Adaptive Neuro-fuzzy Inference System Architecture Research Articles

Experimental study and ANFIS modelling of the thermophysical properties and efficacy of GNP-Al2O3 hybrid nanofluids of different concentrations and temperatures

This study delves into an extensive investigation of the thermophysical properties and heat transfer efficacy of a hybrid nanofluid incorporating graphene nanoplatelets and γ-Al2O3 nanoparticles dispersed in deionised water. The nanofluids were characterised for their viscosity (µ), thermal conductivity (λ), and electrical conductivity (σ) over a 15–40 °C temperature range for varying nanoparticle loading (0.1–0.4 vol%). The experimental results revealed notable enhancements in µ, λ, and σ with increasing nanoparticle concentration, while µ decreased at elevated temperatures as λ and σ increased. At the highest concentration (0.4 vol%), µ increased by 21.74%, while λ and σ exhibited peak enhancements of 17.82% and 393.36% at 40 °C. An Adaptive Neuro-fuzzy Inference System (ANFIS) model was devised to enhance predictive precision by meticulously optimising the number of membership functions (MFs) and input MF type. The ANFIS architecture that exhibited the most remarkable agreement with the experimental data for µ, λ, and σ was found to utilise the Product of Sigmas, Difference of Sigmas, and Generalized Bell MFs, respectively, with corresponding input MF numbers being 2–3, 3–2, and 3–2. The optimal ANFIS model for µ, λ, and σ exhibits a higher prediction accuracy with an R2 value of 0.99965, 0.99424 and 0.99995, respectively. The Figure of Merit analysis using Mouromtseff Number identified an optimal nanoparticle concentration range of 0.1–0.2 vol% for enhanced heat transfer performance with a reasonable µ increase. This range guides practitioners in utilising hybrid nanofluids effectively while managing potential drawbacks.

Prediction of the capacitance of the corn drying process parameter using adaptive- neuro-fuzzy intelligent technique with experimental validation

In this study, innovative, intelligent software based on Adaptive Neuro-Fuzzy Inference System (ANFIS) has been developed to effectively estimate the corn drying process parameters using data from getting experimental study. For this purpose, laboratory-scale corn drier experimental setup was established to get corn drying parameters (air temperature in the different locations, air humidity, and capacitance value). The twelve sensors measure the drying parameters (s1, s2, s3, …, s12 ) such as inlet air temperature T1 with ±0.5 °C sensitivity, relative humidity RH1 with ±0.1 RH, inlet temperatures T2 and Tw2 with the ±0.5 °C, dryer chamber temperature Ti (i = 3,…,7) with the ±0.5 °C, outlet temperatures T8 and Tw8 with the ±0.5 °C, product moisture content with ±0.05 gr. Finally, depending on the twelve sensor parameters, the capacitance value was measured using a capacitive sensor with ±0.05% sensitivity. Accordingly, an Adaptive Neuro-Fuzzy Inference System architecture was developed with twelve inputs and one output. For the training of the Adaptive Neuro-Fuzzy Inference System, 52 data sets from laboratory studies were used, and Adaptive Neuro-Fuzzy Inference System was successfully trained with absolute fraction of variance 0.998 and mean squared error 0.8007. For the testing the performance of the Adaptive Neuro-Fuzzy Inference System 51 dataset was used and testing performance of the ANFIS getting with the absolute fraction of variance 0.9963 and mean squared error 1.47. Finally, the performance of the Adaptive Neuro-Fuzzy Inference System compared with the artificial neural network (ANN), and it is clearly seen that the performance of the fuzzy-neuro hybrid algorithm is higher than the core neuro algorithm with %0.97 for the training and %31.5 for the testing. By forecasting the moisture content of corn under various operating conditions, the Adaptive Neuro-Fuzzy Inference System model can be used to improve the operation of the corn dryer, resulting in increased efficiency, lower energy usage, and higher-quality dried corn.

