Analysis of filtering systems in the experimental research of a transport unit

Abstract

The increase in the speed of development and commissioning of new technology poses a challenge for scientists to conduct research more quickly. There is a need to change approaches to research methods and measuring devices. The main requirement is the speed of the research, the quality and relevance of the information. The use of highly sensitive sensors, on-board measuring devices and the corresponding software solves this problem. But the question arises of the synthesis of measuring sensors, the operation of which is based on the use of fundamentally different physical effects (induction, electromagnetic oscillations, radio waves). Each of these sensors has its own noise spectra; therefore, when conducting research and processing information, it is necessary to use filtering algorithms that can eliminate this drawback.

Recently, many scientists have been conducting experimental research using capacitive accelerometers. Their advantage is high sensitivity, ease of use and low price. But in the general case, readings of this type of accelerometer are subject to significant noise, which is usually caused by design and field conditions, which are characterized by stochastic factors associated with the operating environment: unwanted vibrations, high humidity and temperature, electromagnetic interference from other electromechanical and mechanical elements.

The analysis of the advantages and disadvantages of existing filters is carried out. Justification of the sequence of their application and settings required for field research in real time.

It has been determined that the use of a cascade of active filters is necessary when conducting research in the field, since such filters can be built into the software, which will significantly increase the speed and quality of research. With accurate information about the source of unwanted noise, the application of the principles of Fourier transform allows you to divide the entire signal into components and carry out further processing with those parts that provide the most relevant information in accordance with the research objectives.

Unify the software for various conditions of the experiment in real time, possibly by providing modular connection, or disconnecting individual filters from the main filtering cascade.