Mathematical Modeling of Manufacturing Lines with Distribution by Process: A Markov Chain Approach

Today, there are a wide variety of ways to produce goods in a manufacturing company. Among the most common are mass or line production and process production, both of which are antagonists. In an online production system, materials move from station to station, receiving added value on a well-defined layout. In a production line by process, the materials randomly visit a set of machines strategically located in order to receive a treatment, almost always through metalwork machines, according to the final product of which they will be part. In this case, there is not a predefined layout, as the incoming materials are sectioned and each piece forms a continuous flow through different workstations to receive some process. This activity depends on the function of the product and its final destination as a component of a finished product. In this proposal, Markov chain theory is used to model a manufacturing system by process in order to obtain the expected values of the average production per machine, the total expected production in all the facilities, the leisure per machine and the total productive efficiency of the system, among other indicators. In this research, we assume the existence of historical information about the use of the equipment, its failures, the causes of failure and their repair times; in any factory, this information is available in the area of manufacturing engineering and plant engineering. From this information, statistical frequency indicators are constructed to estimate transition probabilities, from which the results presented here are derived. The proposal is complemented with a numerical example of a real case obtained from a refrigerator factory established in Mexico in order to illustrate the results derived from this research. The results obtained show their feasibility when successfully implemented in the company.

Epileptic Seizures Detection in EEG Signals Using Fusion Handcrafted and Deep Learning Features

Epilepsy is a brain disorder disease that affects people’s quality of life. Electroencephalography (EEG) signals are used to diagnose epileptic seizures. This paper provides a computer-aided diagnosis system (CADS) for the automatic diagnosis of epileptic seizures in EEG signals. The proposed method consists of three steps, including preprocessing, feature extraction, and classification. In order to perform the simulations, the Bonn and Freiburg datasets are used. Firstly, we used a band-pass filter with 0.5–40 Hz cut-off frequency for removal artifacts of the EEG datasets. Tunable-Q Wavelet Transform (TQWT) is used for EEG signal decomposition. In the second step, various linear and nonlinear features are extracted from TQWT sub-bands. In this step, various statistical, frequency, and nonlinear features are extracted from the sub-bands. The nonlinear features used are based on fractal dimensions (FDs) and entropy theories. In the classification step, different approaches based on conventional machine learning (ML) and deep learning (DL) are discussed. In this step, a CNN–RNN-based DL method with the number of layers proposed is applied. The extracted features have been fed to the input of the proposed CNN–RNN model, and satisfactory results have been reported. In the classification step, the K-fold cross-validation with k = 10 is employed to demonstrate the effectiveness of the proposed CNN–RNN classification procedure. The results revealed that the proposed CNN–RNN method for Bonn and Freiburg datasets achieved an accuracy of 99.71% and 99.13%, respectively.

Measuring System Calibration Factors by Unmixing the Excitation–Emission Spectra of One Dish of Cells

Accurate predetermination of the quantum yield ratio (QA/QD) and the extinction coefficient ratio (KA/KD) between acceptor and donor is a prerequisite for quantitative fluorescence resonance energy transfer (FRET) imaging. We here propose a method to measure KA/KD and QA/QD by measuring the excitation–emission spectra (ExEm-spectra) of one dish of cells expressing m (≥3) kinds of FRET constructs. The ExEm-spectra images are unmixed to obtain the weight maps of donor (WD), acceptor (WA), and acceptor sensitization (WS). For each cell, the frequency distribution plots of the WS/WD and WS/WA images are fitted by using a single-Gaussian function to obtain the peak values of WS/WD (SD) and WS/WA (SA). The statistical frequency-SD/SA plots from all cells are fitted by using a multi-Gaussian function to obtain the peak values of both SD and SA, and then the ranges of WS/WD (RSD) and WS/WA (RSA) for each FRET construct are predetermined. Based on the predetermined RSD and RSA values of FRET constructs, our method is capable of automatically classifying cells expressing different FRET constructs. Finally, the WS/WD–WA/WD plot from different kinds of cells is linearly fitted to obtain KA/KD and QA/QD values.

Analisis Penerapan Keselamatan Dan Kesehatan Kerja (K3) Pembangunan Gedung Kuliah Bersama Kampus C Unair Surabaya

