Search

The paper is devoted to the problem of design a contact-less radar MIMO system for the non-destructive testing (NDT) of concrete and reinforced concrete objects using the linearly Frequency-Modulated Continuum Waves signals (FMCW). This paper is the first one in the investigation of the described problem. In the consideration, the greatest interest is in the evaluation of the reinforcement parameters (reinforcing bar diameters and depth of reinforcement frames), thickness estimation, and visualization of the controlled objects in the lack of two-way access (for instance, foundations, slabs, and e.t.c.). The contact-less measurements allow one to speed-up and simplify the testing routine. In the designed device the described features are attained through the using of uniform rectangular array (URA) multi-input-multi-output (MIMO) systems with the FMCW signals and algorithms of signal processing and deep-learning based image processing under investigation. By this we are expect to achieve a high energy of emitting (high signal-to-noise ratio i.e. high depth) and a high space resolution in three dimensions. This paper is devoted to the first stage of the research which is the electrodynamics simulation of the FMCW signals in the investigated system. The results then will be used to signal processing algorithms investigation and gathering the dataset for the deep-neural network training (the second-stage processing and high-resolution visualization and reinforcement parameters investigation). [ABSTRACT FROM AUTHOR]

The paper deals with the numerical simulation of features of most popular up today adaptive signal decomposition techniques, which applied to the signals of the ultrasonic gas flowmeter in pipes with intrusive piezo- transducers with vertical wave penetration and frequency modulated continuum waves. The most attention was paid to the case of small pipe diameters in comparison with the piezo-transducer size (when it has to be considered). Among the investigated methods the most accurate results (with respect to the bias error) was obtained for singular signal analysis (SSA). However, in the trade-off of accuracy to computational complexity, the principal component analysis (PCA) method was recommended to the farther researches. [ABSTRACT FROM AUTHOR]

The article describes a method of processing eddy current sensor signals. We have described a method for measuring the angle between the sensor and the tangent to the shaft surface using only one sensor. Particular attention is paid to the effect of noise on measurement errors. Proposed method allows determining angle between sensor axis and tangent to shaft surface with accuracy better than 0.01 degrees when using only one sensor. [ABSTRACT FROM AUTHOR]

The paper deals with the problem of the using phase-based methods for frequency estimation in the frequency modulated continuum waves (FMCW) short-range radars in the case of high-resolution frequency estimation requirements. The equation is considered for frequency estimation by the least-square method of approximating the phase-to-time relation of the beat signal correlation function in the case of several significant components in the signal (such as valuable and interference signal-like tones). Solution is proposed for the equations by using the parametric (or subspace) decomposition of beat signals (such as eigenvector decomposition, EV). The numerical investigation shows that the bias of the frequency estimations by the proposed solution of the mentioned equation above has statistical properties similar to the method of estimation of signal parameters via rotational invariance techniques (ESPRIT). However, the proposed method does not require the double decompositions, and frequency of each eigenvector can be estimated separately. It is also shown that in the case of the unknown number of signal components the proposed method has the higher statistical properties than for such popular technique as the multiple signal classification method (MUSIC). [ABSTRACT FROM AUTHOR]

The paper deals with the problem of accuracy increasing of small distance changes measurements in radar level gauges with frequency modulated continuous wave signals. In the considered task advantages of estimation of time delay differences as indirect parameter of beat signals full phase are shown. This approach is shown to be more informative in comparison with traditionally applied estimators of time delay difference, which based on a frequency and initial phase estimation as separate parameters. For the model of beat signal of level FMCW radar in the typical interference conditions it is proposed to use the phase-based estimator which was investigated by us in the previous works. The obtained results shown that in the considered conditions the proposed approach allows one to reduce the variance and bias of level increments estimates by several times in comparison with traditionally used estimators. [ABSTRACT FROM AUTHOR]

The paper presents the method for estimation wear value probability by the numerical simulation. The considered task is actual in many technical regions. The designed method is based on an analysis of the probability density of wear distribution as a function of mileage, which considered as stochastic variable. Using this proposition allow one to accurately predict the wear value and its dependence of using lubrication. The proposed method was applied to Sverdlovsk railroad data where the results shown that using of KLS lubricant allow to decrease the wear on 30-50% depending on the mileage value. [ABSTRACT FROM AUTHOR]