based recognition of facial expression has received a lot of attention in recent years because the analysis of facial expression or behavior would be beneficial for different fields such as lawyers, the police, and security agents, who are interested in issues concerning dishonesty and attitude. The eventual goal in research area is the realization of intelligent and transparent communication between human being and machines. Some researchers used only four expressions for analysisusing computer and some used six expressions. We have used JAFFE database with seven expressions for analysis through the computer. Several facial expression recognition methods proposed in literature are for example Facial Action Coding System (FACS)   Facial Point Tracking  and Moment Invariant. Ekman and Friesen  developed the Facial Action Coding System (FACS) for describing expressions, such as anger, disgust, fear, happy, neutral, sad, and surprise. The Maja Pantic  did machine analysis of facial expressions. James J. Lien and Takeo Kanade have developed Automatic facial expression recognition system based on FACS Action Units in 1998 . They have used affine transformation for image normalization and facial feature point tracking method for feature extraction as well as Hidden Marko Model (HMM) used for
The field of fluid mechanics fluid mechanics and heat transfer were greatly benefited from the application of this tool. Almost every major experiment in this area was planned with its help. Using this principle modern experiments can substantially improve their working techniques and be made shorter requiring less time without loss of control.
An artificial neuron is a machine with various inputs and single output. The neuron consist two methods of functions; the training method and the using method. In the training method, the neurons can be taught to fire (or not), for fastidious input sample. In the using method when a taught input sample is identified at the input, their related outputs become the existing output. If the input sample does not belong in the taught record of input sample, the firing rule is used to determine whether to fire or not.
Abstract —Intrusion Detection System (IDS) is used to supervise all tricks which are running on particular machine or network. Also it will give you alert regarding to any attack. However now a day’s these alerts are very large in amount. It is very complicated to examine these attacks. We intend a time and space based alert analysis technique which can strap related alerts without surroundings knowledge and provide attack graph to help the administrator to understand the attack on host or network steps wise clearly and fittingly for analysis. A threat evaluation is given to discover out the most treacherous attack, which decrease administrator’s time and energy in calculating huge amount of alerts. We are analyzing the network traffic in form of attack using Entity Threat Evaluation (ETE) which find out which particular host is attacked, Gadget Threat Evaluation (GTE) which tells us within that host which device is attacked, Network Threat Evaluation (NTE) which tells us which network is attacked, Hit Threat Evaluation (HTE) by giving input as dataset of attack. Main idea is that the distribution of different types of attacks is not balanced. The attacks which are not repeatedly occurs, the learning sample size is too small as compared to high-frequent attacks. It makes ArtificialNeuralNetwork (ANN) not easy to become skilled at the characters of these attacks and therefore detection precision is much worse. To solve such troubles, we propose a new technique for ANN-based IDS, Fuzzy Clustering (FC-ANN), to enhance the detection precision for low-frequent attacks and detection stability.
Various methods of speed control of the induction motor had been employed such as the scalar (or the Volts/Hertz) control where the speed of the motor is controlled by scaling the voltage in proportion to the desired frequency. However, with the introduction of the Field Oriented Control system which improves the dynamics of the motor, various forms of speed control has been devised such as the P-I Control, Fuzzy Logic, Fuzzy Neural and the ArtificialNeuralNetwork Control systems. The ArtificialNeuralNetwork is a mathematical model inspired by biological neural networks. It consists of an interconnected group of artificial neurons, and it processes information using a connectionist approach to computation . It has the special ability of learning which implies that it can estimate the output of a system based on its experience on a set of previously trained data.
Figure 1 compares the measured windspeed at the hub height for WT24 with the out-of-sample forecasted windspeed obtained from bootstrapping an ensemble of ANNs. The best deterministic forecast for the windspeed based on a statistical combination of the individual forecasts in the super-ensemble [determined in accordance to Equations (10) and (13)] is shown in this figure (dot-dashed line labeled “Best forecast”). A comparison of the best deterministic forecast with the measured windspeed shows that the forecast captures adequately the longer temporal trends in the measured windspeed. In addition, the 95% prediction confidence intervals obtained using the two-stage weighted-averaging method is exhibited in Figure 1 as the dotted lines demarcated using an open circle. Note that the 95% prediction uncertainty range appears to cover most of the observations, providing an observation coverage that is consistent (approximately or better) with the quoted level of confidence in the uncertainty interval. A quantitative assessment of the forecast performance in this case is summarized in Table 1, which summarizes the RMSE and MAE in the windspeed prediction using the bootstrapped neural-network methodology. The performance of this forecast methodology can be compared to that obtained from a simple persistence model forecast which uses the current windspeed to predict the value of the future windspeed.
