Earticle

This study compared the performance of the FCM(C-Means)-based ANFIS(Adaptive Neuro-Fuzzy Inference System) model and the CFCM(Context-based Fuzzy C-Means) clustering-based ANFIS model. The FCM-ANFIS model sets the initial Fuzzy Rule through FCM clustering and optimizes the rule through neural network learning. The CFCM-ANFIS model generates more sophisticated rules through CFCM clustering that considers the input and output variable space and learns the neural network. As a result of the experiment, the verification RMSE of the FCM-based ANFIS model was 3.5654 when the number of clusters was 6, and the RMSE of the CFCM clustering-based ANFIS model was 3.3954 in the parameters (P = 6, C = 2), which was higher than the FCM-based ANFIS model. It was confirmed that the CFCM method had better prediction performance than the FCM method, and this study proved that the CFCM-based ANFIS model was more effective in predicting body fat percentage.

초고성능콘크리트(UHPC) 휨부재의 안전한 설계를 위해서는 전단강도에 대한 정확한 평가가 필요하다. 지금까지 UHPC 휨부재의 전단강도에 대한 많은 연구가 수행되었지만, 전단강도가 다양한 인자들에 의하여 영향을 받기 때문에 아직도 UHPC 휨부재의 전단강도를 정확하게 평가하는 것은 매우 어렵다. 이 연구에서는 UHPC 휨부재의 전단강도를 평가하기 위하여 뉴로-퍼지추론 시스템(ANFIS)을 도입하고 기존 실험결과를 수집하여 ANFIS를 훈련시켰다. 훈련된 ANFIS 알고리즘을 통하여 해석된 결과를 프랑스 토목학회(AFGC-SETRA)와 일본 토목학회(JSCE) 설계기준식에 의하여 산정된 결과와 비교하였으며, ANFIS 모델은 평균 0.98, COV 0.15으로써 기준식들에 비해 매우 우수한 정확도를 나타내었다.

For safe design of the reinforced concrete flexural members cast with Ultra High Performance Concrete (UHPC), their shear strengths shall be estimated accurately. Many studies have been performed on the shear strengths of UHPC flexural members so far, but it is still very challenging to estimate their shear strengths accurately because they are considerably influenced by various factors. In this study, an Adaptive Neuro-Fuzzy Inference System (ANFIS) was introduced to estimate their shear strengths, and trained by the test data collected from previous studies. The analysis results by the trained ANFIS algorithm were compared with those calculated by Association Francaise de Genie Civil (AFGC) and Japan Society of Civil Engineers (JSCE) Recommendations, and the ANFIS model provided much better accuracy with an average of 0.98 and a COV of 0.15 than the code provisions.

This study analyzes the performance of ANFIS(Adaptive Neuro-Fuzzy Inference System) based on the input space partitioning method. Using body fat datasets and concrete compressive strength datasets, various fuzzy system configuration methods, such as Grid Partitioning, Subtractive Clustering, and FCM(Fuzzy C-Means), are compared. The results show that the FCM-based ANFIS model demonstrated superior performance, recording the lowest RMSE value. It is confirmed that the initialization method of the fuzzy system significantly influences the performance of ANFIS, and the optimal configuration method may vary depending on the data distribution and complexity.

The scientific and technological achievements trigger people’s desire to learn new things and sometimes it turns into a competition among individuals. The competitive environmental so makes an important psychological pressure on individuals and it may manipulate their choices about their educational careers. Artificial neural networks are widely utilized to reach to a short-cut solution and as a method of decision making. The role of the machine learning techniques and data mining algorithms to define the factors affecting students’ success is important. The aim of this research is to make predictions about these factors by using Adaptive-Network Based Fuzzy Inference Systems (ANFIS). Student Performance data set in UCI platform is used for classification part of this study.

