Adaptive Network-based Fuzzy Inference System Model Research Articles

Evaluation of soft computing algorithms for estimation of spatial transmissivity

This paper explored the potential of inverse technique using adaptive network-based fuzzy inference system (ANFIS), self-organised maps (SOMs) and M5P model tree-based regression approach to estimate the spatial transmissivity of aquifer domain. The study is based on coupling of finite element method (FEM)-soft computing (ANFIS, SOMs, M5P) model, which serve as forward (FEM) and inverse (ANN, SOMs, M5P) models. The root mean square error, coefficient of correlation and Nash-Sutcliffe efficiency index are used as comparison criteria for evaluating the models. The results from this study suggest that M5P model tree-based modelling approach is superior in accuracy in comparison to the ANFIS and SOMs model investigated in this study. This study also suggests that M5P model trees, being analogous to piecewise linear functions, have advantages over other techniques as they offer more insight into the developed model and are very efficient in training, and always converge.

Surface Wind Nowcasting in the Penghu Islands Based on Classified Typhoon Tracks and the Effects of the Central Mountain Range of Taiwan

Abstract The purposes of this study were to forecast the hourly typhoon wind velocity over the Penghu Islands, and to discuss the effects of the terrain of the Central Mountain Range (CMR) of Taiwan over the Penghu Islands based on typhoon tracks. On average, a destructive typhoon hits the Penghu Islands every 15–20 yr. As a typhoon approaches the Penghu Islands, its track and intensity are influenced by the CMR topography. Therefore, CMR complicates the wind forecast of the Penghu Islands. Six main typhoon tracks (classes I–VI) are classified based on typhoon directions, as follows: (I) the direction of direct westward movement across the CMR of Taiwan, (II) the direction of northward movement along the eastern coast of Taiwan, (III) the direction of northward movement traveling through Taiwan Strait, (IV) the direction of westward movement traveling through Luzon Strait, (V) the direction of westward movement traveling through the southern East China Sea (near northern Taiwan), and (VI) the irregular track direction. The adaptive network-based fuzzy inference system (ANFIS) and multilayer perceptron neural network (MLPNN) were used as the forecasting technique for predicting the wind velocity. A total of 49 typhoons from 2000 to 2012 were analyzed. Results showed that the ANFIS models provided high-reliability predictions for wind velocity, and the ANFIS achieved more favorable performance than did the MLPNN. In addition, a detailed discussion on the interaction of the CMR with the Penghu Islands based on various track directions is provided. For class I, the CMR is observed to have significantly influenced variations in wind speed when typhoons approached the Penghu Islands. In addition, the winds on the Penghu Islands were observed to have been influenced by the distance from the typhoon center to the Penghu Islands for all classes except class II.

Prediction of shear capacity of channel shear connectors using the ANFIS model

Due to recent advancements in the area of Artificial Intelligence (AI) and computational intelligence, the application of these technologies in the construction industry and structural analysis has been made feasible. With the use of the Adaptive-Network-based Fuzzy Inference System (ANFIS) as a modelling tool, this study aims at predicting the shear strength of channel shear connectors in steel concrete composite beam. A total of 1200 experimental data was collected, with the input data being achieved based on the results of the push-out test and the output data being the corresponding shear strength which were recorded at all loading stages. The results derived from the use of ANFIS and the classical linear regressions (LR) were then compared. The outcome shows that the use of ANFIS produces highly accurate, precise and satisfactory results as opposed to the LR.

Modeling Potential Future Energy Demand for Turkey in 2034 by Using an Integrated Fuzzy Methodology

Abstract Since the decisions made regarding the future include uncertainty for operations, alternative predictions are needed to be developed in such decision-making processes. Accurate forecasting is a great help for companies in making the best decisions in terms of unit commitment, production, and maintenance planning. It is necessary for the companies to have prior foresight of future demand with adequate accuracy. Some data mining algorithms play the greatest role in predicting the demand forecasting. As a regular data-driven method, artificial neural networks (ANNs) are popular in energy forecasting. This paper investigated the application of the adaptive network based fuzzy inference system (ANFIS) as a forecasting tool for predicting the energy demand in Turkey. The benefit of the proposed model is forecasting energy needs through the evaluation of ANFIS application using data sets processed with principle component analysis (PCA) and collected from the energy forecast shootout (EFS) and Ministry of Natural Resources Turkey. The results showed that the hybrid ANFIS model based upon fuzzy logic (FL) and ANN performed efficiently in term of forecast accuracy. Thus, it could be regarded as an alternative method in energy forecasting. Finally, the application of ANFIS in a long term energy forecasting was provided, and the results were interpreted.

