Co-active Neuro-fuzzy Inference System Research Articles

Predicting the Performance of Gorgan Wastewater Treatment Plant Using ANN-GA, CANFIS, and ANN Models

A reliable model for any wastewater treatment plant (WWTP) is essential to predict its performance and form a basis for controlling the operation of the process. This would minimize the operation costs and assess the stability of environmental balance. This study applied artificial neural network-genetic algorithm (ANN-GA) and co-active neuro-fuzzy logic inference system (CANFIS) in comparison with ANN for predicting the performance of WWTP. The result indicated that the GA produces more accurate results than fuzzy logic technique. It was found that GA components increased the ANN ability in predicting WWTP performance. The normalized root mean square error (NRMSE) for ANN-GA in predicting chemical oxygen demand (COD), total suspended solids (TSS) and biochemical oxygen demand (BOD) were 0.15, 0.19 and 0.15, respectively. The corresponding correlation coefficients were 0.891, 0.930 and 0.890, respectively. Comparing these results with other studies showed that despite the slightly lower performance of the current model, its requirement for a lower number of input parameters can save the extra cost of sampling.

CANFIS‐Based Semi‐active Vibration Control of Stochastically Excited High‐Rise Civil Engineering Structures with Nonlinearities and Uncertainties

AbstractThis study introduces a co‐active neuro‐fuzzy inference system (CANFIS) based vibration control strategy for high‐rise civil engineering structures, such as buildings, wind turbines and towers. The proposed approach uses semi‐active dampers, which are distributed in the structure. Attached to the structural members, semi‐active dampers generate adaptive restoring forces to reduce the structural vibrations by autonomously tuning their parameters. The nonlinear behavior and the uncertainties involved in the mathematical models of both the semi‐active dampers and the structures require sophisticated control strategies. Previous studies have shown that fuzzy logic controllers (FLCs) can overcome these challenges and perform effectively with semi‐active dampers to mitigate the structural vibrations. Nevertheless, FLCs require a predefined parameter tuning of their fuzzy rules and membership functions, which must be performed based on the knowledge of a human expert. Combining FLCs with artificial neural networks (ANNs), hybrid systems are proposed, which can determine the parameters of FLCs by using learning based training and adaptation strategies of ANNs. However, there is only limited research investigating the performance of these adaptive systems in the field of vibration control of civil engineering structures. The present work is concerned with a numerical formulation of such an adaptive system, in particular the CANFIS, and its implementation on a non‐linear seismically excited three‐story steel benchmark building with semi‐active dampers. To determine the premise parameters of the membership functions and the consequent parameters of the fuzzy rules a hybrid training method is performed using the recursive least‐square algorithm and the steepest descent method. Investigations with historic earthquakes show a significant reduction of the roof displacements of the building. For instance, during the Hachinohe earthquake the peak value of the displacement is reduced by 87 %.

Modelling of soil permeability using different data driven algorithms based on physical properties of soil

Soil permeability is an important parameter for assessment of infiltration, runoff, ground water, drainage and structures design. In the current research, five different data driven algorithms namely Multilayer Perceptron (MLP), Co-Active Neuro-Fuzzy Inference System (CANFIS), Support Vector Machine (SVM), Decision Tree (DT) and Random Forest (RF) algorithms and also, their wavelets (W-MLP, W-CANFIS, W-SVM, W-DT and W-RF algorithms) were used to predict soil permeability based on physical properties of soil. Also, reliable information/input vectors were assessed based on Gamma Test (GT). Sand, silt, clay and organic content (OC) parameters were chosen as information vectors based on gamma test. The potential of data driven algorithms were evaluated based on different statistical indices during model development and validation phase. It was found that wavelet based algorithms viz. W-MLP, W-CANFIS, W-SVM, W-DT and W-RF simulated better results of soil permeability compared to non-wavelet (MLP, CANFIS, SVM, DT and RF) algorithms. Among all wavelet and non-wavelet algorithms, W-RF algorithm had the highest accuracy and efficiency of model. The results of sensitivity analysis indicated that clay > silt > sand > OC > BD > PD was the order of sensitive parameters for soil permeability prediction based on data driven algorithms.