Predicting infarction growth rate II using ANFIS-based binary particle swarm optimization technique in ischemic stroke

Ischemic stroke, a severe medical condition triggered by a blockage of blood flow to the brain, leads to cell death and serious health complications. One key challenge in this field is accurately predicting infarction growth - the progressive expansion of damaged brain tissue post-stroke. Recent advancements in artificial intelligence (AI) have improved this prediction, offering crucial insights into the progression dynamics of ischemic stroke. One such promising technique, the Adaptive Neuro-Fuzzy Inference System (ANFIS), has shown potential, but it faces the 'curse of dimensionality' and long training times as the number of features increased. This paper introduces an innovative, automatic method that combines Binary Particle Swarm Optimization (BPSO) with ANFIS architecture, achieves reduction in dimensionality by reducing the number of rules and training time. By analyzing the Pearson correlation coefficients and P-values, we selected clinically relevant features strongly correlated with the Infarction Growth Rate (IGR II), extracted after one CT scan. We compared our model's performance with conventional ANFIS and other machine learning techniques, including Support Vector Regressor (SVR), shallow Neural Networks, and Linear Regression.•Inputs: Real data about ischemic stroke represented by clinically relevant features.•Output: An innovative model for more accurate and efficient prediction of the second infarction growth after the first CT scan.•Results: The model achieved commendable statistical metrics, which include a Root Mean Square Error of 0.091, a Mean Squared Error of 0.0086, a Mean Absolute Error of 0.064, and a Cosine distance of 0.074.

Optimal Adaptive Neuro-Fuzzy Inference System Architecture for Time Series Forecasting with Calendar Effect

This paper discusses a procedure for model selection in ANFIS for time series forecasting with a calendar effect. Calendar effect is different from the usual trend and seasonal effects. Therefore, when it occurs, it will affect economic activity during that period and create new patterns that will result in inaccurate forecasts for decision making if not considered. The focus is on the model selection strategy to find the appropriate input variable and the number of membership functions (MFs) based on the Lagrange Multiplier (LM) test. The ARIMAX stochastic model is used at the preprocessing stage to capture calendar variations in the data. The calendar effect observed is the Eid al-Fitr holiday in Indonesia, a country with the largest Muslim population in the world. The data of Tanjung Priok port passengers used as a case study. The result shows that hybrid ARIMAX-ANFIS based on the LM test can be an effective procedure for model selection in ANFIS for time series with calendar effect forecasting. Empirical results show that the use of the calendar effect variable provides more accurate predictions as indicated by smaller RMSE and MAPE values than without the calendar effect variable.

A Multiple Adaptive Neuro-Fuzzy Inference System for Predicting ERP Implementation Success

The implementation of modern ERP solutions has introduced tremendous opportunities as well as challenges into the realm of intensely competent businesses. The ERP implementation phase is a very costly and time-consuming process. The failure of the implementation may result in the entire business to fail or to become incompetent. This fact along with the complexity of data streams has led the researchers to develop a hierarchical multi-level predictive solution to automatically predict the implementation success of ERP solution. This study exploits the strength and precision of the Adaptive Neuro-Fuzzy Inference System (ANFIS) for predicting the implementation success of ERP solutions before the initiation of the implementation phase. This capability is obtained by training the ANFIS system with a data set containing a large number of ERP implementation efforts that have led to success, failure, or a mid-range implementation. In the initial section of the paper, a brief review of the recent ERP solutions as well as ANFIS architecture and validation procedure is provided. After that, the major factors of ERP implementation success are deeply studied and extracted from the previous literature. The major influential implementation factors in the businesses are titled as Change Orchestration (CO), Implementation Guide (IG), and Requirements Coverage (RC). The factors represent the major elements that guide the implementation project to a final success or to a possible failure if mismanaged. Accordingly, three initial ANFIS models are designed, trained, and validated for the factors. The models are developed by gathering data from 414 SMEs located in the Islamic Republic of Iran during a three-year period. Each model is capable of identifying the weaknesses and predicting the successful implementation of major factors. After this step, a final ANFIS model is developed that integrates the outputs of three initial ANFIS models into a final fuzzy inference system, which predicts the overall success of the ERP implementation project before the initiation phase. This model provides the opportunity of embedding the previous precious experiences into a unified system that can reduce the heavy burden of implementation failure.

Identification of the important variables for prediction of individual medical costs billed by health insurance

The cost of health care insurance is one of the most important factors in the health care development. To establish a better health care system, there is a need to estimate the cost of health insurance. The prediction of the cost is one possibility to improve health care development. There is a need for more advanced methods other than traditional regression approaches, because the prediction of the health insurance costs are now a big data problem. To simplify the prediction process in this study, a selection procedure was performed to identify the most important factors for the prediction of the health care insurance costs. Artificial neural network, namely adaptive neuro fuzzy inference system (ANFIS), was used for the identification procedure. ANFIS architecture was employed to model nonlinear relationships between data samples. Five input factors were considered in the analyzing (age of primary beneficiary, insurance contractor gender, Body mass index, Number of children covered by health insurance, and smoking). The obtained results showed that smoking has the highest impact on the cost of health insurance. Moreover, prediction accuracy is acceptable and could be used for future management of health care development.

