signal processing

Predictive maintenance is a widely applied maintenance program that requires extensive support of computer aided systems. The program uses specific procedures that are to be addressed when developing predictive maintenance software solutions. Despite the fact that software solutions for predictive maintenance were introduced almost at the same time as the program emerged, it still remains a very actual field for the application of computer aided systems.

The work is devoted to the problem of frequency-selective signal conversion in microelectronic sensor devices. It has been shown that the signal path of such devices, in particular, sensor nodes in the concept of the Internet of Things, must meet the requirements of embedded systems using a mixed analog-digital front end. The analysis of the signal transformation of photovoltaic sensors, in particular the problem of a significant parasitic influence of extraneous non-informative optical radiation and electromagnetic interference, has been carried out.

The possibility of creating inexpensive and widely available real-time signal processing devices to solve scientific and technical tasks based on a microcontroller, which is an alternative to the use of specialized digital signal processors, has been investigated. It is shown that the bit rate, clock speed, instruction speed are the basic parameters that are necessary for the implementation of algorithms for digital signal processing based on the microcontroller.

The accuracy analysis method for determining the signal phase of the induction encoder depending on the digit capacity of the input data obtained from the sensor via analog-to-digital converter presented in this article. The phase determining method based on a known method using the invers tangent and invers cotangent functions, which found the maximum and RMS errors depending on the digit capacity of the input data.

The design of mathematical models and corresponding functional block-diagrams of discrete-time neural networks for Internet information retrieval, parallel sorting, and rankorder filtering is proposed. The networks are based on the discrete-time dynamical K-winnerstake-all (KWTA) neural circuits which can identify the K largest from N input signals, where 1£ < K N is a positive integer. Implementation prospects of the networks in an up-to date digital hardware are outlined.