The viability of co-active fuzzy inference system model for monthly reference evapotranspiration estimation: case study of Uttarakhand State

Abstract Reference evapotranspiration (ETo) is a major component of the hydrological cycle linking the irrigation water requirement and planning and management of water resources. In this research, the potential of co-active neuro-fuzzy inference system (CANFIS) was investigated against the multilayer perceptron neural network (MLPNN), radial basis neural network (RBNN), self-organizing map neural network (SOMNN) and multiple linear regression (MLR) to estimate the monthly ETo at Pantnagar and Ranichauri stations, located in the foothills of Indian central Himalayas of Uttarakhand State, India. The significant combination of input variables for implemented techniques was decided by the Gamma test (GT). The results obtained by CANFIS models were compared with MLPNN, RBNN, SOMNN and MLR models based on performance evaluation indicators and visual inspection using line, scatter and Taylor plots for both the stations. The results of comparison revealed that CANFIS-5/CANFIS-9 models (RMSE = 0.0978/0.1394, SI = 0.0261/0.0475, COE = 0.9963/0.9846, PCC = 0.9982/0.9942 and WI = 0.9991/0.9959) with three and five input variables provide superior results for estimating monthly ETo at Pantnagar and Ranichauri stations, respectively. Also, the adopted modelling strategy can build a truthful expert intelligent system for estimating the monthly ETo at the study stations.

Application of Heuristic Approaches for Prediction of Hydrological Drought Using Multi-scalar Streamflow Drought Index

Quantification and prediction of drought events are important for planning and management of water resources in coping with climate change scenarios at global and local scales. In this study, heuristic approaches including Co-Active Neuro Fuzzy Inference System (CANFIS), Multi-Layer Perceptron Neural Network (MLPNN) and Multiple Linear Regression (MLR) were utilized to predict the hydrological drought based on multi-scalar Streamflow Drought Index (SDI) at Naula and Kedar stations located in upper Ramganga River basin, Uttarakhand State, India. The SDI was calculated on 1-, 3-, 6-, 9-, 12- and 24-month time scales (SDI-1, SDI-3, SDI-6, SDI-9, SDI-12, and SDI-24) using monthly streamflow data of 33 years (1975-2007). The significant input variables (lags) for CANFIS, MLPNN, and MLR models were derived using autocorrelation and partial autocorrelation functions (ACF P while MLPNN was the best at 1- and 9-month time scales for Naula station, and at 6- and 9-month time scales for Kedar station. The MLR model was found to be the best at 24-month time scale for Naula station only. The results of this study could be helpful in prediction of hydrological drought on multiple time scales and decision making for remedial schemes to cope with hydrological drought at Naula and Kedar stations.

Viability of the advanced adaptive neuro-fuzzy inference system model on reservoir evaporation process simulation: case study of Nasser Lake in Egypt

Reliable prediction of evaporative losses from reservoirs is an essential component of reservoir management and operation. Conventional models generally used for evaporation prediction have a number of drawbacks as they are based on several assumptions. A novel approach called the co-active neuro-fuzzy inference system (CANFIS) is proposed in this study for the modeling of evaporation from meteorological variables. CANFIS provides a center-weighted set rather than global weight sets for predictor–predictand relationship mapping and thus it can provide a higher prediction accuracy. In the present study, adjustments are made in the back-propagation algorithm of CANFIS for automatic updating of membership rules and further enhancement of its prediction accuracy. The predictive ability of the CANFIS model is validated with three well-established artificial intelligence (AI) models. Different statistical metrics are computed to investigate the prediction efficacy. The results reveal higher accuracy of the CANFIS model in predicting evaporation compared to the other AI models. CANFIS is found to be capable of modeling evaporation from mean temperature and relative humidity only, with a Nash–Sutcliffe efficiency of 0.93, which is much higher than that of the other models. Furthermore, CANFIS improves the prediction accuracy by 9.2–55.4% compared to the other AI models.