Prediction of maximum in-cylinder pressure by adaptive neuro-fuzzy inference system method

A widely used method to substituting expensive experimental method in order to optimizing different parameters of technological application of equipment is using of modelling these phenomena by intelligent techniques. Hence, in this paper, an ANFIS (adaptive neuro-fuzzy inference system architecture) model has been used to predict one of the most important of the diesel engine which is cylinder pressure. Measurement of this parameter requires expensive and time consuming methods. Therefore, application of the mathematical method to prediction of this parameter is necessary. The inputs of this model are injection time, engine speed and engine load. The testing performance of the proposed ANFIS model revealed a good predictive capacity to yield acceptable error measures with, R2=0.99 and MSE=6.8. This model is not developed based on complicated mathematical formula and is easy to use. The result of study recommends that the ANFIS model can be successfully used to perdition of cylinder pressure according to effective parameters.

Migration modelling of phthalate from non-alcoholic beer bottles by adaptive neuro-fuzzy inference system.

There are limitations to the basic knowledge regarding various ways by which packaging components migrate into food as well as ways by which various conditions, elements and molecules related to this phenomenon are analysed. This research aimed to model phthalate migration from polyethylene terephthalate bottles containing non-alcoholic beer by performing adaptive neuro-fuzzy inference system (ANFIS) analysis. The data showed that storage temperature, contact surface and storage period correlates with the rate of migration. Migration of phthalate increases with storage duration gradually and reduces under different temperatures and contact surface. Moreover, increased temperature and storage duration resulted in an increase in migration level ranging from 0.6 μg L-1 to 2.9 μg L-1 . In summary, the present study used an ANFIS architecture which consists of three inputs (temperature, surface and storage period), Gauss-bell membership functions for each input variable and one output layer, which represent the migration level. The validation and training models showed an excellent match between the experimental and predicted values of ANFIS. Analysis of the model showed that ANFIS is a powerful tool for predicting phthalate migration from bottles containing non-alcoholic beer. © 2017 Society of Chemical Industry.

Navigational strategy for underwater mobile robot based on adaptive neuro-fuzzy inference system model embedded with shuffled frog leaping algorithm–based hybrid learning approach

In this research article, a novel navigational approach has been introduced for underwater robot based on learning and self-adaptation ability of adaptive neuro-fuzzy inference system. For avoiding obstacles during three-dimensional navigation, two adaptive neuro-fuzzy inference system models have been coupled to find out required change in heading angles of underwater robot in horizontal and vertical planes, respectively. A new hybrid learning scheme has been proposed for adaptive neuro-fuzzy inference system. Here, memetic approach based shuffled frog leaping algorithm has been used to tune the premise parameters and consequent parameters has been estimated through recursive least square estimation. Minimization of error in output of adaptive neuro-fuzzy inference system model has been treated as major objective of evolutionary-based training algorithm. Preliminary robotic behaviors of underwater robot have been successfully executed by implementing such well-trained adaptive neuro-fuzzy inference system architecture within three-dimensional unspecified workspace. Navigational performance of adaptive neuro-fuzzy inference system trained with the proposed hybrid learning algorithm has been compared with other three-dimensional navigational approaches in simulation mode for authentication purpose. Experimental verification has also been carried out to validate the feasibility and efficiency of the proposed navigational strategy.

Artificial Neural Network and Adaptive Neuro-Fuzzy Interface System Modeling of Supercritical CO2 Extraction of Glycyrrhizic Acid from Glycyrrhiza glabra L

Abstract In this study, the extraction of Glycyrrhizic acid (GA) from Glycyrrhiza glabra (licorice) root was investigated by Soxhlet extraction and modified supercritical CO2 with water as co-solvents and 30 min of static extraction time. The high performance liquid chromatography (HPLC) was used to identify and quantitatively determine the amount of extracted GA recovery of supercritical CO2 extraction of GA. The extraction recovery was modeled by adaptive neuro-fuzzy inference system (ANFIS) and artificial neural network (ANN). Different ANFIS networks (by changing the type of membership functions) were compared with evaluation of networks accuracy in GA recovery prediction and subsequently the suitable network was determined. A three-layer artificial neural network was also developed for modeling of GA extraction from licorice plant root. In this regard, different networks (by changing the number of neurons in the hidden layer and algorithm of network training) were compared with evaluation of networks accuracy in extraction recovery prediction. One-step secant back propagation algorithm with six neurons in hidden layer was found to be the most suitable network and the coefficient of determination (R2) was 98.5 %. Gaussian combination membership function (gauss2mf) using 2 membership function to each input was obtained to be optimum ANFIS architecture with mean square error (MSE) of 0.05,0.17 and 0.07 for training, testing and checking data, respectively.