A GA-Weighted Adaptive Neuro-Fuzzy Model to Predict the Behaviour of Magnetorheological Damper

Magnetorheological damper is a controllable device in semi-active suspension system to absorb unwanted movement. The accuracy of magnetorheological damper model will affect performance of the control system. In this paper, a combination of genetic algorithm (GA) and adaptive-network-based fuzzy inference system (ANFIS) approaches is utilized to model the magnetorheological damper using experimental results. GA algorithm is implemented to modify the weights of the trained ANFIS model. The proposed method is compared with ANFIS and artificial neural network (ANN) methods to evaluate the prediction performance. The result illustrates that the proposed GA-weighted adaptive neuro-fuzzy model has successfully predicted the magnetorheological damper behaviour and outperformed other compared methods.

Application of a fuzzy inference system for the prediction of longshore sediment transport

A fuzzy inference system (FIS) and a hybrid adaptive network-based fuzzy inference system (ANFIS), which combines a fuzzy inference system and a neural network, are used to predict and model longshore sediment transport (LST). The measurement data (field and experimental data) obtained from Kamphuis [1] and Smith et al. [2] were used to develop the model. The FIS and ANFIS models employ five inputs (breaking wave height, breaking wave angle, slope at the breaking point, peak wave period and median grain size) and one output (longshore sediment transport rate). The criteria used to measure the performances of the models include the bias, the root mean square error, the scatter index and the coefficients of determination and correlation. The results indicate that the ANFIS model is superior to the FIS model for predicting LST rates. To verify the ANFIS model, the model was applied to the Karaburun coastal region, which is located along the southwestern coast of the Black Sea. The LST rates obtained from the ANFIS model were compared with the field measurements, the CERC [3] formula, the Kamphuis [1] formula and the numerical model (LITPACK). The percentages of error between the measured rates and the calculated LST rates based on the ANFIS method, the CERC formula (Ksig=0.39), the calibrated CERC formula (Ksig=0.08), the Kamphuis [1] formula and the numerical model (LITPACK) are 6.5%, 413.9%, 6.9%, 15.3% and 18.1%, respectively. The comparison of the results suggests that the ANFIS model is superior to the FIS model for predicting LST rates and performs significantly better than the tested empirical formulas and the numerical model.

Modeling Ultraviolet Protection Factor of Polyester-Cotton Blended Woven Fabrics Using Soft Computing Approaches

Ultraviolet protection factor (UPF) of woven fabrics is modeled by using two soft computing approaches, namely adaptive network based fuzzy inference system (ANFIS) and artificial neural network (ANN). Three fabric parameters: proportion of polyester in weft yarns, weft count, and pick density are used as input parameters for predicting fabric UPF. Two levels (low and high) of membership function for each of the input parameters are used to reduce the complexity of ANFIS. The eight linguistic fuzzy rules trained by ANFIS are able to explain the relationship between fabric parameters and UPF. A comparison between ANFIS and ANN models is also presented. Both the models predict the UPF of fabrics with very good prediction accuracy in the testing data sets.

Modeling of TiN Coating Thickness Using ANFIS

In this paper, an approach in predicting thickness of Titanium Aluminum Nitrite (TiN) coatings using Adaptive Network Based Fuzzy Inference System (ANFIS) is implemented. The TiN coatings were coated on tungsten carbide (WC) using Physical Vapor Deposition (PVD) magnetron sputtering process. The N2 pressure, argon pressure and turntable speed were selected as the input parameters and the coating thickness as an output of the process. Response Surface Methodology (RSM) was used to design the matrix in collecting the experimental data. In the ANFIS structure, three bell shapes were used as input membership function (MFs). The collected experimental data was used to train the ANFIS model. Then, the ANFIS model was validated with confirmatory test data and compared with other prediction models in terms of the root mean square error (RMSE), residual error and prediction accuracy. The result indicated that the developed ANFIS model result was the lowest RMSE7 and average residual error, besides the highest in prediction accuracy compared to the other models. The result also indicated that the limited experimental data could be used in training the ANFIS model and resulting accurate predictive result.