Computer aided automatic detection of glioblastoma tumour in the brain using CANFIS classifier

Detection and diagnosis of brain tumour is complicated due to its similar characteristics between tumour pixels and non-tumour pixels in brain image. This paper proposes an efficient methodology for the detection and segmentation of glioblastoma tumour region in the brain. The proposed methodology for glioblastoma tumour classifications has the following stages as noise reduction, image fusion, feature extraction and classification. The median filter is used to remove the noises in the brain images and pixel level image fusion is applied to obtain the enhanced brain image. The features are extracted from the fused image and co-active neuro fuzzy inference system (CANFIS) classifier is used to classify the brain image into either benign or malignant. Further, morphological operations are applied on the classified malignant brain image in order to segment the glioblastoma tumour region. The proposed methodology achieves 96.43% sensitivity, 99.99% specificity and 99.91% accuracy with respect to ground truth images.

A new wavelet conjunction approach for estimation of relative humidity: wavelet principal component analysis combined with ANN

Relative humidity (RH) has an important effect on precipitation, especially in arid and semiarid regions. Prediction of RH has been the focus of attention of climate change researchers as well. In this investigation, the accuracy of six intelligent models, including an artificial neural network (ANN), a co-active neuro-fuzzy inference system (CANFIS), principal component analysis (PCA) combined with ANN (PCA–ANN) and three hybrid wavelet-artificial intelligence models, including WANN, WCANFIS and WPCA–ANN, was evaluated in daily RH prediction. Thirty weather stations located in different climates in Iran for the period 2000–2010 were selected for the evaluation and comparison of these models. The performance of each model was evaluated using correlation coefficient (r) and normal root mean square error (NRMSE). Based on the statistical evaluation criteria, the accuracy ranks of the six models were: WPCA–ANN, WCANFIS, WANN, PCA–ANN, ANN and CANFIS. The results indicated that the WPCA–ANN model was the optimal model for estimation of RH, and the range of NRMSE and r values were from 0.009 to 0.080 and from 0.996 to 0.999, respectively. Overall, WPCA–ANN is a new approach that can be successfully applied to predict RH with a high accuracy.

Performance improvement in vertical heterogeneous handoff methodology using CANFIS classification approaches

SummaryDetection of malicious Base Stations (BS) in heterogeneous wireless networks is significant for developing an efficient system in wireless networks. In this paper, the handoff is performed between two different networks such as Wi‐Fi and WiMax, which may be located in different locations in the same region. The trust features, Cumulative Binary Features and Cost Index Features of each BS in different wireless network environment, are derived individually, which differentiates the behavior of the normal BS and malicious BS in different networks using Co‐Active Neuro Fuzzy Inference System (CANFIS) classification approach. The proposed methodology is compared with other state‐of‐arts methods for evaluating the individuality and efficiency. The proposed heterogeneous handoff methodology achieves 96.02% of packet delivery ratio, 90% detection rate, 94.6% of precision, and 15.99 ms of latency.

Assessment of groundwater utilization status and prediction of water table depth using different heuristic models in an Indian interbasin

The knowledge of fluctuation of water table depth is highly required for proper planning and sustainable development of available water resources. This study intends to study groundwater behaviour under changing scenario in the lower part of Ganga–Ramganga interbasin. It also investigates the comparative performance of soft computing techniques, i.e. co-active neuro-fuzzy inference system (CANFIS), fuzzy logic and radial basis function network (RBFN), which are used for prediction of water table depth in the study area. Components of groundwater recharge and discharge along with seasonal water table depth covering a period of 23 years (1990–2012) are used to develop four combination sets of input variables. Different CANFIS structures, fuzzy logic rules and RBFN structures are applied to these combinations of input variables, and the best combinations are selected on the basis of the values of different performance indicators such as coefficient of determination (R2), mean absolute deviation, root mean square error, coefficient of variation of error residuals, Nash–Sutcliff efficiency, correlation coefficient (r), absolute prediction error and performance index. The result of this study indicates the superiority of fuzzy logic rule-based model than of CANFIS models and RBFN model in predicting water table depth.

