Wyniki 1-3 spośród 3 dla zapytania: authorDesc:"Paweł RYMARCZYK"

Detection of seepages in flood embankments using the ElasticNET method DOI:10.15199/48.2019.01.40

Czytaj za darmo! »

Electric tomography is based on the transformation of data taken from the surface of the tested object into the image of its cross-section. There are many methods to optimize the obtained image by solving the appropriate objective function [1-5,13,15,16,20-25,32]. The algorithm based on the ElasticNET presented in this article is a new proposal in tomography. Fig. 1. Model of measuremnt system. The way of working of electrical impedance tomography (EIT) consists in introducing electrical voltage to the tested object by means of a set of electrodes located on the surface of the object. Next, the measured values of electrical potentials between individual electrode pairs are collected. Conductance of individual sections of the crosssection of the tested object is reconstructed on the basis of known values of voltages and measured values of potentials. Reconstruction of the image obtained by electrical tomography requires sophisticated modeling. This method of imaging consists in the fact that the conductivity distribution of the tested object is estimated on the basis of measurements of electrical voltages and electrode potentials on the surface of their contact with the tested object. In order to obtain quantitative data on changes in the conductivity inside an object, it is more effective to apply a non-linear model in differential imaging [1,6-12,14,17- 199,26-31]. In Fig. 1 shows the model of the measurement system. ElasticNET Let’s consider the problem of recognizing linear dependencies (1) Y  X   where Y Rn , X Rnk1 are the observation matrices of a output variable and predictive variables respectively,   Rk1means a matrix of structural parameters, while  Rn vector of independent random variables. The wellknown method of least squares consists in estimating unknown parameters &[...]

Machine learning in image reconstruction by multi-sensor electrodes DOI:10.15199/48.2019.12.42

Czytaj za darmo! »

In electrical impedance tomography (EIT), voltage or current is injected into an object using a set of electrodes attached to its surface [1-6]. Voltage values are measured on the remaining electrodes. The conductivity of the object is reconstructed based on the measurements taken and the appropriate algorithm to solve the inverse problem [7-19]. Measurement system The measuring system was based on an electric tomography device, which assumes the use of two measuring methods and allows measurements to be made to 32 channels. Figure 1 shows laboratory measurement models with one and two multi-sensor electrodes. The device measures the tested object based on measurements of the potential distribution on the surface. The system collects measured data from the electrodes. The solution assumes the use of two measuring methods and allows measurements to be made to 32 channels. The device consists of several separate modules: power generator, measuring block, multiplexer and controller. You can connect more than one multiplexer to the device. Data acquisition systems require equipment for voltage measurement, filtering, demodulation and conversion to digital units and signal processing in order to send data to the computing unit. Depending on the method of use, the system can be adapted to measurements in large areas or small sections. In both cases, all infrastructure is centrally managed to collect data and record measurements. The system allows you to collect measurements, manage data and monitor devices. An electric tomograph constructed for this type of measurement is shown in Fig. 2. Methods There are many numerical methods used in optimization problems [20-35]. This article uses machine learning methods [36-39] that use information to learn directly from data without first determining a mathematical model. The machine learning algorithms used utilize the availability of large amounts of data and computing resources. The [...]

Application of multi-source data for process analysis in electrical tomography DOI:10.15199/48.2019.12.43

Czytaj za darmo! »

Industrial tomography solves the inverse problem that the optimization, identification or synthesis process. Such problems are difficult to analyze [1-5]. They do not have clear solutions. Knowledge of the process can make image reconstruction more resistant to incomplete information [6- 9]. There are many numerical methods [10-26].Tomography enables the analysis of processes taking place in the facility without interfering with them [27-30]. It enables better understanding and monitoring of industrial processes and facilitates real-time process control. Figure 1 shows the structure of the system using an electric tomograph with image reconstruction. Fig. 1. The idea of system structure with a hybrid tomography scanner with flow measurement, image processing in the cloud computing. Methods Electrical tomography is an imaging technique that uses different electrical properties of materials. In this method, the current source is connected to the tested object, and then the voltage distribution on its edge is measured (Fig, 2). The collected information is processed and reconstructed by a specific algorithm. The basic theory can be obtained from Maxwell's equations. A complex 'admittivity' can define as follows: (1) 􀟛 􀵌 􀟪 􀵅 i􀟱􀟝 where ε is the permittivity, σ is the electrical conductivity, ω is the angular frequency. Potential distribution in a heterogeneous, isotropic area: (2) 􀗏[...]

 Strona 1