Particles flow identification in pipeline using adaptive network-based fuzzy inference system and electrodynamic sensors

Purpose – An identification model for materials flow through a pipeline is presented in this paper. The development of the model involves fuzzy C-means clustering, in which different flow regimes can be identified by every adaptive network-based fuzzy inference system (ANFIS). The paper aims to discuss these issues. Design/methodology/approach – For experimentation, 16 electrodynamic sensors were used to monitor and measure the charge carried by dense particles flow through a pipeline in a vertical gravity flow rig system. Four ANFIS models were also used simultaneously to provide the expected output on thresh-holding and were evaluated for ten different flow regimes, which produced satisfactory results at high flow rate. Findings – The observations made on the four ANFIS models in the flow identification experimentation (in ten different flow regimes) have shown convincing and satisfactory results at high-flow rate of the particles. Originality/value – Electrodynamic sensors have shown strong sensing capability in identification of dense-particle flows within a conveyor; and also proven capability to operate effectively in harsh industrial environments due to their firm and simple structures. Moreover, it has been verified that these sensors can conveniently be applied in flow regime identification of solid particles.

ANFIS Modeling For a Water-Methanol System in a Spiral Plate Heat Exchanger

In this paper, an experimental investigation of Water- Methanol system in a Spiral plate Heat Exchanger (SHE) is presented. Experiments have been conducted by varying the mass flow rate of cold fluid (Methanol), the mass flow rate of hot fluid (Water) and inlet temperature of the hot fluid, by keeping inlet temperature of the cold fluid constant. The effects of relevant parameters on the performance of spiral plate heat exchanger are studied. Also, in this paper, an attempt is made to propose Adaptive Network based Fuzzy Inference System (ANFIS) and Artificial Neural Network (ANN) models for the analysis of SHE. The data required to train the models are obtained from the experimental data based on Response Surface Methodology (RSM). The ANN models are developed using Back Propagation Network (BPN) algorithm, incorporating Levenberg-Marquardt (L-M) training method. The ANFIS models are developed based on advanced neural- fuzzy technology. The ANFIS model possesses the robustness of fuzzy system, the learning ability of neural networks and can adapt to various situations. The accuracy of the trained networks are verified according to their ability to predict unseen data by minimizing root mean square error (RMSE), average percentage error (APE) and correlation coefficient (%R 2 ) value. The prediction of the parameters can be obtained without using charts and complicated equations. The data obtained from ANFIS and ANN models for overall heat transfer coefficient (U) and pumping power (Wp) are compared with those of experimental data. It is observed that the accuracy between the ANFIS model's predictions, NN model's predictions and experimental values are achieved with minimum % error, APE, RMSE and Correlation coefficient (%R 2 ). Also it is proved that ANFIS models give better performance

Wavelet-based models for air pollution modelling around coal mining sites in Jharkhand for 1, 3 and 5 day lead time

Coal dust is a major cause of air pollution in areas around opencast coal mining sites. Prediction of ambient concentration of pollutant should be known to implement any control techniques or to install any control equipment. Combining adaptive network based fuzzy inference system (ANFIS) and linear regression (LR) with discrete wavelet transform (DWT), two types of hybrid models were developed to forecast concentration of suspended particulate matter (SPM) for 1, 3 and 5-day-ahead around open cast coal mining sites in Jharkhand, India using DWT-obtained sub-time series of SPM and other relevant parameters, i.e., distance of point of forecast from core zone (x), ambient temperature (T), wind speed (w) and relative humidity (h). For the purpose of comparison ANFIS and LR models were also derived for the same lead time using original antecedent time series of SPM, T, w and h and distance of point of forecasting from the core zone. When compared, wavelet-based hybrid models predicted SPM with greater accuracy than ANFIS and LR models. Between the two types of hybrid models, the WANFIS type models gave slightly better results than the WR type.

DETERMINATION OF REFLECTANCE VALUES OF HYPERICUM'S LEAVES UNDER STRESS CONDITIONS USING ADAPTIVE NETWORK BASED FUZZY INFERENCE SYSTEM

The effects of water stress and salt levels on hypericum's leaves were examined on greenhouse-grown plants of Hypericum perforatum L. by spectral reflectance. Salt levels and irrigation levels were applied 0, 1, 2.5 and 4 deci Siemens per meter (dS/m), 80%, 100% and 120% respectively. Adaptive Network based Fuzzy Inference System (ANFIS) was performed to estimate the effects of water stress and salt levels on spectral reflectance. As a result of ANFIS, it was found that there was close relationship between actual and predicted reflectance values in Hypericum perforatum L. leaves. Performance of ANFIS was examined under different numbers of epoch and rules. On the other hand, RMSE, correlation and analysis time values were found as outputs. Correlation was 99%. The estimation of optimal ANFIS model was determined in 3*3*3 number of rules with 400 epochs.