Simulation of suspended sediment based on gamma test, heuristic, and regression-based techniques

In the present study, four different heuristic techniques viz. multi-layer perceptron (MLP), radial basis function (RBF), self-organizing maps (SOM), and co-active neuro-fuzzy inference system (CANFIS) with hyperbolic tangent and sigmoid transfer functions and two regression-based techniques, i.e., multiple linear regression (MLR) and sediment-rating curve (SRC), were used for suspended sediment modeling. Gamma test (GT), correlation function (CF), M test, and trail–error procedure were applied for estimation of appropriate input variables as well as training data length. The results of the GT and CF suggested the five input variables (Qt, Qt−1, Qt−2, St−1, and St−2, where Qt−1 and St−1 indicate the discharge and sediment values of one previous day) as the best combination. The optimal training data length (75% of total data) was estimated by M test and trail–error procedure for development of the applied models. The MLP with sigmoid transfer function (M-2) performed better than the all other models. The results of sensitivity analysis indicated that the present-day discharge (Qt), 1-day lag discharge (Qt−1) and 1-day lag suspended sediment (St−1) are the most influenced parameters in modeling current day suspended sediment (St).

Rainfall-runoff modeling in hilly watershed using heuristic approaches with gamma test

In this study, daily rainfall-runoff modeling was done using co-active neuro-fuzzy inference system (CANFIS) and multi-layer perceptron neural network (MLPNN) approaches in the hilly Naula watershed of Ramganga River in Uttarakhand, India. The daily observed rainfall and runoff data from June 1, 2000, to October 31, 2004, were used for training and testing of the applied models. Before starting the modeling process, the gamma test (GT) was used to select the best combination of input variables for each model. The simulated values of runoff from CANFIS and MLPNN models were compared with the observed ones with respect to root mean squared error (RMSE), Nash-Sutcliffe efficiency (CE), Pearson correlation coefficient (PCC). This study provides a conclusive evidence that the CANFIS shows better accuracy than the MLPNN models. Therefore, according to the best fitting CANFIS-10 model, the runoff of the present day depends on rainfall and runoff of current and previous 2 days for the studied area.

Computer Aided Segmentation of Blockages in Coronary Heart Images Using Canfis Classifier

Coronary vessel blockage segmentation is the fundamental component which extracts significant features from angiogram images to detect heart disease. This paper proposes an automated method of blockage segmentation from coronary angiogram images using coactive adaptive neuro fuzzy inference system (CANFIS) classifier. The proposed method consists of preprocessing, feature extraction and classification. The vessels in the coronary image are enhanced using preprocessing technique and then features are extracted from these images which are given to the CANFIS classifier. This classifier classifies the given test coronary image into either normal or abnormal. Further, the blockage is detected and segmented if the proposed system classifies the test image as abnormal. The proposed method achieves 99.76% sensitivity, 99.9% specificity and 99.9% accuracy for blockage vessel pixel detection.

Monthly pan-evaporation estimation in Indian central Himalayas using different heuristic approaches and climate based models

Estimation of pan-evaporation has a vital importance in water resources planning and management especially in arid/semi-arid regions. In this study, four heuristic approaches, multi-layer perceptron neural network (MLPNN), co-active neuro-fuzzy inference system (CANFIS), radial basis neural network (RBNN) and self-organizing map neural network (SOMNN) were utilized to estimate monthly pan-evaporation (EPm) at two locations, Pantnagar and Ranichauri, in the foothills of Indian central Himalayas. The monthly climatic data, minimum and maximum air temperatures, relative humidity in the morning and afternoon, wind speed, sun-shine hours and pan-evaporation, were used for model calibration and validation. The combination of appropriate input variables for the applied models was decided using gamma test. The results obtained by MLPNN, CANFIS, RBNN and SOMNN models were compared with climate-based empirical models, such as Stephens-Stewart (SS) and Griffith’s (G), on the basis of root mean squared error, coefficient of efficiency and coefficient of correlation. The results indicated that the performance of CANFIS (RMSE = 0.627 mm, COE = 0.936, COC = 0.979) and MLPNN (RMSE = 0.214 mm, COE = 0.989, COC = 0.970) models with six input variables was superior than the others models in estimating monthly pan evaporation at Pantnagar and Ranichauri stations.

