Thesis for the degree of candidate of technical sciences in specialty 05.11.13 - devices and methods for monitoring and determining the composition of substances. - IvanoFrankivsk National Technical University of Oil and Gas, Ivano-Frankivsk, 2021.
The advantages and disadvantages of methods for diagnosing a gas pumping unit with an automatic control system are analyzed, the absence of a systematic approach to their development is established and the possibility of using methods for identifying objects, as well as modern information technologies for processing vibroacoustic processes when modernizing existing and developing new diagnostic methods for a gas pumping unit with an automatic management. The theoretical foundations of the method for diagnosing the technical state of the GTK-25-I automatic control system and by their acceleration characteristics have been developed. In this case, the technical state of the automatic control system of the gas pumping unit is determined by the change in the area formed by the transient characteristic and is a diagnostic feature that characterizes the change in the technical state of the gas pumping unit. To determine the technical state of GTK-25-i, it is proposed to use two-layer hierarchical neural networks of feedforward propagation, training according to the error backpropagation algorithm, and modeling the "defective" and "nominal" states based on the obtained measurement results and processing the characteristics of acoustic processes generated by the gas pumping unit in during operation.
Methodological, technical and software support has been developed for conducting experimental studies, taking into account the peculiarities and operating conditions of the GTK-25-i, the selection and installation of additional sensors of technological parameters is justified, an acoustic monitoring system of its state and a complex software package for monitoring the state of the GTK-25-I have been done.
The statistical characteristics of the acoustic processes of GTK-25-I certain diagnostic features that can be used as the basis for methods for monitoring its technical condition have been investigated. Based on the results of experimental studies, the following have been developed: a method for parametric diagnostics of the technical state of GTK -25-I, based on the determination of the highest values of the discriminant functions of the GTK-25-I under study and for each of its technical states for 16 technological parameters; method, the diagnostic feature of which is the value of the norm of the wavelet component of the fifth order detailing D5, obtained by processing the acoustic processes accompanying the operation of the GTK-25-I; an artificial neural network that was trained on experimental data, the results of which make it possible to clearly distinguish between two states "unsatisfactory" and "nominal". The procedure for determining the readiness indicator on the basis of statistical data on failures of the GTK-25-i automatic control system is considered and its value Ri = 0.93 is calculated, which confirms the effectiveness of the developed method for diagnosing the GTK-25-i automatic control system.
Key words: diagnostic methods, diagnostic feature, technical condition, acceleration characteristic, experimental data, gas pumping unit, automatic control system, artificial neural networks, wavelet transformations, discriminant analysis, acoustic process, diagnostics system.