Evaluation of ANFIS, ANN, and geostatistical models to spatial distribution of groundwater quality (case study: Mashhad plain in Iran)

Groundwater is one of the major sources of exploitation in arid and semiarid regions. Spatial and temporal quality distribution is an important factor in groundwater management. Thus for protecting groundwater quality, data on spatial and temporal distribution are important. Geostatistical models are the most advanced techniques for interpolation and spatial prediction of groundwater parameters. Determining the best and the most suitable model is also very essential which is the main aim in this study. In this research, inverse distance weighted (IDW), kriging, and cokriging methods in geostatistical and artificial neural network (ANN) and adaptive network-based fuzzy inference system (ANFIS) models were used for predicting the spatial distribution of groundwater electrical conductivity (EC) and those were compared together. EC and chloride (Cl−) are two important indicators for water quality assessment. Data were related to 120 wells in Mashhad plain (Iran). Groundwater resources have an important role in this region due to surface water deficit. After normalization of data, to geostatistical models, variograme was drawn; for selecting a suitable model for fitness on experimental variograme, less RSS value was used. Then using cross-validation and root mean square error (RMSE), the best method for interpolation was selected. To compare these three models, we used 25 % of observation data and determined the R2, RMSE, and MAE parameters. Different ANFIS structures were examined. Also in ANFIS method, different types of membership function, such as Gaussian, bell shape, and trapezoid for inputs of model, were used. Results showed that for interpolation of groundwater quality, cokriging method is superior to kriging method in geostatistical model. In cokriging method, Cl parameter was selected as auxiliary variable which had the highest correlation with EC. Results showed that ANN model had the best accuracy (R2 = 0.932, RMSE = 367.9, MAE = 265.78 μmos/cm) than ANFIS and geostatistical models.

Estimation of Monthly Mean Reference Evapotranspiration in Turkey

Monthly mean reference evapotranspiration (ET0) is estimated using adaptive network based fuzzy inference system (ANFIS) and artificial neural network (ANN) models. Various combinations of long-term average monthly climatic data of wind speed, air temperature, relative humidity, and solar radiation, recorded at stations in Turkey, are used as inputs to the ANFIS and ANN models so as to calculate ET0 given by the FAO-56 PM (Penman-Monteith) equation. First, a comparison is made among the estimates provided by the ANFIS and ANN models and those by the empirical methods of Hargreaves and Ritchie. Next, the empirical models are calibrated using the ET0 values given by FAO-56 PM, and the estimates by the ANFIS and ANN techniques are compared with those of the calibrated models. Mean square error, mean absolute error, and determination coefficient statistics are used as comparison criteria for evaluation of performances of all the models considered. Based on these evaluations, it is found that the ANFIS and ANN schemes can be employed successfully in modeling the monthly mean ET0, because both approaches yield better estimates than the classical methods, and yet ANFIS being slightly more successful than ANN.

Prediction of concrete compressive strength: Research on hybrid models genetic based algorithms and ANFIS

The management of concrete quality is an important task of concrete industry. This paper researched on the structured and unstructured factors which affect the concrete quality. Compressive strength of concrete is one of the most essential qualities of concrete, conventional regression models to predict the concrete strength could not achieve an expected result due to the unstructured factors. For this reason, two hybrid models were proposed in this paper, one was the genetic based algorithm the other was the adaptive network-based fuzzy inference system (ANFIS). For the genetic based algorithm, genetic algorithm (GA) was applied to optimize the weights and thresholds of back-propagation artificial neural network (BP-ANN). For the ANFIS model, two building methods were explored. By adopting these predicting methods, considerable cost and time-consuming laboratory tests could be saved. The result showed that both of these two hybrid models have good performance in desirable accuracy and applicability in practical production, endowing them high potential to substitute the conventional regression models in real engineering practice.

Stress–strain modeling of high-strength concrete by the adaptive network-based fuzzy inference system (ANFIS) approach

In this study, an adaptive network-based fuzzy inference system (ANFIS) approach is presented for modeling of high-strength concrete under uniaxial loading. The ANFIS approach applied to test the data of concrete cylinder test is available in the literature. In this paper, the stress–strain behavior of high-strength concrete subjected to axial load is obtained by using the ANFIS model. It is shown that the present model can predict the stress–strain behavior of concrete accurately by taking into account the effective parameters. The advantage of the proposed approach is that the stress–strain behavior of high-strength concrete can be predicted easily. The results of ANFIS approach are compared with the analytical models given in various studies concerning cylinder tests. The ANFIS approach results given show a good agreement with the experimental results.