Reservoir inflow forecasting with a modified coactive neuro-fuzzy inference system: a case study for a semi-arid region

Existing forecast models applied for reservoir inflow forecasting encounter several drawbacks, due to the difficulty of the underlying mathematical procedures being to cope with and to mimic the naturalization and stochasticity of the inflow data patterns. In this study, appropriate adjustments to the conventional coactive neuro-fuzzy inference system (CANFIS) method are proposed to improve the mathematical procedure, thus enabling a better detection of the high nonlinearity patterns found in the reservoir inflow training data. This modification includes the updating of the back propagation algorithm, leading to a consequent update of the membership rules and the induction of the centre-weighted set rather than the global weighted set used in feature extraction. The modification also aids in constructing an integrated model that is able to not only detect the nonlinearity in the training data but also the wide range of features within the training data records used to simulate the forecasting model. To demonstrate the model’s efficacy, the proposed CANFIS method has been applied to forecast monthly inflow data at Aswan High Dam (AHD), located in southern Egypt. Comparative analyses of the forecasting skill of the modified CANFIS and the conventional ANFIS model are carried out with statistical score indicators to assess the reliability of the developed method. The statistical metrics support the better performance of the developed CANFIS model, which significantly outperforms the ANFIS model to attain a low relative error value (23%), mean absolute error (1.4 BCM month−1), root mean square error (1.14 BCM month−1), and a relative large coefficient of determination (0.94). The present study ascertains the better utility of the modified CANFIS model in respect to the traditional ANFIS model applied in reservoir inflow forecasting for a semi-arid region.

Daily suspended sediment concentration simulation using hydrological data of Pranhita River Basin, India

Simulation of suspended sediment concentration (SSC) in a river is very important for planning and management of water resources. In this study, co-active neuro-fuzzy inference system (CANFIS), multi-layer perceptron neural network (MLPNN), multiple linear and non-linear regressions (MLR and MNLR), and sediment rating curve (SRC) techniques were applied for simulating the daily SSC at Tekra gauging site on Pranhita River, a major tributary of Godavari River basin, Andhra Pradesh, India. The daily data of discharge (m3/s) and SSC (g/l) from June 2000 to November 2003 were used for SSC simulation. The appropriate combination of input variables for CANFIS, MLPNN, MLR and MNLR models were decided using the gamma test (GT). The outcomes from CANFIS, MLPNN, MLR, MNLR and SRC models were compared to observed values of SSC on the basis of root mean squared error (RMSE), Pearson correlation coefficient (r) and coefficient of efficiency (CE). The results indicate the superiority of CANFIS model than the MLPNN, MLR, MNLR and SRC models in simulating daily SSC for the study location.

A method of groundwater quality assessment based on fuzzy network-CANFIS and geographic information system (GIS)

The process of water quality testing is money/time-consuming, quite important and difficult stage for routine measurements. Therefore, use of models has become commonplace in simulating water quality. In this study, the coactive neuro-fuzzy inference system (CANFIS) was used to simulate groundwater quality. Further, geographic information system (GIS) was used as the pre-processor and post-processor tool to demonstrate spatial variation of groundwater quality. All important factors were quantified and groundwater quality index (GWQI) was developed. The proposed model was trained and validated by taking a case study of Mazandaran Plain located in northern part of Iran. The factors affecting groundwater quality were the input variables for the simulation, whereas GWQI index was the output. The developed model was validated to simulate groundwater quality. Network validation was performed via comparison between the estimated and actual GWQI values. In GIS, the study area was separated to raster format in the pixel dimensions of 1 km and also by incorporation of input data layers of the Fuzzy Network-CANFIS model; the geo-referenced layers of the effective factors in groundwater quality were earned. Therefore, numeric values of each pixel with geographical coordinates were entered to the Fuzzy Network-CANFIS model and thus simulation of groundwater quality was accessed in the study area. Finally, the simulated GWQI indices using the Fuzzy Network-CANFIS model were entered into GIS, and hence groundwater quality map (raster layer) based on the results of the network simulation was earned. The study’s results confirm the high efficiency of incorporation of neuro-fuzzy techniques and GIS. It is also worth noting that the general quality of the groundwater in the most studied plain is fairly low.

