Subtractive Clustering Research Articles

An experimental investigation and neuro-fuzzy modeling to ascertain metal deposition parameters for the wire arc additive manufacturing of Incoloy 825

Additive manufacturing is critical to repair, rebuild, and rapidly manufacture major components in today's manufacturing environment. Wire arc additive manufacturing technique using metal inert gas welding, which is a better way to produce large-scale parts at a subsidized cost was experimented on Incoloy 825 due to its many industrial applications in the construction of large structures. As bead deposition is vital for wire arc additive manufacturing, single bead depositions were performed. Metrological features of all deposited beads are evaluated using the CNC Video Measuring System by capturing the macrostructures. The effects of input parameters on bead geometry are statistically analyzed, and regression models are derived to forecast bead geometry. An adaptive Neural Fuzzy Inference System was implemented to predict geometrical characteristics, and the results were compared with experimental results. Subtractive clustering fuzzy inference model obtained the lowest percentage error difference in predicting deposited bead geometry as a function of process parameters. Further, the multi-objective optimization technique determined optimal input parameters for the best bead geometry. Finally, using identified optimum parameters, wire arc additive manufacturing of a single-pass multi-layer wall of Incoloy 825 was successfully constructed for validation, which revealed that the microstructure of the built structure was of appreciable quality without cracks and porosity.

Research on a new power distribution control strategy of hybrid energy storage system for hybrid electric vehicles based on the subtractive clustering and adaptive fuzzy neural network

Research on a new power distribution control strategy of hybrid energy storage system for hybrid electric vehicles based on the subtractive clustering and adaptive fuzzy neural network

A modified convolutional neural network architecture for diabetic retinopathy screening using SVDD

Automatic diabetic retinopathy diagnostic methods are proposed to facilitate the examination process and act as the physician’s helper. Most of the traditional convolution neural network (CNN) algorithms use only spatial features for image category recognition. This approach may not be optimal for the screening diabetic retinopathy because the retinal images have generally the same feature maps with minor differences in spatial domain. We propose a new high level image understanding using a modified CNN architecture mixed with modified support vector domain description (SVDD) as a classifier. This new innovative architecture uses two pathways extracting features of the retinal images in both spatial and spectral domains. The standard pre-trained AlexNet is chosen for modification to avoid the time complexity of the training algorithms. In spite the advantages of the modified AlexNet with two pathways configuration and standard SVDD classification, the different SVDD kernel functions affect the performance of the proposed algorithm. By using the appropriate transformed data into two or three dimensional feature spaces, the proposed SVDD can obtain more flexible and more accurate image descriptions. Also, we compared the performance of our approach with that of the commonly used as classification methods such as K-Means, subtractive and FCM clustering. Our proposed architecture achieves more than 98% precision and sensitivity for two class classification.

Regional Flood Frequency Analysis Using the FCM-ANFIS Algorithm: A Case Study in South-Eastern Australia

Regional flood frequency analysis (RFFA) is widely used to estimate design floods in ungauged catchments. Both linear and non-linear methods are adopted in RFFA. The development of the non-linear RFFA method Adaptive Neuro-fuzzy Inference System (ANFIS) using data from 181 gauged catchments in south-eastern Australia is presented in this study. Three different types of ANFIS models, Fuzzy C-mean (FCM), Subtractive Clustering (SC), and Grid Partitioning (GP) were adopted, and the results were compared with the Quantile Regression Technique (QRT). It was found that FCM performs better (with relative error (RE) values in the range of 38–60%) than the SC (RE of 44–69%) and GP (RE of 42–78%) models. The FCM performs better for smaller to medium ARIs (2 to 20 years) (ARI of five years having the best performance), and in New South Wales, over Victoria. In many aspects, the QRT and FCM models perform very similarly. These developed RFFA models can be used in south-eastern Australia to derive more accurate flood quantiles. The developed method can easily be adapted to other parts of Australia and other countries. The results of this study will assist in updating the Australian Rainfall Runoff (national guide)-recommended RFFA technique.

Urban growth modelling in Qazvin, Iran: an investigation into the performance of three ANFIS methods

ABSTRACT Determining the rules that lead to the expansion of urban areas has always been a challenging factor in urban modeling. To overcome this issue, an ANFIS model is proposed to enhance the simulation of urban growth through automatic production of transition rules. Hence, 22 ANFIS models were trained using different division methods and a Cellular Automata-based Markov Chain (CA-MC) was developed to examine their efficiencies. The results of accuracies reveal that the ANFIS accompanied by subtractive clustering (ANFIS-SC) simulating urban growth with a Kappa coefficient of 0.76 and overall accuracy of 93.41% is superior to other ANFIS and CA-MC models.

