• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.3
• /
• pp.639-647
• /
• 2011

In this paper, we introduce a design methodology of data-centroid Radial Basis Function neural networks with extended polynomial function. The two underlying design mechanisms of such networks involve K-means clustering method and Particle Swarm Optimization(PSO). The proposed algorithm is based on K-means clustering method for efficient processing of data and the optimization of model was carried out using PSO. In this paper, as the connection weight of RBF neural networks, we are able to use four types of polynomials such as simplified, linear, quadratic, and modified quadratic. Using K-means clustering, the center values of Gaussian function as activation function are selected. And the PSO-based RBF neural networks results in a structurally optimized structure and comes with a higher level of flexibility than the one encountered in the conventional RBF neural networks. The PSO-based design procedure being applied at each node of RBF neural networks leads to the selection of preferred parameters with specific local characteristics (such as the number of input variables, a specific set of input variables, and the distribution constant value in activation function) available within the RBF neural networks. To evaluate the performance of the proposed data-centroid RBF neural network with extended polynomial function, the model is experimented with using the nonlinear process data(2-Dimensional synthetic data and Mackey-Glass time series process data) and the Machine Learning dataset(NOx emission process data in gas turbine plant, Automobile Miles per Gallon(MPG) data, and Boston housing data). For the characteristic analysis of the given entire dataset with non-linearity as well as the efficient construction and evaluation of the dynamic network model, the partition of the given entire dataset distinguishes between two cases of Division I(training dataset and testing dataset) and Division II(training dataset, validation dataset, and testing dataset). A comparative analysis shows that the proposed RBF neural networks produces model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Elastomers and Composites
• /
• v.45 no.4
• /
• pp.263-271
• /
• 2010

Intermittent CSR testing was used to investigate the degradation of an FKM O-ring, also the prediction of its life-time. An intermittent CSR jig was designed taking into consideration the O-ring's environment under use. The testing allowed observation of the effects of friction, heat loss, and stress relaxation by the Mullins effect. Degradation of O-rings by thermal aging was observed between 60 and $160^{\circ}C$. In the high temperature of range ($100-160^{\circ}C$) O-rings showed linear degradation behavior and satisfied the Arrhenius relationship. The activation energy was about 60.2 kJ/mol. From Arrhenius plots, predicted life-times were 43.3 years and 69.9 years for 50% and 40% failure conditions, respectively. Based on TTS (time-temperature superposition) principle, degradation was observed at $60^{\circ}C$, and could save testing time. Between 60 and $100^{\circ}C$ the activation energy decreased to 48.3 kJ/mol. WLF(William-Landel-Ferry) plot confirmed that O-rings show non-linear degradation behavior under $80^{\circ}C$. The life-time of O-rings predicted by TTS principle was 19.1 years and 25.2 years for each failure condition. The life-time predicted by TTS principle is more conservative than that from the Arrhenius relationship.

• Journal of Sensor Science and Technology
• /
• v.25 no.6
• /
• pp.440-446
• /
• 2016

This study introduces the characteristics of current-voltage (I-V) and impedance variance for $ZnO-Zn_2BiVO_6-Co_3O_4$ (ZZCo), which is sintered at $900^{\circ}C$, according to temperature changes. ZZCo varistor demonstrates dramatic improvement of non-linear coefficient, ${\alpha}=66$, with lower leakage current and higher insulating resistivity than those of ZZ ($ZnO-Zn_2BiVO_6$) from the aspect of I-V curves. While both systems are thermally stable up to $125^{\circ}C$, ZZCo represents a higher grain boundary activation energy with 1.05 eV and 0.94 eV of J-E-T and from IS & MS, respectively, than that of ZZ with 0.73 eV and 0.82 eV of J-E-T and from IS & MS, respectively, in the region above $180^{\circ}C$. It could be attributed to the formation of $V^*_o$(0.41~0.47 eV) as dominant defect in two systems, as well as the defect-induced capacitance increase from 781 pF to 1 nF in accordance with increasing temperature. On the other hand, both the grain boundary capacitances of ZZ and ZZCo are shown to decrease to 357 pF and 349 pF, respectively, while the resistances systems decreased exponentially, in accordance with increasing temperature. So, this paper suggests that the application of newly formed liquid phases as sintering additives in both $Zn_2BiVO_6$ and the ZZCo-based varistors would be helpful in developing commercialized devices such as chips, disk-type ZnO varistors in the future.