Online detection and measurements of drill wear for the drilling of stainless steel parts

Thrust and torque have been selected in this research for online detection and measurements of drill wear for the drilling of stainless steel parts because cutting forces are closely related to the drilling process and give the best indirect indication of drill conditions. Using thrust and torque information to monitor the drill conditions for control of the drilling process can decrease the operation cost and enhance the product quality. To find the most important feature(s), the feature selection technique is used in this research. Sequential forward search algorithm is used for feature selection. To reduce the dimension of the measurement vector, it is necessary to retain only those components of the extracted features which show a high sensitivity to drill wear and low sensitivity to process parameters. The best feature selected is the peak of torque in the drilling process. Adaptive neuro-fuzzy inference system (ANFIS) is a neuro-fuzzy system. It includes input layer, output layer, and layers between them. ANFIS can construct fuzzy rules with membership functions to generate an input–output pair. A 1 × 6 ANFIS architecture with generalized bell membership function can achieve a success rate of 100 % for the online detection of drill states. A 1 × 6 ANFIS architecture with product of sigmoid membership functions can measure the drill wear online with an error as low as 0.15 %. Furthermore, the detection and measurement of drill wear is performed under different drilling conditions as compared with the training process. This shows that ANFIS has the capability of generalization.

Adaptive Neuro-Fuzzy Inference System for antepartum antenatal care using phonocardiography

This work discusses development of a model based on Adaptive Neuro-Fuzzy Inference System (ANFIS) for evaluation of fetal health status using phonocardiography. The model integrates adaptable fuzzy inputs with a modular neural network to deal with the imprecision and uncertainty in the interpretation of the FHR data from phonocardiographic signals. A zero-order Takagi-Sugeno model is chosen for designing ANFIS architecture. The diagnostic parameters e.g., Baseline FHR, Baseline Variability, Acceleration and Deceleration of the FHR are derived from the fPCG signals for training and testing of the model. The elicited fuzzy rules derived from clinical guidelines and other resources are implemented into the ANFIS expert model. The performance of the ANFIS model is evaluated in terms of sensitivity and overall accuracy. The results have indicated that the ANFIS can be implemented effectively and provides high accuracy for antepartum antenatal care through phonocardiography.

Modeling of dimmable High Power LED illumination distribution using ANFIS on the isolated area

High power light emitting diodes (HP-LEDs) are more suitable for energy saving applications and have becoming replacing traditional fluorescent and incandescent bulbs for its energy efficient. Therefore, HP-LED lighting has been regarded in the next-generation lighting. In this study, illumination distribution of white color HP-LED is modeled by adaptive neuro-fuzzy inference system (ANFIS) approach on the isolated area while LED head is fixed. Subtractive clustering with hybrid learning approach is used to train the realized ANFIS architectures. End of the numerous experiment we finally concluded that, ANFIS could be used to modeling the illumination distribution applications perfectly.

Using fuzzy inference system to improve neural network for predicting hospital wastewater treatment plant effluent

In this study, three types of adaptive neuro fuzzy inference system (ANFIS) and artificial neural network (ANN) were employed to predict suspended solids (SS eff) and chemical oxygen demand (COD eff) in the effluent from a hospital wastewater treatment plant. The results indicated that ANFIS statistically outperforms ANN in terms of effluent prediction. The minimum mean absolute percentage errors of 11.99% and 12.75% for SS eff and COD eff could be achieved using ANFIS. The maximum values of correlation coefficient for SS eff and COD eff were 0.75 and 0.92, respectively. The minimum mean square errors of 0.17 and 19.58, and the minimum root mean square errors of 0.41 and 4.42 for SS eff and COD eff could also be achieved. ANFIS's architecture consists of both ANN and fuzzy logic including linguistic expression of membership functions and if–then rules, so it can overcome the limitations of traditional neural network and increase the prediction performance.