• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.5
• /
• pp.705-711
• /
• 2013

In this paper, we introduce an design of multi-output fuzzy neural networks based on Interval Type-2 fuzzy set. The proposed Interval Type-2 fuzzy set-based fuzzy neural networks with multi-output (IT2FS-based FNNm) comprise the network structure generated by dividing the input space individually. The premise part of the fuzzy rules of the network reflects the individuality of the division space for the entire input space and the consequent part of the fuzzy rules expresses three types of polynomial functions with interval sets such as constant, linear, and modified quadratic inference for pattern recognition. The learning of fuzzy neural networks is realized by adjusting connections of the neurons in the consequent part of the fuzzy rules, and it follows a back-propagation algorithm. In addition, in order to optimize the network, the parameters of the network such as apexes of membership functions, uncertainty factor, learning rate and momentum coefficient were automatically optimized by using real-coded genetic algorithm. The proposed model is evaluated with the use of numerical experimentation.

• Earthquakes and Structures
• /
• v.12 no.2
• /
• pp.153-163
• /
• 2017

The underlying goal of the present paper is to investigate soil and structural uncertainties on impedance functions and structural response of soil-shallow foundation-structure (SSFS) system using Monte Carlo simulations. The impedance functions of a rigid massless circular foundation resting on the surface of a random soil layer underlain by a homogeneous half-space are obtained using 1-D wave propagation in cones with reflection and refraction occurring at the layer-basement interface and free surface. Firstly, two distribution functions (lognormal and gamma) were used to generate random numbers of soil parameters (layer's thickness and shear wave velocity) for both horizontal and rocking modes of vibration with coefficients of variation ranging between 5 and 20%, for each distribution and each parameter. Secondly, the influence of uncertainties of soil parameters (layer's thickness, and shear wave velocity), as well as structural parameters (height of the superstructure, and radius of the foundation) on the response of the coupled system using lognormal distribution was investigated. This study illustrated that uncertainties on soil and structure properties, especially shear wave velocity and thickness of the layer, height of the structure and the foundation radius significantly affect the impedance functions, and in same time the response of the coupled system.

• Structural Monitoring and Maintenance
• /
• v.3 no.1
• /
• pp.33-49
• /
• 2016

Structural Health Monitoring (SHM) has been attracting numerous research efforts around the world because it targets at monitoring structural conditions and performance to prevent catastrophic failure, and to provide quantitative data for engineers and infrastructure owners to design a reliable and economical asset management strategy. In the past decade, with supports from Australian Research Council (ARC), Cooperative Research Center for Infrastructure and Engineering Asset Management (CIEAM), CSIRO and industry partners, intensive research works have been conducted in the School of Civil, Environmental and Mining Engineering, University of Western Australia and Centre for Infrastructural Monitoring and Protection, Curtin University on various techniques of SHM. The researches include the development of hardware, software and various algorithms, such as various signal processing techniques for operational modal analysis, modal analysis toolbox, non-model based methods for assessing the shear connection in composite bridges and identifying the free spanning and supports conditions of pipelines, vibration based structural damage identification and model updating approaches considering uncertainty and noise effects, structural identification under moving loads, guided wave propagation technique for detecting debonding damage, and relative displacement sensors for SHM in composite and steel truss bridges. This paper aims at summarizing and reviewing the recent research advances on SHM of civil infrastructure in Western Australia.

• Smart Structures and Systems
• /
• v.20 no.2
• /
• pp.207-217
• /
• 2017

Because of the inevitable uncertainties such as structural parameters, external excitations and measurement noises, the effects of uncertainties should be taken into consideration in structural damage detection. In this paper, two probabilistic structural damage detection approaches are proposed to account for the underlying uncertainties in structural parameters and external excitation. The first approach adopts the statistical moment-based structural damage detection (SMBDD) algorithm together with the sensitivity analysis of the damage vector to the uncertain parameters. The approach takes the advantage of the strength SMBDD, so it is robust to measurement noise. However, it requests the number of measured responses is not less than that of unknown structural parameters. To reduce the number of measurements requested by the SMBDD algorithm, another probabilistic structural damage detection approach is proposed. It is based on the integration of structural damage detection using temporal moments in each time segment of measured response time history with the sensitivity analysis of the damage vector to the uncertain parameters. In both approaches, probability distribution of damage vector is estimated from those of uncertain parameters based on stochastic finite element model updating and probabilistic propagation. By comparing the two probability distribution characteristics for the undamaged and damaged models, probability of damage existence and damage extent at structural element level can be detected. Some numerical examples are used to demonstrate the performances of the two proposed approaches, respectively.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.6 no.5
• /
• pp.1400-1420
• /
• 2012

Object detection and tracking using visual sensors is a critical component of surveillance systems, which presents many challenges. This paper addresses the enhancement of object detection and tracking via the combination of multiple visual sensors. The enhancement method we introduce compensates for missed object detection based on the partial detection of objects by multiple visual sensors. When one detects an object or more visual sensors, the detected object's local positions transformed into a global object position. Local and global information exchange allows a missed local object's position to recover. However, the exchange of the information may degrade the detection and tracking performance by incorrectly recovering the local object position, which propagated by false object detection. Furthermore, local object positions corresponding to an identical object can transformed into nonequivalent global object positions because of detection uncertainty such as shadows or other artifacts. We improved the performance by preventing the propagation of false object detection. In addition, we present an evaluation method for the final global object position. The proposed method analyzed and evaluated using case studies.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.05a
• /
• pp.47-50
• /
• 2000