Prediction of bio-heat and mass transportation in radiative MHD Walter-B nanofluid using MANFIS model

This research article discusses the heat and mass transportation of magnetohydrodynamic (MHD) flow of Walter-B nanofluid containing the gyrotactic microorganisms over a curved surface with warm radiation. The mathematical model of the physical flow problem has been formulated, and the analog equations have been normalized into dimensionless equations through similarity practice. The succeeding equations resolved analytically by Homotopy Perturbation Method (HPM) and numerically by Runge–Kutta–Fehlberg’s fourth fifth-order (RKF-45) technique by adopting shooting method via MATLAB and MAPLE software The impacts of various involved parameters on speed, temperature, focus and motile microorganisms of density distributions have been delineated through graphs. The patterns of skin friction coefficient Cfx, Nusselt number (Nux), Sherwood number (Shx) and motile microorganisms’ flux (Nhx) have been examined via tables and graphs. Finally, analyses of Cfx, Nux, Shx and Nhx executed with the aid of Multi output adaptive neuro fuzzy inference system (MANFIS). In MANFIS, the grid partitioning method (GPM) and subtractive clustering method (SCM) have been used to calculate the physical quantities (Cfx, Nux, Shx and Nhx), and the stability analysis has been performed with the assistance of GPM and SCM. The predicted results show that MANFIS-SCM is more accurate and efficient compared to the other existing technique.

Predicting Australia’s Domestic Airline Passenger Demand using an Anfis Approach

Abstract The forecasting of future airline passenger demand is critical task for airline management. The objective of the present study was to develop an adaptive neuro-fuzzy inference system (ANFIS) for predicting Australia’s domestic airline passenger demand. The ANFIS model was trained, tested, and validated in the study. Sugeno fuzzy rules were used in the ANFIS structure and Gaussian membership function, and linear membership functions were also developed. The hybrid learning algorithm and the subtractive clustering partition method were used to generate the optimum ANFIS models. The results found that the mean absolute percentage error (MAPE) for the overall data set of the ANFIS model was 3.25% demonstrating that the ANFIS model has high predictive capabilities. The ANFIS model could be used in other domestic air travel markets.

Detecting fishing URLs using fuzzy clustering algorithms with global optimization

An algorithm for detecting phishing URLs (classifier) using fuzzy clustering is proposed, which includes choosing the type of intelligent classifier and justifying its parameters using global optimization methods. The following were studied as intellectual classifiers: subtractive clustering and fuzzy clustering of C-means. To find (adjust) the optimal (for a specific task) parameters of intelligent classifiers, the use of global optimization methods is justified, including genetic algorithm, direct random search, annealing simulation method, multicriteria optimization and threshold acceptance method. As a criterion of global optimization, a combined criterion was used, which includes the definition of the regularity criterion calculated on the test sample and the bias (minimum shift) criterion based on the analysis of solutions. By modeling in the Matlab environment with the help of standard and developed programs, the evaluated efficiency of using the proposed algorithm is evaluated on the example of experimental data – a set of 150 phishing and 150 secure URLs. The set of experimental data included information about the domain name registrar, the lifetime of the domain, the geolocation of the hosting server, the presence of a secure connection with a valid certificate. By simulation it is established that the fuzzy classifier with the subtractive clustering algorithm and using the Sugeno structure and 6 clusters meets the minimum of the combined criterion. All phishing URLs that were mistakenly classified as secure were found to have a secure con-nection with a valid certificate. Thus, further research should be aimed at exploring additional informative attributes (features) that could allow better separation of phishing and secure URLs.

An Hour-Ahead PV Power Forecasting Method Based on an RNN-LSTM Model for Three Different PV Plants

Incorporating solar energy into a grid necessitates an accurate power production forecast for photovoltaic (PV) facilities. In this research, output PV power was predicted at an hour ahead on yearly basis for three different PV plants based on polycrystalline (p-si), monocrystalline (m-si), and thin-film (a-si) technologies over a four-year period. Wind speed, module temperature, ambiance, and solar irradiation were among the input characteristics taken into account. Each PV plant power output was the output parameter. A deep learning method (RNN-LSTM) was developed and evaluated against existing techniques to forecast the PV output power of the selected PV plant. The proposed technique was compared with regression (GPR, GPR (PCA)), hybrid ANFIS (grid partitioning, subtractive clustering and FCM) and machine learning (ANN, SVR, SVR (PCA)) methods. Furthermore, different LSTM structures were also investigated, with recurrent neural networks (RNN) based on 2019 data to determine the best structure. The following parameters of prediction accuracy measure were considered: RMSE, MSE, MAE, correlation (r) and determination (R2) coefficients. In comparison to all other approaches, RNN-LSTM had higher prediction accuracy on the basis of minimum (RMSE and MSE) and maximum (r and R2). The p-si, m-si and a-si PV plants showed the lowest RMSE values of 26.85 W/m2, 19.78 W/m2 and 39.2 W/m2 respectively. Moreover, the proposed method was found to be robust and flexible in forecasting the output power of the three considered different photovoltaic plants.