It is well known that fossil fuels are being depleted at a very fast rate which motivates to supplement the power generation from the renewable sources such as wind, solar, tidal and fuel cells etc. Among the several renewable power generating systems, the wind power generation dominates the other sources of renewable power. However, integration of wind power system to the existing power system possess a number of problems in view of achieving good power quality, stability and power dispatching issues, due to the fact it is non-dispatchable and volatility. These problems can be resolved if one could forecast the windspeed and wind power. It may be noted that as wind power is a function of windspeed, forecasting of wind power can be accomplished though windspeed forecast. But wind power generation depends on the availability of the wind. It is estimated that by 2020, about 12% of the world’s electricity will be available through wind generation .
The main advantage of Artificial intelligence (AI) is solving complex problems in less time and with high precision, such as using optimization methods to solve the complex control problem. Also, AI can easily predict and take the correct decisions with little margin of error. It can be used for predicting the change in windspeed and its impact on stability of power system. In this paper, AI has been used to predict and determine the optimal value of the control gains of SSSC which can enhance the performance of CWF. On other side, AI represents high technology so that it is storage costly. In last years, Artificial Intelligence (AI) has been used extensively in improving the performance of FACTS and enhancing the performance of wind farms interconnected grid. A genetic algorithm has been implemented to tune different type of FACTS interconnected wind farms and photovoltaic solar plant in . In ref   multi-objective genetic algorithm is used to improve the performance of DFIG. Also, multi-objective genetic algorithm is used to find the optimal gains of SSSC in . Adaptive-network-based fuzzy inference system (ANFIS), ANN and genetic algorithm are proposed in  to improve the reactive power control of STATCOM. The whale optimization algorithm, genetic algorithm and ANN were used in  to determine the optimal parameters of STATCOM integrated with CWF. In Ref  particle swarm optimization is used to tune and damp power system oscillation of DFIG wind farms integrated with SSSC. A new control strategy based on ANFIS is proposed in  to improve the performance of DFIG wind farm integrated with SSSC.
The learning algorithm plays an important role in pattern classification particularly if the image object varies its features. The topographical features used for training and classification are central peak and domed crater types. These two craters have an outer ring with a center feature. Classification of these two patterns is discussed here with artificialneuralnetworkusing back propagation algorithm. Training of the neuralnetwork uses topographical images with the consideration of different transformations like rotation and scaling. This neuralnetwork trained images with 10 x 10 pixels and more. The analysis of training of image pattern, classification of pattern is studied here.
Cukai adalah salah satu cara untuk membiayai perbelanjaan kerajaan dan ia memainkan peranan penting dalam meningkatkan hasil kerajaan. Jumlah cukai yang dikutip sebenarnya bergantung kepada kecekapan system cukai sesebuah negara. Apabila sistem cukai tidak berkesan, kebanyakan orang ramai akan mengambil peluang ini untuk mengelak daripada membayar cukai dan aktiviti pengelakan cukai akan terus meluas. Dalam keadaan ini, kerajaan akan menghadapi kesukaran dalam menentukan hasil aktiviti-aktiviti yang dilakukannya dan tidak dapat menyediakan perkhidmatan yang diperlukan oleh masyarakat. Menyedari akan kepentingan kesan aktiviti pengelakan cukai ke atas ekonomi negara, kajian ini cuba menentukan faktor-faktor utama yang menyebabkan berlakunya aktiviti pengelakan cukai dan kepentingan relatifnya. Kajian ini menggunakan kaedah ArtificialNeuralNetwork dalam menganalisis data Malaysia dari tahun 1963 hingga 2011. Keputusan kajian menunjukkan bahawa beban cukai, saiz kerajaan dan kadar inflasi member kesan positif ke atas aktiviti pengelakan cukai. Walau bagaimanapun, pendapatan pembayar cukai dan keterbukaan perdagangan memberi kesan negative kepada aktiviti pengelakan cukai. Keputusan kajian juga menunjukkan bahawa pendapatan pembayar cukai adalah secara relatifnya lebih penting daripada faktor-faktor lain yang mempengaruhi aktiviti pengelakan cukai.
