ANFIS optimaization Research Papers - Academia.edu (original) (raw)

Despite of acquiring popularity among researchers, the implementations of ANFIS-based models face problems when the number of rules surge dramatically and increase the network complexity, which consequently adds computational cost. Essentially, not all the rules in ANFIS knowledge-base are the potential ones. They contain those rules which have either minor or no contribution to overall decision. Thus, removing such rules will not only reduce complexity of the network, but also cut computational cost. Thus, there are various rule-base optimization techniques, proposed in literature, which are presented in motivation to simultaneously obtain rule-base minimization and accuracy maximization. This paper analyzes some of those approaches and important issues related to achieving both the contradictory objectives simultaneously. In this paper, Hyperplane Clustering, Subtractive Clustering, and the approach based on selecting and pruning rules are analyzed in terms of optimizing ANFIS rule-base. The optimized rule-base is observed in connection with providing high accuracy. The results and analysis, presented in this paper, suggest that the clustering approaches are proficient in minimizing ANFIS rule-base with maximum accuracy. Although, other approaches, like putting threshold on rules’ firing strength, can also be improved using metaheuristic algorithms.

- by
- •
- Optimization techniques, ANFIS optimaization, Fuzzy and Neuro-Fuzzy Systems, ANFIS

Adaptive neural fuzzy inference system (ANFIS) is an intelligent neuro-fuzzy technique used for modeling and control of uncertain systems. In this paper, we proposed an ANFIS based modeling approach (called MLANFIS) where the number of data pairs employed for training was adjusted by application of clustering method. By employing this method, the number of data required for learning step and thereby its complexity were significantly reduced. The results obtained were compared with those obtained by using artificial neural networks (ANNs). Inputs to the first group were feed supply, fuel and machinery and the ones to second cluster were pullet, electricity and labor energies. Finally, the outputs of aforementioned networks were considered as inputs to ANFIS 3 network and predicted values of egg yield were derived. The coefficient of determination (R 2), root mean square error (RMSE) and mean absolute percentage error (MAPE) parameters of ANFIS 3 network were calculated as 0.92, 448.126, 0.014, respectively showing that ANFIS compared with ANNs with statistical parameters as 0.81, 751.96 and 0.019 respectively, can properly predict the egg yield of poultry farms. As a recommendation for future studies, ANFIS models with multi-layered structures can be developed to find the optimum number of layers.

- by K.W. Chau
- •
- Poultry Science, Energy Consumption, Energy Production & Consumption, Poultry Nutrition

This research suggests new experimental outcomes regarding the viscosity and thermal conductivity of silver, copper and titanium oxide nanoparticles dispersed in mineral insulating oil by high-pressure homogenization process without using any additives or surfactants. Later, via employing non-linear regression, an adaptive neuro-fuzzy inference system (ANFIS) and achieved experimental data, new models were evolved to predict the viscosity besides thermal conductivity of nanofluids. For modelling, viscosity as well as thermal conductivity of nanofluids was picked as the target factor, and the volume concentration in addition to types of nanoparticles was regarded as the design (input) factors and all experimental data was classified into a train and a test data set. The model was conducted through the train set and the outcomes were contrasted with the experimental data set. Predicted thermal conductivities as well as viscosities were compared with experimental data for three different nanofluids, having nanoparticles volume concentrations of 0.00125% and 0.050%. A comparison was made between the ANFIS and regression outcomes. To evaluate the results, the coefficient of determination (R 2) and root-mean-square error (RMSE) are reported. The achieved results of this research indicate that thermal conductivity of nanofluids enhance by nanoparticles concentration increment. Thermal conductivity of silver is higher compared to thermal conductivity of titanium oxide and copper nanoparticles. According to the ANFIS and non-linear regression outputs, two sets of correlations for calculating the dynamic viscosity as well as thermal conductivity were suggested. Comparing the experimental data with suggested correlations demonstrate very good agreement between the suggested correlations and experimental data. However, equations of previous researches would not be perfectly able to predict the experimental data of present study.

- by Mohammad Reza Safaei and +1
- •
- Nanomaterials Characterization, Heat Transfer, Nanomaterials, Nanoparticles

Dengue Hemorrhagic Fever is one of the dangerous infectious diseases that can cause death within a short time and often cause epidemic. The spread of dengue fever outbreaks globally with frequency levels tend to be higher during the period of last 50 years gave rise to an idea that systematic prevention. The purpose of this paper was to design an application to predict the number of dengue hemorrhagic fever patients with ANFIS method. Weather factors such as air humidity, air temperature, rainfall and number of rain days is used as the factors that influence the incidence of dengue hemorrhagic fever. In this paper using three methods for establishment of FIS: Grid Partition, Substractive Clustering and Fuzzy C Means. By simulating three methods for maximum predicted results, it was found that the ANFIS method with Grid Partition as the establishment of FIS is the best model to generate value with the smallest RMSE testing is 0.71. It indicates That ANFIS models is well proven to be used in predicting The cases of dengue fever.

This paper presents an intelligent control technique for the Maximum Power Point Tracking (MPPT) of a photovoltaic (PV) system using adaptive neuro-fuzzy inference system (ANFIS) under variable solar irradiation conditions. The MXS 60 PV Module specifications is considered for the analysis and models of solar PV module and a DC/DC Boost converter are developed in MATLA/SIMULINK environment. Initially, an MPPT controller is designed using Perturb and Observe (P&O) method. However, this conventional method cannot track rapid changes in the solar irradiation effectively. Hence, an intelligent controller is designed using ANFIS which draws much energy and fast response under continuously changing operating conditions. The PV module with proposed MPPT controller is analyzed in stand-alone mode. The major disadvantage with PV system is its uncertain and intermittent power output which depends on weather conditions. PV module alone cannot supply reliable power to the isolated load effectively. To overcome this, PV module can be connected to the grid. It serves two purposes; in case of deficit solar irradiation, power can be taken from the grid and when there is surplus irradiation, power can be fed to the grid. In order to predict the power supplied to the load and grid under different operating conditions sensitivity analysis has been carried out for the PV system with designed MPPT controller is simulated using HOMER Pro Software.