철근콘크리트 부재의 전단전달 매커니즘이 매우 복잡하기 때문에 이론적 접근방법이 매우 어렵고 많은 불확실성을 내포하고 있다. 이러한 이유로 설계기준들을 포함한 여러 연구들에서는 대부분 반경험적으로 결정된 전단강도 평가식을 사용하고 있으며, 철근콘크리트 부재의 전단거동에 대한 이해 및 접근 방법은 더욱 제한적이다. 따라서, 이 연구에서는 발전된 뉴로-퍼지 정보시스템(Neuro-Fuzzy information system)인 ANFIS (Adaptive nuero-fuzzy inference system)를 철근콘크리트 패널부재의 전단거동평가에 도입하였으며, 역전파알고리즘을 통하여 전단저항 메커니즘이 내포하고 있는 불확실성을 크게 감소시킬 수 있는 ANFIS 해석모델을 도출하였다. ANFIS의 훈련과정을 통하여 최적화된 ANFIS 전단거동해석모델에 의한 해석결과와 실험결과와의 비교를 통하여 분석한 결과, 제안모델은 패널 실험체들의 전단거동을 매우 근사하게 모사하였으며, 균열전 탄성거동, 전단균열강도, 균열후 거동, 극한강도 뿐만 아니라 일부 실험체들에서 관측된 최대강도이후의 연화거동도 잘 평가하는 것으로 나타났다.

The complicated shear transfer mechanisms in reinforced concrete (RC) members make it hard to develope theoritical approaches and have many uncertainties. Consequently, shear strength estimation models in the various design codes and existing studies mostly have semi-empirical basis, and our understanding and approaches in shear behavior of RC members are even more limited. Thus, ANFIS(Adaptive neuro-fuzzy inference system), an extended Neuro-Fuzzy information system, was introduced to evaluate the shear behavior of reinforced concrete panel members, and many inherent uncertainties in shear transfer mechanisms were reduced by back propagation method in the ANFIS analysis model developed in this study. The ANFIS shear behavior model optimized through the training process were verified by comparison of experimental test results, which showed that it properly estimated overall shear behavior of panel specimens including elastic behavior before cracking, shear cracking strength, post-cracking behavior, ultimate strength as well as softening effects in post-peak behavior, if any.

We are sometimes interacting with people who we know nothing and facing with the difficult task of making decisions involving risk in social network. To reduce risk, the topic of building Web of trust is receiving considerable attention in social network. The easiest approach to build Web of trust will be to ask users to represent level of trust explicitly toward another users. However, there exists sparsity issue in Web of trust which is represented explicitly by users as well as it is difficult to urge users to express their level of trustworthiness. We propose a fuzzy-based inference model for Web of trust using user behavior information in social network. According to the experiment result which is applied in Epinions.com. the proposed model show improved connectivity in resulting Web of trust as well as reduced prediction error of trustworthiness compared to existing computational model.

Mine Blast Algorithm (MBA) is newly developed metaheuristic technique. It has outperformed Genetic Algorithm (GA), Particle Swarm Optimization (PSO) and their variants when solving various engineering optimization problems. MBA has been improved by IMBA, which is modified in this paper to accelerate its convergence speed furthermore. The proposed variant, so called Accelerated MBA (AMBA), replaces the previous best solution with the available candidate solution in IMBA. ANFIS accuracy depends on the parameters it is trained with. Keeping in view the drawbacks of gradients based learning of ANFIS using gradient descent and least square methods in two-pass learning algorithm, many have trained ANFIS using metaheuristic algorithms. In this paper, for getting high performance, the parameters of ANFIS are trained by the proposed AMBA. The experimental results of real-world benchmark problems reveal that AMBA can be used as an efficient optimization technique. Moreover, the results also indicate that AMBA converges earlier than its other counterparts MBA and IMBA.

Fuzzy Neural Networks (FNNs) techniques have been effectively used in applications that range from medical to mechanical engineering, to business and economics. Despite of attracting researchers in recent years and outperforming other fuzzy systems, Adaptive Neuro-Fuzzy Inference System (ANFIS) still needs effective parameter training and rule-base optimization methods to perform efficiently when the number of inputs increase. Moreover, the standard gradient based learning via two pass learning algorithm is prone slow and prone to get stuck in local minima. Therefore many researchers have trained ANFIS parameters using metaheuristic algorithms however very few have considered optimizing the ANFIS rule-base. Mostly Particle Swarm Optimization (PSO) and its variants have been applied for training approaches used. Other than that, Genetic Algorithm (GA), Firefly Algorithm (FA), Ant Bee Colony (ABC) optimization methods have been employed for effective training of ANFIS networks when solving various problems in the field of business and finance.