A neuro-fuzzy model to predict the inflow to the guardialfiera multipurpose dam (Southern Italy) at medium-long time scales

Intelligent computing tools based on fuzzy logic and artificial neural networks have been successfully applied in various problems with superior performances. A new approach of combining these two powerful tools, known as neuro-fuzzy systems, has increasingly attracted scientists in different fields. Few studies have been undertaken to evaluate their performances in hydrologic modeling. Specifically are available rainfall-runoff modeling typically at very short time scales (hourly, daily or event for the real-time forecasting of floods) with in input precipitation and past runoff (i.e. inflow rate) and in few cases models for the prediction of the monthly inflows to a dam using the past inflows as input. This study presents an application of an Adaptive Network-based Fuzzy Inference System (ANFIS), as a neuro-fuzzy-computational technique, in the forecasting of the inflow to the Guardialfiera multipurpose dam (CB, Italy) at the weekly and monthly time scale. The latter has been performed both directly at monthly scale (monthly input data) and iterating the weekly model. Twenty-nine years of rainfall, temperature, water level in the reservoir and releases to the different uses were available. In all simulations meteorological input data were used and in some cases also the past inflows. The performance of the defined ANFIS models were established by different efficiency and correlation indices. The results at the weekly time scale can be considered good, with a Nash- Sutcliffe efficiency index E = 0.724 in the testing phase. At the monthly time scale, satisfactory results were obtained with the iteration of the weekly model for the prediction of the incoming volume up to 3 weeks ahead (E = 0.574), while the direct simulation of monthly inflows gave barely satisfactory results (E = 0.502). The greatest difficulties encountered in the analysis were related to the reliability of the available data. The results of this study demonstrate the promising potential of ANFIS in the forecasting of the short term inflows to a reservoir and in the simulation of different scenarios for the water resources management in the longer term.

Voltage Sag and Swell Estimation Using ANFIS for Power System Applications

Power quality is a term that is now extensively used in power systems applications, and in this context the voltage, current, and phase angle are discussed widely. In particular, different algorithms that are capable of detecting the voltage sag and swell information in a real time environment have been proposed and developed. Voltage sag and swell play an important role in determining the stability, quality, and operation of a power system. This paper presents ANFIS (Adaptive Network based Fuzzy Inference System) models with different membership functions to build the voltage shape with the knowledge of known system parameters, and detect voltage sag and swell accurately. The performance of each method has been compared with each other/other methods to determine the effectiveness of the different models, and the results are presented.

ANFIS Model for Time Series Prediction

This paper describes an architecture of ANFIS (adaptive network based fuzzy inference system), to the prediction of chaotic time series, where the goal is to minimize the prediction error. We consider the stock data as the time series. This paper focuses on how the stock data affect the prediction performance. In the experiments we changed the number of data as input of the ANFIS model, the type of membership functions and the desired goal error, thereby increasing the complexity of the training.

Prediction of wave parameters by using fuzzy inference system and the parametric models along the south coasts of the Black Sea

Forecasting of sea-state characteristics has a great importance in coastal and ocean engineering studies. Therefore, the purpose of this study was to investigate performances of Adaptive-Network-Based Fuzzy Inference System (ANFIS) and several parametric methods in the Black Sea. For this purpose, different fuzzy models with different input combinations were developed for two different wind data sources (TSMS and ECMWF) at two offshore buoy stations. It also aimed to apply several approaches to event-based data sets for wave predictions. Generally, in literature the tendency is to use time series data for wave predictions. In this kind of prediction approach, lagged time series data are taken as inputs and current or future variables are taken as output. In this study, event-based data for each independent storm were extracted from time series data. Simultaneous or concurrent data of wind speed, blowing duration, fetch length and wave characteristics were detected for each single storm. These event data were then used to set up models. The hindcast results were validated with significant wave height and mean wave period data recorded in Hopa and Sinop buoy stations. The performance of developed fuzzy models were also compared with that of four different parametric methods (Wilson, SPM, Jonswap, and CEM methods) applied for two wind data sources at both buoy stations. Finally, it was determined that in the prediction of both wave parameters (Hs and Tz) the ANFIS models (R = 0.66, squared correlation coefficient, and MAE = 0.37 m, mean absolute error, for the best model in prediction of Hs) were more accurate than the parametric methods (R = 0.63 and MAE = 0.75 m for the best model in prediction of Hs).