It is important to conduct this study to get all the variables as many as possible in developing the accident prediction model and to propose an accident prediction with improvement of accident statistical model. ArtificialNeural Networks (ANN) model is usually used for prediction cases. By using ANN model, these factors can be determined by collecting the input data from the critical road or highway. The input data can be process by the ANN applied software to get the predict result for the forecasting purposes for the road or highway. This ANN applied software is also easy and ready to use for any level of users which they can implement or analyze all the parameters and accident data for the future prediction.
neural networks, in other words, is an emulation of biological neural system. As its biological predecessor, an artificialneuralnetwork is an adaptive system. By adaptive, it means that each parameter is changed during its operation and it is deployed for solving the problem in matter. This is called the training phase. Simulation results show that STATCOM devices significantly improve the performance of the wind farm and power network. This paper also presents Algorithm and Program in MATLAB for training of ArtificialNeuralNetwork for calculation of weights and biases and then using these weights and biases Program in MATLAB is presented for forecasting the voltage values on bus 1 and bus2 of the power network on different values of Loads L1 and L2 at different time.
The third parameter, the momentum; it correlation values 0.005 have no more than 50% overall accuracy. This means that the momentum value that is too small is less used as a reference for network simulation, in this case the critical land classification. The fourth parameter, the number of iteration; assessing the accuracy with the addition of iterations is based on the idea that the more data studied then the system will take longer to repeat the learning. The same logic is used in the human nervous system; with much data that people need more repetition to remember the data studied compared to when the data studied is less. some simulations have achieved above 70% accuracy with a larger iteration, and low accuracy occurs at iteration 5000. However, increasing the number of iterations is also strongly influenced by other parameters. As the number of iterations and the RMS error, the smaller the iteration or the less repetition is done then the higher RMS error on the training and testing. Conversely, the more the number of iterations is done, then the smaller the RMS error (Fig.2)
Non-invasive genetic sampling is increasingly being used for monitoring mammalian carnivore populations. However, environmental conditions in the tropics challenge researchers’ ability to collect samples. We present the results of a preliminary study on the feasibility of using scent- baited hair traps for population monitoring of mammalian carnivores in PeninsularMalaysia. Stations were baited using either fatty acid scent or male cologne applied to hair traps. Video camera traps were also used to monitor carnivore reactions to the scent stations. We recorded 19 visits by seven carnivore species over 764 camera trap nights. Cheek-rubbing and scent- marking behaviour was recorded only for single individuals of two species: the Malayan tiger (Panthera tigris jacksoni) and clouded leopard (Neofelis nebulosa). This study suggests that scent-baited hair traps hold some promise for ecological issues requiring DNA analysis in PeninsularMalaysia. Additional research is needed to develop its full potential for conservation monitoring of large carnivores.
increase in the chances of water quality deterioration which adversely affects sustaining life. Modelling of water quality is one of the most prominent area of study in the present scenario. The objective of the present study was to model Biochemical Oxygen Demand (BOD) and Chemical Oxygen Demand (COD) usingArtificialNeural Networks (ANNs). Among twenty four parameters, twenty two parameters were taken as input values. Three different station of Korapuzha river in Kozhikode, Kerala have been taken as the study area. The three sampling points are Purakatteri, Kanyankode and Korapuzha. The data considered for the study was monthly of the duration 2006 to 2015. Since there were large number of input parameters, Factor Analysis was done in order to identify the most prominent input parameters for the BOD and COD modelling. The factors which were having a fine correlation with that of the output parameter were used for the development of ANN model. Different trials for the three sampling points were done with different input parameters depending upon the correlation value from the factor analysis result. The algorithm used in the ANN model is feed – forward back propagation . The comparison of the models were done by observing the variation of predicted values from the observed values. The best result was obtained for the input parameter combination having high correlation with that of output parameter. The results obtained suggested that ANN technique provide better results when combined with factor analysis for data reduction.
The typical structure of an ArtificialNeuralNetworkmodel for univariate or multivariate has three different layers as shown in figure 4. The first layer is the input layer consist input node depending on the number of variables used in the model. The second layer is the hidden layer and the number of nodes in hidden layer is depends on the user. The user can modified the number of nodes in order to find the least error for the model. Based on the past research paper, the suitable number of nodes for hidden layer is between five to ten nodes only. Building complex model will not ensure better forecasting result. The third layer is the output layer where the values of each node in hidden layer are summed and multiplied with their associated weights. For each of the input and hidden layer, they have their own bias value to stabilize the weightage of nodes in each layer.