Basically, there are six types of warranty that can be offered either by manufacturer or dealers of a product or services which are basic warranty, extended warranty, warranty for used, repair limit warranty, service warranty and lifetime warranty. The different types of warranty policy have been established in order to fulfill the demand of manufacturers and the requirement of buyers so that a win-win situation could be acquired. However, when considering these warranty policies, an important concept to keep in mind is warranty reclaims. Warranty reclaims extend the scope of warranty activities beyond the walls of a single company to encompass suppliers, manufacturers, OEMs, distributors, dealers, repair centers, policy carriers, and customers. This paper presents a methodology to adapt historical maintenance warranty reclaims data with Adaptive Neuro Fuzzy Inference System (ANFIS) approach. The main motivations for conducting this paper are simplicity and less computational mass of the ANFIS based on linear generating functions with regard to warranty reclaims prediction.

Comparing to the big volume, large weight and high power consumption of the conventional samplers which are fixed on the lunar rover, the paper firstly described a novel flexible mini lunar sampling robot. Then the nonlinear dynamics resonance broken system is built to model the contact between the sampling robot and the lunar regolith. It is found to be suitable for drilling when the sampling robot is in the resonance condition. For the nonlinear time-varying system of the dynamic modeling of the sampler in drilling, we presented the method of the frequency neural-fuzzy adaptive control based on the dynamic resonant frequency prediction of the flexible sampling robot using neural networks. Firstly the algorithm predicts the dynamic resonant frequency of the sampling robot by GRNN. Then a neural-fuzzy adaptive control system is established, in which the frequency prediction error, the amplitude and its variable are adopted as the input and the sweep frequency bandwidth as the output, to adjust the frequency bandwidth dynamically. What’s more, the simulation results verify the effectiveness of the control strategy. Finally, the experimental results show that the control algorithm can improve the drilling depth, drilling efficiency and the discarding efficiency by 66.7%, 65.2% and 67.4%, respectively, in stimulant lunar regolith.

A smart home refers to a house or building equipped with advanced systems that allow users to remotely control various electronic devices. This paper proposes a novel approach utilizing an Adaptive Network-Based Fuzzy Inference System (ANFIS) to address the nonlinear challenges faced by conventional sensor-based smart home security systems. The proposed system uses multiple sensors to detect internal threats (e.g., fire, gas leaks) and external threats (e.g., theft, intrusion), analyzing sensor data to identify abnormalities. Designed with a multi-output ANFIS model, the system dynamically responds to various scenarios and provides optimal security decisions. Additionally, it features real-time transmission of security analysis results to relevant agencies or users via the internet. The findings demonstrate that ANFIS significantly enhances the efficiency and accuracy of smart home security systems compared to traditional fuzzy logic-based methods.

Trend Impact Analysis is a prominent hybrid method has been used in future studies with a modified surprise-free forecast. It considers experts' perceptions about how future events may change the surprise-free forecast. It is an advanced forecasting tool used in futures studies for identifying, understanding and analyzing the consequences of unprecedented events on future trends. In this paper, we propose an advanced mechanism to generate more justifiable estimates to the probability of occurrence of an unprecedented event as a function of time with different degrees of severity using adaptive neuro-fuzzy inference system (ANFIS). The key idea of the paper is to enhance the generic process of reasoning with fuzzy logic and neural network by adding the additional step of attributes simulation, as unprecedented events do not occur all of a sudden but rather their occurrence is affected by change in the values of a set of attributes. An ANFIS approach is used to identify the occurrence and severity of an event, depending on the values of its trigger attributes.

Vehicle License Plate Images Segmentation is a substantial stage for developing an Automatic License Plate Recognition (ALPR) system. In this paper, it is considered an efficient segmentation algorithm for extracting vehicle license plate images using Cellular Neural Networks (CNN). The learning CNN templates values are formulated as an optimization problem to achieve the desired performances which can be found by means of Adaptive Fuzzy (AF) algorithm and Neuro-Fuzzy (NF) algorithm techniques. The main objective of the paper is to compare the performances of standard CNN, Adaptive Fuzzy (AF), and Neuro-Fuzzy (NF) on real data of several vehicle license plate images of standard Indonesia License Plates. The results are then compared with ideal vehicle license plate images. Quantitative analysis between ideal vehicle license plate images and segmented vehicle license plate images is presented in terms of Peak signal-to-noise ratio (PSNR), Mean Squared Error (MSE) and Root Mean Squared Error (RMSE). From the performance analysis, the CNN template optimized by ANFIS algorithm is more recommended than the standard CNN edge detector or the CNN template optimized by Adaptive Fuzzy algorithm in vehicle license plate image segmentation. It is shown from the calculation that PSNR is 80% better than the standard CNN, and the resulted MSE and RMSE are 70% better than the standard CNN. Whereas the CNN template optimized by Adaptive Fuzzy algorithm achieves the PSNR 90% better than the standard CNN, but it yields the MSE and RMSE 40% worse than the standard CNN.

