Achieving productive reliability through applying statistical distribution functions

Purpose The purpose of this paper is to find the proper statistical distribution function, which can cover the failure time of a single machine or a group of machines. To this end, an innovative program is written in an Excel software, capable of assessing at least six statistical distribution functions. This research study intends to show the advantages of applying statistical distribution functions in an integrated model format to create or increase productive reliability machines. Productive reliability is a simultaneous combination of efficiency and effectiveness in reliability. Design/methodology/approach The method of theoretical research methodology comprises data collection tools, reference books and articles in addition to exploiting written reports of the Iranian Center for Defence’s Standards. The practical research method includes deploying and assessing the proposed model for a selected machine (in this case a computerized numerical control machine). Findings A comprehensive program in an Excel software having the capability of assessing at least six statistical distribution functions was developed to find the most efficient option for covering the failure times of each machine in the shortest time with the highest precision. This is regarded as the most important achievement of the present study. Furthermore, the advantages of applying the developed model are discussed and a large group of which have direct influences on the productivity of equipment reliability. Originality/value The originality of the research was ascertained by managers and experts working in maintenance issues at the different levels of the Defense Industries Organization.

Performance Probability Distribution Function for Modelling Solar Radiation in South Southern Nigeria: A Case Study of Yenagoa

This study has attempted to assess the performance of the most suitable statistical distribution function for modelling solar radiation over Yenagoa, Bayelsa State in Nigeria. The probability distribution functions are tested based on eleven years (2007-2017) solar radiation data obtained from National Aeronautics and Space Administration (NASA). Six probability distribution functions are tested to ascertain the most appropriate one based on four different statistical tools and fitting accuracy. The associated parameters of the most appropriate fitted probability distribution function are calculated and the trends in the characteristic of the solar radiation are deduced. The result shows that logistic distribution presents the most suitable probability distribution function for modelling solar radiation over the selected environment with RMSE of 1.500 KWh/m2/day, MAE of 1.260 KWh/m2/day, MAPE of 22.000% and R2 of 0.880.When compared with the other five distribution functions, the same trend could be seen although with different values of RMSE, MAE, MAPE and R2. The estimated distribution location and scale parameters of the model vary with month and season. The overall result will be useful for predicting future solar radiation over the studied environment. It will also be a good reference point for the design of large solar power projects in Yenagoa in particular and south southern Nigerian environments at large.

Microstructure Design for Fast Lifetime Measurements of Magnetic Tunneling Junctions

The estimation of the reliability of magnetic field sensors against failure is a critical point concerning their application for industrial purposes. Due to the physical stochastic nature of the failure events, this can only be done by means of a statistical approach which is extremely time consuming and prevents a continuous observation of the production. Here, we present a novel microstructure design for a parallel measurement of the lifetime characteristics of a sensor population. By making use of two alternative designs and the Weibull statistical distribution function, we are able to measure the lifetime characteristics of a CoFeB/MgO/CoFeB tunneling junction population. The main parameters governing the time evolution of the failure rate are estimated and discussed and the suitability of the microstructure for highly reliable sensor application is proven.

Effect of Neighboring Phase Properties on Measured Indentation Data

The paper focuses on the variation of mechanical properties near the interface of two materials. The aim is to determine the properties of the individual phases and the size of the region influenced by the adjacent material. Nanoindentation measurements were performed in the vicinity of a sharp interface between tungsten and steel, which was a plane located parallel to the loading force direction. The probability of indentation in the affected area as a function of depth of penetration and distance from the interface was fitted by a proper statistical distribution function. The intrinsic properties of the individual phases can subsequently be extracted from the experimental indentation data.

New estimation method of degree of chaos in biosystems

. Potential application of traditional and stochastic methods in assessment of parameters of complex biosystem – complexity is presented. Their amplitude-frequency characteristics, autocorrelation functions A(t), Lyapunov exponents, statistical distribution function f(x) constantly change. In spite of such chaotic dynamics in recorded parameters of tremor, tapping, cardiograms, myograms and other parameters of homeostasis, an order in the dynamics of these processes can be observed. The order is revealed in a change of numbers of sample overlaps that are obtained as a result of some processes. Owing to all the stochastic characteristics constantly change, calculation method for statistical distribution functions with repetitions of identical experimental sets is proposed. In this case, number of sample overlaps (their belonging to the same general population) will numerically present the transition mechanism to order out of chaos or the variation of chaos degree in movement formation and electrobiological muscular activity. The current work shows some typical examples of different physiological human states presented in a form of matrices of paired comparison.

A Statistical Local Binary Fitting Model for Blood Vessel Segmentation

Network structure such as blood vessels in medical images are important features for computer-aided diagnosis and follow-up of many diseases. In this study, a new model-based segmentation method is proposed to detect blood vessels in medical images. The Local Binary Fitting (LBF) model with statistical distribution function is used for this purpose. The brain tissues and cerebral vessels in the image are modeled by Gaussian distribution and uniform distribution respectively. The region distribution combined with the LBF model is used in curve evolution. And the level set method is developed to implement the curve evolution to assure high efficiency of the cerebrovascular segmentation. Comparisons with the LBF method show that our model can achieve better results.