Fault location in series-compensated transmission lines using adaptive network-based fuzzy inference system

In this paper a new and accurate method for fault location in series-compensated transmission lines (SCTLs) is presented using adaptive network-based fuzzy inference system (ANFIS). In the proposed approach, the desired features are obtained using the Least-Squares Estimation (LSE) from two cycles of voltage and current data in only one line terminal. After that, extracted features, including the decaying DC offset component of the current, the fundamental components of current and voltage, and the phase difference between current and voltage for three-phase current and voltage signals, are normalized and applied as inputs to the fuzzy neural network for fault location. The subtractive clustering technique has been utilized to design the basic ANFIS. This study has scrutinized different cases, including two different ANFIS arrangements, implementing a single network to perform fault location on the entire line and a separate network for each line segment. Moreover, the efficiency of feature selection has been analyzed using the RReliefF algorithm. Various training patterns and tests have been provided in a test transmission system for different system conditions, such as different fault inception angles, fault locations, fault resistances, and structural conditions. The results validate that the proposed approach has accurately located the fault under a wide range of system variations.

Automatic Generation of Fuzzy Classification Rules from Data

In this paper, we propose a method for automatic generation of fuzzy rules for data classification. The proposed method is based on subtractive clustering optimized using genetic algorithm. It searches for the FIS structure and number of rules that have the highest fitness value. Multiple performance measures are incorporated into the fitness function to address the problem of imbalanced data. Fitness function includes both training and validation to avoid data over-fitting. Classification performance of the proposed method is evaluated using different data sets and results are compared to those of a number of models generated by fuzzy cmeans clustering with various cluster numbers. Results show that the proposed method has better accuracy and a well compromised sensitivity and specificity.

Artificial intelligence in energy industry: forecasting electricity consumption through cohort intelligence & adaptive neural fuzzy inference system

ABSTRACT Demand forecasting is critical for energy systems, as energy is difficult to store and should only be supplied as needed. Researchers attempted to improve forecasts of energy consumption. However, they assume independent factors increase at a constant growth rate, which is unrealistic. Existing methods are designed to determine annual consumption, whereas energy-planning organizations rely on short- or medium-term consumption values. Therefore, we propose a new forecasting framework that introduces new models and scenarios. We apply a cohort intelligence-based adaptive neuro-fuzzy inference system (CI-ANFIS) with a subtractive clustering and grid partition approach to forecast net electricity consumption. One challenge in accurately predicting electricity consumption for specific projection intervals is missing values for factors independent of those known for existing net consumption. Then, we utilize a regression equation scenario approach. We test our framework using a real-world energy consumption dataset and show that our proposed framework outperforms the existing methods.

ANFIS with input space division for modeling magnetorheological energy absorber

A magnetorheological energy absorber (MREA) is a promising actuator for vibration and shock mitigation due to its continuous and controllable damping force. To take full advantage of this feature, a high-fidelity model that can accurately predict the MREA dynamic force is required. Therefore, this study focuses on proposing adaptive neural-fuzzy inference system (ANFIS) models for MREA that use the impact speed, displacement, velocity, acceleration and current as inputs and the MREA force as output. Three types of ANFIS models are developed based on different input space division methods, namely, the grid partition (GP) method, subtractive clustering (SC) algorithm and fuzzy c-means (FCM) algorithm. The best ANFIS model is chosen from each category by six criteria, i.e., mean absolute error (MAE), mean absolute percentage error (MAPE), root mean square error (RMSE), coefficient of variation (cov), coefficient of determination (R2) and training time. In addition, the desirability function is proposed to further decide the best model because these criteria do not have a uniform trend. Then, a comprehensive comparative analysis of the three best models is performed. The results show that the ANFIS model with 53 FIS rules generated by the FCM algorithm has the largest global desirability of 0.66782, demonstrating that it is the best one to predict the dynamic force of MREA under high-speed impacts. Accordingly, the values of MAE, MAPE, RMSE, cov and R2 of this model varied in ranges of 0.31679-0.57118, 4.34181-8.43942, 0.44808-1.03095, 6.14134-15.23283 and 0.98973-0.99834, respectively, for various data sets.