Wind power energy is receiving attention in recent years. The properties of wind are very hard to predict because of its heavy nonlinear characteristics. This paper predicts the wind speed by ANFIS and FCM clustering. The data were measured in the region of islands in Jeonnam Shinan. One year and 10 minute interval makes 52,560 samples of data but use 48,240 samples instead for stable operation. For prediction of wind speed, the covariance was examined. As a result, the input domain consists of lunar date and wind direction. This input domain has so big range of wind direction and lunar date. Therefore the whole range is partitioned by clusters. For experiments, two type are chosen. one is 4 clusters and the other is 6 clusters. . The error of cluster-6 is 7.5 % lower than cluster-4. This means that the prediction of cluster-6 is more accurate than cluster-4. With four Gaussian bell membership functions, ANFIS is trained over 200 epochs by clustered data. After training, ANFIS could predict the wind speed by lunar date and wind direction. Even if heavy nonlinear system can be predicted by ANFIS and FCM clustering.

This research contains the simulation and designing of Quadcopter using Adaptive Neuro Fuzzy Controller to control the altitude of quadcopter and obstacle detection. Now a day’s advancement in technology has made it possible to develop low power and lightweight with accurate sensors which are used with controllers for controlling, which have high processing power but small power consumption. This has been allowed for the development of complex and difficult control systems that can be implemented onboard UAV. With this combination of high precision and light weight, real-time onboard navigation or guidance and autonomous flights are now practical. This research work used a Fuzzy controller to control the pitch angle of quadcopter and avoiding obstacles. The fuzzy controller receives the sensory data and adjust the pitch accordingly until unless it finds the clear path. For detecting obstacles the IR sensors are used. For designing a fuzzy inference system used Sugeno model and used mat lab commands to design ANFIS as we have another method for designing by using a Simulink as well. ANFIS designed is based on kinematics and dynamics equation of quadcopter that will be able to control the pitch of quadcopter. Simulations results in mat lab show that by using ANFIS the performance of Quadcopter will be improved significantly.

In this paper we presented an architecture and basic learning process underlying in fuzzy inference system and adaptive neuro fuzzy inference system which is a hybrid network implemented in framework of adaptive network. In real world computing environment, soft computing techniques including neural network, fuzzy logic algorithms have been widely used to derive an actual decision using given input or output data attributes, ANFIS can construct mapping based on both human knowledge and hybrid learning algorithms. This study involves study of ANFIS strategy ANFIS strategy is employed to model nonlinear functions, to control one of the most important parameters of the induction machine and predict a chaotic time series, all yielding more effective, faster response or settling times.

This paper proposes a hybrid learning algorithm for tuning an Adaptive Network based Fuzzy Inference System ANFIS. The proposed scheme of adapting parameters in ANFIS employs evolutionary techniques PSO and GA to adjust the antecedent parameters. The least-squares (LSE) algorithm is used to adjust consequent parameters. The number of fuzzy rules is fixed and given by using the Xie Beni’s index. This new approach is applied to identify and control nonlinear systems with an on-line strategy. The obtained results are compared to similar ANFIS using gradient descent method GD as antecedent parameters of training algorithm and other methods applied in the same problems.

Several models have been created for Smart Grid resource-allocation problem. The principal purpose of the models is to connect power sources with appropriate sinks when considering the input parameters of power balance and consumption size, etc. Fuzzy logic is representative of these models. When creating the fuzzy model, the parameters and rule construction play the most significant role. For the fuzzy Logic model, the rule base has been constructed considering the operator’s general knowledge, the operator’s activity and the experience. However, the fuzzy model did not have any clear boundary for the price, power, and distance values, so the model output was largely dependent on the contributions from individual portions. This paper introduces an Adaptive Neuro Fuzzy Inference System (ANFIS) approach to the smart grid problem. Learning is another attribute that can be incorporated into the current model. ANFIS is one way that the current model can be extended to. In ANFIS, the learning is done with the incorporation of a neural network. The Fuzzy system is trained over time to become self-adaptive. Once the training is complete, the system is capable of making intelligent changes based on a neural network. As a result, ANFIS has a distinct boundary for each segment. Hence, a little change along the boundary value pushed the value into the next segment.