Implementation of multi storey building construction projects is very prone of work accidents, so the application of Occupational Safety and Health must be strictly considered. This study purposes are to determine the most dominant occupational safety and health implementation measures and the level of implementation of occupational safety and health in the Shared Lecture Building Project Of Campus C Airlangga University Surabaya. This research uses quantitative descriptive method. The research populations are 150 workers, which include: security, workers, foremen, safety officer, project implementers and management staff. The sample was determined by the slovin technique with the results of 60 respondents. Data collection by distributing questionnaires to respondents. The results of data collection were tested for validity, realibility testing and statistical frequency analysts using IBM SPSS Statistics 25 software. In this study the most dominant results of the application of Occupational Safety and Health (K3) is checking the condition of PPE and the provision of PPE that is periodically complete with a value of 91.70%. The application level of Occupational Safety and Health (K3) in the Shared Lecture Building Project Of Campus C Airlangga University Surabaya has a percentage of 77.84%, so it can be classified in the VERY GOOD category. Pelaksanaan proyek konstruksi gedung bertingkat sangat rawan akan terjadinya kecelakaan kerja, sehingga penerapan Keselamatan dan Kesehatan Kerja harus benar-benar diperhatikan. Tujuan dilakukan penelitian ini untuk mengetahui tindakan penerapan Keselamatan dan Kesehatan Kerja yang paling dominan dan tingkat penerapan Keselamatan dan Kesehatan Kerja pada proyek pembangunan Gedung Kuliah Bersama Kampus C UNAIR Surabaya. Metode penelitian ini menggunakan deskriptif kuantitatif. Populasi penelitian berjumlah 150 tenaga kerja yang meliputi: satpam, pekerja, mandor, pelaksana K3, pelaksana proyek dan staff manajemen. Sampel ditentukan dengan teknik slovin dengan hasil 60 responden. Pengumpulan data dengan membagikan kuisoner kepada responden. Hasil pengumpulan data dilakukan uji validitas, uji reliabilitas dan analis frequensi statistic menggunakan software IBM SPSS Statistic 25. Pada penelitian ini di dapatkan hasil tindakan penerapan Keselamatan dan Kesehatan Kerja (K3) yang paling dominan adalah Pengecekan Kondisi APD dan Penyediaan APD yang lengkap secara berkala dengan nilai 91,70%. Tingkat penerapan Keselamatan dan Kesehatan Kerja (K3) Pada Proyek Pembangunan Gedung Kuliah Bersama Kampus C UNAIR memiliki prosentase 77,84 %, sehingga dapat di klasifikasikan dalam kategori SANGAT BAIK.

Increasing frequency stability of HF signals in the transmitting device of the localiser based on statistical estimation of phases

The article based on the method of statistical frequency stabilization deals with the issues of increasing frequency stability and synchronization of the forming HF signals in a transmitting device of a localizer using a multichannel variant of construction. It was demonstrated that the available digital unit of frequency and phase correction allows easy application of the proposed method. Two main features of the localizer operation affecting frequency stability and phase synchronization of HF signals are noted. The first factor is determined by deviation of the present -HF signal frequency (on the measurement interval) from the average (i) frequency value in n-channel. The second one is related to average frequency variation of each of the forming HF signals and its deviation within the (i) value from the nominal value during localizer operation. On the basis of HF signals description in channels of the transmitting device of a localizer ratios are obtained determining optimal values in terms of the method of least square method as (i) deviations of the present frequency values from the average value as well as variations of average frequency values during localizer operation. The article considers the most significant, from an applicatory point of view, case of assessment covering only deviations of the present HF signals frequency values from the average value on the measurement interval. It is shown that application of the method of statistical frequency stabilization allows the transmitting device of a localizer including N channels of HF signals formation to increase frequency stability and HF signals phase synchronization times. That enables to improve accuracy of forming integrated and difference directivity diagrams and setting heading in the runway direction as well. Apart from that, on the basis of the received values of frequency parameters estimation and relative instability of the forming HF signals a decision can be made about the condition of the controlled parameter by the criterion STANDARD-DETERIORATION-ACCIDENT.

Characterizing Corrosion Defects With Apparent High Growth Rates on Transmission Pipelines

Consecutive in-line inspections of transmission pipelines enable a comparison between the inspection results to characterize corrosion growth. Despite the high levels of in-line inspection tool accuracy and detection capabilities, corrosion defects with low calculated burst capacities may be detected on a subsequent inspection that were not reported in a previous inspection. These newly reported defects can pose a substantial challenge as the apparent growth rates between inspections of these defects can potentially drive unnecessary repair digs. This paper characterizes the contributing factors that can explain these phenomena, including: • Typical corrosion growth rates and their associated statistical frequency • The diminishing detection capability of inspection tools for smaller defects • The inspection tool minimum reporting threshold • The measurement accuracy of inspection tools. A statistical analysis was developed to quantify this interacting set of factors using Monte Carlo simulations that work retrospectively, covering a range of observed measured defect depths and then simulating the processes that could lead to newly reported defects being un-matched in a previous inspection. This analysis can be used to quantify the likelihood that a defect of a specific measured size would have been unreported in an earlier inspection due only to the performance characteristics of the inspection tool, and not as a result of defect growth that initiated since the time of the previous inspection. A set of case studies covering a range of pipeline inspection intervals ranging from 2 to 10 years are presented to demonstrate how this approach can be used to quantify appropriate growth rates that may be applied to these un-matched defects when assessing the remaining life or predicted probability of failure.

A Complete Characterization of Projectivity for Statistical Relational Models

A generative probabilistic model for relational data consists of a family of probability distributions for relational structures over domains of different sizes. In most existing statistical relational learning (SRL) frameworks, these models are not projective in the sense that the marginal of the distribution for size-n structures on induced substructures of size k<n is equal to the given distribution for size-k structures. Projectivity is very beneficial in that it directly enables lifted inference and statistically consistent learning from sub-sampled relational structures. In earlier work some simple fragments of SRL languages have been identified that represent projective models. However, no complete characterization of, and representation framework for projective models has been given. In this paper we fill this gap: exploiting representation theorems for infinite exchangeable arrays we introduce a class of directed graphical latent variable models that precisely correspond to the class of projective relational models. As a by-product we also obtain a characterization for when a given distribution over size-k structures is the statistical frequency distribution of size-k substructures in much larger size-n structures. These results shed new light onto the old open problem of how to apply Halpern et al.'s ``random worlds approach'' for probabilistic inference to general relational signatures.