Parameter Estimation-Based Slime Mold Algorithm of Photocatalytic Methane Reforming Process for Hydrogen Production

The key contribution of this paper is to determine the optimal operating parameters of the methane reforming process for hydrogen production. The proposed strategy contained two phases: ANFIS modelling and optimization. Four input controlling parameters were considered to increase the hydrogen: irradiation time (min), metal loading, methane concentration, and steam concentration. In the first phase, an ANFIS model was created with the help of the experimental data samples. The subtractive clustering (SC) technique was used to generate the fuzzy rules. In addition, the Gaussian-type and weighed average were used for the fuzzification and defuzzification methods, respectively. The reliability of the resulting model was assessed statistically by RMSE and the correlation (R2) measures. The small RMSE value and high R2 value of testing samples assured the correctness of the modelling phase, as they reached 0.0668 and 0.981, respectively. Based on the robust model, the optimization phase was applied. The slime mold algorithm (SMA), as a recent as well as simple optimizer, was applied to look for the best set of parameters that maximizes hydrogen production. The resulting values were compared by the findings of three competitive optimizers, namely particle swarm optimization (PSO), Harris hawks optimization (HHO), and evolutionary strategy HHO (EESHHO). By running the optimizers 30 times, the statistical results showed that the SMA obtained the maximum value with high mean, standard deviation, and median. Furthermore, the proposed strategy of combining the ANFIS modelling and the SMA optimizer produced an increase in the hydrogen production by 15.7% in comparison to both the experimental and traditional RSM techniques.

Clustered adaptive neuro-fuzzy inference system models of net radiation flux using temperature series

This study proposes three clustered adaptive neuro-fuzzy inference system (ANFIS) models for the prediction of net radiation flux over climatic regions in Nigeria. Five temperature indices data obtained from the Nigerian Meteorological Agency, Lagos covering a period 1983 to 2013 were used as input variables to train and test the ANFIS models. Comparative analyses of the three ANFIS models—grid partitioning clustering (GPC), the subtractive clustering (SC), and the fuzzy c-means clustering (FCM) using the minimum values of coefficient of uncertainty (UII)—showed that ANFIS-GPC performed best in the semi-arid and sub-humid humid regions with values of 0.089 and 0.092, respectively. ANFIS-FCM performed best in the sub-humid dry region with a value of 0.074. ANFIS-SC performed best in the humid region with a value of 0.080. It can be concluded that the proposed models are regional dependent, and they are suitable for the prediction of net radiation flux over respective tropical regions for practical purposes.

Machine learning algorithms for soil moisture estimation using Sentinel-1: Model development and implementation

The present study provided the first-time comprehensive evaluation of 12 advanced statistical and machine learning (ML) algorithms for the Soil Moisture (SM) estimation from dual polarimetric Sentinel-1 radar backscatter. The ML algorithms namely support vector machine (SVM) with linear, polynomial, radial and sigmoid kernel, random forest (RF), multi-layer perceptron (MLP), radial basis function (RBF), Wang and Mendel’s (WM), subtractive clustering (SBC), adaptive neuro fuzzy inference system (ANFIS), hybrid fuzzy interference system (HyFIS), and dynamic evolving neural fuzzy inference system (DENFIS) were used. Extensive field samplings were performed for collection of in-situ SM data and other parameters from the selected sites for seven different dates and at two different locations (Varanasi and Guntur District, India), concurrent to Sentinel-1 overpasses. The backscattering coefficients were considered as input variables and SM as output variable for the training, validation and testing of the ML algorithms. The site at Varanasi was used for the training, validation and testing of the models. On the other hand, the Guntur site was used as an independent site for checking the model performance, before finalizing the algorithms. The performances of different trained algorithms were evaluated in terms of correlation coefficient ( r ), root mean square error (RMSE) (in m 3 /m 3 ) and bias (in m 3 /m 3 ). The study identified the RF, SBC and ANFIS as the top three best performing models with comparable and promising SM estimation. In order to test the robustness of these best models (RF, SBC and ANFIS), further performance analysis was performed to the independent datasets of the Varanasi and Guntur test sites, which indicates that the performance of these three models were consistent and SBC can be recommended as the best among all for SM estimation.