In this paper, an optimal identification method using Multi-FNN(Fuzzy-Neural Network) is proposed for model ins of nonlinear complex system. In order to control of nonlinear process with complexity and uncertainty of data, proposed model use a HCM clustering algorithm which carry out the input-output data preprocessing function and Genetic Algorithm which carry out optimization of model. The proposed Multi-FNN is based on Yamakawa's FNN and it uses simplified inference as fuzzy inference method and Error Back Propagation Algorithm as learning rules. HCM clustering method which carry out the data preprocessing function for system modeling, is utilized to determine the structure of Multi-FNN by means of the divisions of input-output space. Also, the parameters of Multi-FNN model such as apexes of membership function, learning rates and momentum coefficients are adjusted using genetic algorithms. Also, a performance index with a weighting factor is presented to achieve a sound balance between approximation and generalization abilities of the model, To evaluate the performance of the proposed model, we use the time series data for gas furnace and the numerical data of nonlinear function.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.12
• /
• pp.2313-2321
• /
• 2008

In order to develop reliable on-site partial discharge(PD) pattern recognition algorithm, the fuzzy neural network based on fuzzy set(FNN) and the polynomial network pattern classifier based on fuzzy Inference(PNC) were investigated and designed. Using PD data measured from laboratory defect models, these algorithms were learned and tested. Considering on-site situation where it is not easy to obtain voltage phases in PRPDA(Phase Resolved Partial Discharge Analysis), the measured PD data were artificially changed with shifted voltage phases for the test of the proposed algorithms. As input vectors of the algorithms, PRPD data themselves were adopted instead of using statistical parameters such as skewness and kurtotis, to improve uncertainty of statistical parameters, even though the number of input vectors were considerably increased. Also, results of the proposed neuro-fuzzy algorithms were compared with that of conventional BP-NN(Back Propagation Neural Networks) algorithm using the same data. The FNN and PNC algorithms proposed in this study were appeared to have better performance than BP-NN algorithm.

• Computers and Concrete
• /
• v.4 no.5
• /
• pp.347-362
• /
• 2007

Due to heterogeneities at all scales, concrete exhibits significant variability in mechanical behavior from sample to sample. An understanding of the fundamental mechanical performance of concrete must therefore be embedded in a stochastic framework. The current work attempts to address the connection between a two-dimensional concrete mesostructure and the random local material properties associated within that mesostructure. This work builds on previous work that has focused on the random configuration of concrete mesostructures. This was accomplished by developing an understanding of the effects of variations in the mortar strength and the mortar-aggregate interfacial strength in given deterministic mesostructural configurations. The results are assessed through direct tension tests that are validated by comparing experimental results of two different, pre-arranged mesostructures, with the intent of isolating the effect of local variations in strength. Agreement is shown both in mechanical property values as well as the qualitative nature of crack initiation and propagation.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.23 no.6
• /
• pp.329-339
• /
• 2019

The strategy for the management of earthquakes is shifting from post recovery to prevention; therefore, seismic performance management requires quantitative predictions of damage and the establishment of strategies for initial responses to earthquakes. Currently, seismic performance evaluation for seismic management in Korea consists of two stages: preliminary evaluation and detailed evaluation. Also, the priority of seismic performance management is determined in accordance with the preliminary evaluation. As a deterministic method, preliminary evaluation quantifies the physical condition and socio-economic importance of a facility by various predetermined indices, and the priority is decided by the relative value of the indices; however, with the deterministic method it is difficult to consider any uncertainty related to the return-year, epicenter, and propagation of seismic energy. Also this method cannot support tasks such as quantitative socio-economic damage and the provision of data for initial responses to earthquakes. Moreover, indirect damage is often greater than direct damage; therefore, a method to quantify damage is needed to enhance accuracy. In this paper, a Seismic Risk Assessment is used to quantify the cost of damage of road facilities in Pohang city and to support decision making.

• Smart Structures and Systems
• /
• v.24 no.6
• /
• pp.733-744
• /
• 2019

Wind speed forecasting is critical for a variety of engineering tasks, such as wind energy harvesting, scheduling of a wind power system, and dynamic control of structures (e.g., wind turbine, bridge, and building). Wind speed, which has characteristics of random, nonlinear and uncertainty, is difficult to forecast. Nowadays, machine learning approaches (generalized regression neural network (GRNN), back propagation neural network (BPNN), and extreme learning machine (ELM)) are widely used for wind speed forecasting. In this study, two schemes are proposed to improve the forecasting performance of machine learning approaches. One is that optimization algorithms, i.e., cross validation (CV), genetic algorithm (GA), and particle swarm optimization (PSO), are used to automatically find the optimal model parameters. The other is that the combination of different machine learning methods is proposed by finite mixture (FM) method. Specifically, CV-GRNN, GA-BPNN, PSO-ELM belong to optimization algorithm-assisted machine learning approaches, and FM is a hybrid machine learning approach consisting of GRNN, BPNN, and ELM. The effectiveness of these machine learning methods in wind speed forecasting are fully investigated by one-year field monitoring data, and their performance is comprehensively compared.