Fuzzy supervised online coactive neuro‐fuzzy inference system‐based rotor position control of brushless DC motor

In this study, fuzzy supervised online coactive neuro-fuzzy inference system (CANFIS)-based rotor position controller is presented for brushless DC (BLDC) motor. An online learning algorithm is employed for updating premises and consequent parameters of the CANFIS controller. The rotor position control of BLDC motor is simulated using MATLAB/Simulink Toolbox. The dynamic response of the BLDC motor with proposed controller is measured for standard sinusoidal reference input. The effectiveness of the proposed controller performance is compared with fuzzy proportional-integral derivative controller, adaptive neuro-fuzzy inference system controller and supervised recurrent fuzzy neural network controller. The proposed controller is able to solve the problem of non-linearities and uncertainty due to reference input changes of BLDC motor and guarantees fast and accurate dynamic response to a remarkable steady-state performance. Also, experimental hardware results are presented to demonstrate the validity and effectiveness of the proposed control scheme using field programmable gate array chip. Experimental results show that the proposed control scheme can achieve a more favourable tracking performance without the chattering phenomena in the control effort.

Soil temperature estimation using an artificial neural network and co-active neuro-fuzzy inference system in two different climates

Soil temperature is an important meteorological parameter which determines the rates of physical, chemical, and biological reactions in the soil. However, measured values are very sparse in space and time and often not available for a given site. In this study, two intelligent neural models including artificial neural networks (ANNs) and co-active neuro-fuzzy inference system (CANFIS) were used for the estimation of soil temperatures at six depths (5, 10, 20, 30, 50, and 100 cm) with minimum input data (mean air temperature). For this purpose, use was made of the 14-year meteorological data obtained for the two regions of Gorgan in northern Iran with a humid climate and Zabol in southeastern Iran with a dry climate. Comparisons of the model performances in arid and humid regions showed that both ANNs and CANFIS models performed better in arid regions. The accuracy of the soil temperature predictions by both ANNs and CANFIS models gradually decreased from the surface down to the various depths. The results also indicated the capabilities of the ANNs in predicting soil temperature in arid and humid regions.

Estimating Sediment Discharge Using Sediment Rating Curves and Artificial Neural Networks in the Shiwen River, Taiwan

Sediment in river is usually transported during extreme events related to intense rainfall and high river flows. The conventional means of collecting data in such events are risky and costly compared to water discharge measurements. Hence, the lack of sediment data has prompted the use of sediment rating curves (SRC). The aim of this study is to explore the abilities of artificial neural networks (ANNs) in advancing the precision of stream flow-suspended discharge relationships during storm events in the Shiwen River, located in southern Taiwan. The ANNs used were multilayer perceptrons (MLP), the coactive neurofuzzy inference system model (CANFISM), time lagged recurrent networks (TLRN), fully recurrent neural networks (FRNN) and the radial basis function (RBF). A comparison is made between SRC and the ANNs. Hourly based water and sediment discharge during 8 storms were manually collected and used as inputs for the SRC and the ANNs. Results have shown that the ANN models were superior in reproducing hourly sediment discharge compared to SRC. The findings further suggest that MLP can provide the most accurate estimates of sediment discharge, (R2 of 0.903) compared to CANFISM, TLRN, FRNN and RBF. SRC had the lowest R2 (0.765), and resulted in underestimations of peak sediment discharge (−47%).