Application of hybrid ANFIS-based non-linear regression modeling to predict the %oil yield from grape peels: Effect of process parameters and FIS generation techniques

The present study evaluates the extraction process parameters of grape peel oil using non-linear hybrid Adaptive Neuro-Fuzzy Inference System (ANFIS) model. The experiments were conducted at temperature (80–100 °C), and time (5–9 h) at 5 levels with output parameter as oil yield. The experimental result and factor interaction analysis shows that temperature had the most significant effect on the oil yield. The oil yield estimation performance indicators for the Fuzzy Inference (FIS) generation techniques are: Fuzzy C-means Clustering (FCM), [R 2 = 0.9102, MSE = 0.278, RMSE = 0.527], Subtractive Clustering (SC), [R 2 = 0.9090, MSE = 0.300, RMSE = 0.548] and Grid Partitioning (GP), [R 2 = 0.827, MSE = 0.668, RMSE = 0.817]. This indicates that Fuzzy C-means FIS generation fits better than Subtractive Clustering and Grid Partitioning techniques. For future research, a substantial volume of data is recommended for successful ANFIS training. • Oil yield extraction from grape peels using Soxhlet method. • Prediction of oil yield using hybrid ANFIS non-linear regression model. • Analysis of the effect of process parameters on oil yield. • FIS generating strategies utilized in ANFIS modeling are compared.

Clustering of Marine Vessel Trajectory Data for Routes Planning through Water Areas with Heavy Traffic

The article is devoted to the problem of ensuring the safety of vessel traffic. One of the elements of the traffic organization in areas with heavy navigation is the system of establishing the routes of vessels. This system is a set of restrictions imposed by a certain traffic pattern and rules adopted in a particular water area. The paper considers the problem of planning a transition route for water areas with heavy marine traffic. The planning of the vessels transition route during the movement of the water area with established routes must be carried out taking into account the specified restrictions. A possible way to identify these restrictions is to isolate the movement patterns of a particular sea area from the retrospective information about its traffic. Model representations of such a problem can be formulated on the basis of the idea of clustering the parameters of traffic. The route planning problem model is based on finding the shortest path on a weighted graph. Several ways of constructing such a graph are proposed: a regular grid of vertices and edges; a layered ore random grid of vertices and edges; vertices and edges based on retrospective data. The weight of the edges is proposed to be set as a function of the “desirability” of a particular course of the vessel for each point of the water area, taking into account the identified movement patterns. The paper discusses possible clustering methods and makes a choice in favor of subtractive clustering.

Research for Fault Diagnosis Method and System for Diesel Engine Based on ANFIS

<p>After compliance verification, operating vehicles can enter the road transportation market. Diesel engine is the main power source of these vehicles, there will be some typical faults during the use of diesel engine, which will affect the technical status of vehicles. According to the fault diagnosis problem of diesel engine, a fault diagnosis method based on Adaptive-Network-Based Fuzzy Inference System(ANFIS) was proposed, Subtractive clustering algorithm was used to confirm the original structure of fuzzy inference model, and ANFIS was used to build an original fault diagnosis model of diesel engine. Hybrid algorithm is used to train the parameter of fuzzy rule, and the final model is established. Simulation experiment results show that the modeling algorithm based on subtractive clustering-ANFIS is effective. It has been found that the average error is 7%, the recognition accuracy is 93.33%. Simulation results show that the fitting ability, convergence speed and recognition accuracy of ANFIS model are all superior to back propagation neural networks (BPNN), and much more suitable as diesel engine fault diagnosis model. Finally, an effective fault diagnosis system is developed by using the given method.</p> <p> </p>

Penerapan Data Mining Dalam Pengelompokan Data Member Card Mitra10 Untuk Meningkatkan Rewards Terhadap Konsumen dengan Metode Fuzzy Subtractive Clustering

Data clustering of Mitra10 member card, it is only based on the payment day by customer. And the total point only based on 0,25% calcutation of each payment. This is not so efficient knowing that there are so many ungrouping data that can be use for increase the rewards to the customer. In order to clustering the data of member card total point, it need a method of data clustering. The method election really affecting the result of data clustering. After all of the member card Mitra10 data transformed to the number form, then the datas can be grouping to fuzzy subtractive clustering algorythm. The data need to divided to some cluster : Select the amount of the cluster. In this research the data will be divided to 3 cluster. Select the center point of each cluster and fisrt center point selected randomly to generate main center point of each cluster. One data will be part of one cluster that has the smallest distance from the center cluster. Example for the first data, smallest distance get by cluster 1, so that the first data will be member of cluster 1. And so for the second data , smallest distance in cluster 3, so it will be member of cluster 3. In this iteration, center point of each cluster does not changed and there is no moving data to another cluster. Cluster 1 result: Has center (1,413, 1,195, 1,695) cluster 1 dominated by the bronze group. Cluster 2 result: Has center (2,658, 1,219, 1,585) cluster 2 dominated by the silver group. Cluster 3 result: Has center (4, 1,8, 3) cluster 3 dominated by the silver group