Wyniki wyszukiwania - BazTech

1

Residual life estimation of fabricated humidity sensors using different artificial intelligence techniques

Bhargava C., Aggarwal J., Sharma P. K.

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2019

|

Vol. 67, nr 1

147--154

EN

Background: a humidity sensor is used to sense and measure the relative humidity of air. A new composite system has been fabricated using environmental pollutants such as carbon black and low-cost zinc oxide, and it acts as a humidity sensor. Residual life of the sensor is calculated and an expert system is modelled. For properties and nature confirmation, characterization is performed, and a sensing material is fabricated. Methodology: characterization is performed on the fabricated material. Complex impedance spectroscopy (CIS), Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD) and scanning electron microscopy (SEM) are all used to confirm the surface roughness, its composite nature as well as the morphology of the composite. The residual lifetime of the fabricated humidity sensor is calculated by means of accelerated life testing. An intelligent model is designed using artificial intelligence techniques, including the artificial neural network (ANN), fuzzy inference system (FIS) and adaptive neuro-fuzzy inference system (ANFIS). Results: maximum conductivity obtained is 6.4£10−3 S/cm when zinc oxide is doped with 80% of carbon black. Conclusion: the solid composite obtained possesses good humidity-sensing capability in the range of 30–95%. ANFIS exhibits the maximum prediction accuracy, with an error rate of just 1.1%.

2

Study of the characteristics about insulation damage based on the accelerated life tests

Yang Y. H., Wang K.-S.

Eksploatacja i Niezawodność

|

2013

|

Vol. 15, no. 4

325--331

EN

In this study the Maximum Likelihood Estimator is taken to identify the characteristics of insulation failure about the class-H motors by considering the accelerated life testing data under censored situation from Nelson. Based on the Weibull survival modeling the failure is expressed as the series connection of three modes, namely the turn, phase, and ground, respectively, the so-called competing failure. The main concern in present investigation is about the variation of shape parameters, β with the temperature. The Gompertz-type relation of βi(T) is suggested with the reference temperature, Tri for the i-th failure mode. It is found that the Tri ’s not only distinguish the characteristics of cumulative damage process about the insulation, but also involve the estimation of mean-time-to-failure (MTTF). Physically Tri denotes the turning point of varied βi as the i-th failure mode becomes moderate in a sense of less capability about the accumulation of insulation damage at higher temperature where corresponds the thermal degradation process. The numerical results indicate that the insulation technique used is acceptable as the operation temperature kept in the use condition 363K. According to the predicted lifetime as the temperature rises up to 440K, which still within the allowed range in application, the turn structure needs to be rearranged primarily, then the phase next. The ground mode has only influence on the failure at much higher temperature.

3

On a general standby system and its optimization

Cha J. H., Yun W. Y.

Journal of Polish Safety and Reliability Association

|

2009

|

Vol. 1

95--100

EN

Redundancy or standby is a technique that has been widely applied to improving system reliability and availability in the stage of system design. In this paper, we consider a standby system with two units in which the first unit (unit 1) starts its operation under active state and the other unit (unit 2) is under cold standby state at the starting point. After a specified time s (switching time), the state of unit 2 is changed to warm standby state and, as soon as the operating unit 1 fails, the state of unit 2 is changed to active state. If unit 1 fails before times, the system fails. Units can fail at both active and warm standby states. A general method for modeling the standby system is adopted and, based on it, system performance measures (system reliability and mean life) are derived. We consider the problem of determining optimal switching time which maximizes the expected system life. Some numerical examples are studied.

4

On statistical modelling in accelerated life testing

Gerville-Réache L., Nikulin M.

Eksploatacja i Niezawodność

|

2006

|

nr 2

48-51

EN

The aim of this paper is to present some models used in accelerated life testing. The AFT model, the Sedyakin model, the Power Generalized Weibull model and the CHSS model are discussed. Many recent references are given in order to help readers in there choices.