RU2049320C1 - Method of vibrodiagnostics of mechanism - Google Patents

Abstract

FIELD: vibromeasuring engineering. SUBSTANCE: spectrum of vibrations of mechanism is divided into N frequency fields and N groups of frequency components are filtered off; selected as m vibroparameter of mechanismm is total (integral) intensity of vibration spectrum components of m frequency field; after determination of respective approximating functiona, respective final difference of first order is determined in function of discrete time which is compared with preset threshold level and defect is determined in mechanism at preset moment of time; prediction for this defect is made; technical state of mechanismm is estimated according to values Δ_m(t_i) = σ $\genfrac{}{}{0ex}{}{\text{2}}{\text{m}}$ (t_i-σ $\genfrac{}{}{0ex}{}{\text{2}}{\text{m}}$ (t_i-1) for each present and future moments of time. EFFECT: wide informativity and extended functional capabilities through prediction of spread of defect. 1 dwg

Description

 The invention relates to a vibration measuring technique and can be used to assess the technical condition of mechanisms.

The closest in technical essence to the proposed one is the method of vibration diagnostics of the technical condition of the mechanism, which consists in the fact that at successive times t ₁ , t ₂ , t _k measure the current values of the vibration parameters of the mechanism, determine the informative components of the vibration spectrum, for the measured informative components of the spectrum build approximating functions determine the maximum permissible values of informative components and evaluate the technical condition of the mechanism.

 However, the method is characterized by insufficiently high information content and limited functionality in view of ignoring possible changes in the technical state of the mechanism in the interval of a series of consecutive measurements.

 The purpose of the invention is the expansion of information content and functionality by predicting the development over time of each individual identified defect.

≅ N для каждого текущего и будущих моментов времени, где σ² _m(t_i) суммарная интенсивность составляющих спектра вибраций из m-й частотной области в дискретном времени t_i, i 1,2,
При работе сложных механизмов возбуждается сложный спектр вибраций в некоторой фиксированной полосе частот. Возникающие в процессе работы механизма дефекты проявляются в искажениях соответствующих областей спектра вибраций за счет увеличения интенсивности или средней мощности их спектральных составляющих. Каждому конкретному дефекту можно поставить в соответствие вибропараметр механизма, определяемый по спектральным составляющим из соответствующих М частотных областей. Анализируя изменения вибропараметра во времени на интервале наблюдения [t₁, t_k] можно по одной из известных методик построить его аппроксимирующую функцию и прогнозировать дальнейшее (вне интервала наблюдения) развитие дефекта в работе механизма. Решение такой задачи достигается в предлагаемом способе.The goal is achieved by the fact that in the method of vibration diagnostics of the technical state of the mechanism, which consists in the fact that at successive times the current values of the vibration parameter of the mechanism are measured, the corresponding approximating function is determined from the changed sequences of the values of the vibration parameter, while the vibration spectrum of the mechanism is divided into N frequency regions, measurement of the vibration parameter and the determination of the approximating function is performed in each of the N frequency domains, as the mechanism vibration parameter is selected ummarnuyu intensity components of the vibration spectrum after determining respective approximating functions define the corresponding finite difference of first order Δ _m (t _i) at discrete time function, compare it with a predetermined threshold level and from the comparison determine the presence of a defect and its development in the mechanism, and technical the state of the mechanism is estimated from the following relationships:
Δ _m (t _i ) σ $\genfrac{}{}{0ex}{}{\text{2}}{\text{m}}$ (t _i -σ $\genfrac{}{}{0ex}{}{\text{2}}{\text{m}}$ (t _i-1 ), m

≅ N for each current and future time points, where σ ² _m (t _i ) is the total intensity of the components of the vibration spectrum from the m-th frequency domain in the discrete time t _i , i 1,2,
When complex mechanisms are operating, a complex spectrum of vibrations is excited in a certain fixed frequency band. Defects arising during the operation of the mechanism manifest themselves in distortions of the corresponding regions of the vibration spectrum due to an increase in the intensity or average power of their spectral components. Each specific defect can be associated with a vibration parameter of the mechanism, determined by the spectral components from the corresponding M frequency regions. By analyzing the changes in the vibration parameter over time on the observation interval [t ₁ , t _k ], one of the known methods can construct its approximating function and predict further (outside the observation interval) development of the defect in the mechanism. The solution to this problem is achieved in the proposed method.

 The method is as follows.

 The vibration spectrum of the mechanism is divided into N frequency domains and the corresponding N groups of frequency components are distinguished.

(t_k+n-t_k), n 1,2
Последовательность отсчетов аппроксимирующей функции в моменты i > k выполняет роль соответствующего прогнозируемого значения для вибропараметра механизма σ_m ². После этого определяют соответствующую конечную разность первого порядка
Δ_m (t_i)σ² _m (t_i) σ² _m (t_i-1),
i= k, k+1. как численную характеристику скорости увеличения вибропараметра σ² _m, которую сравнивают по величине с соответствующим пороговым уровнем α_m> 0. Полностью исправному механизму соответствует случай близких к нулю величин Δ_m(t_i) для всех m

Как результат, выполняется система неравенств
Δ_m(t_i) < α_m ∀t_i∈[t_k;t_k+n]
При возникновении в текущий момент времени t_k некоторого дефекта в работе механизма возрастает интегральная интенсивность σ_m ²спектральных составляющих из соответствующей m-й области спектра вибраций и одновременно возрастает соответствующая разность первого порядка Δ_m(t_i). При условии Δ_m (t_i) ≥ α_m возникший дефект автоматически регистрируют для дальнейшей оценки его развития по соответствующим будущим моментам времени t_i > t_k значениям аппроксимирующей функции σ² _m(t_i). По моменту t_k+n, когда достигается равенство Δ_m (t_k+n) α_m ^*, где α_m ^* (2-3) α_m предельно допустимое значение скорости увеличения вибропараметра σ _m ² определяют прогнозируемый остаточный ресурс работы механизма.The total or integral intensity σ ² _{m of the} spectral components of the vibrations from the m-th frequency domain, where m 1,2, M ≅ N, is selected as the vibration parameter of the mechanism. Measuring its values from the vibration of the mechanism at successive times t ₁ , t ₂ , t _k , determine the corresponding time series σ _{m, 1} ² , σ _{m, 2} ² , σ _{m, k} ² , which is used to construct the approximating function σ _m ² (t _i ) in the discrete time t _i , i 1,2, and the condition σ _m ² (t _i) = σ _{m, i} ² for i≅ k. An example is a linear approximation of the form
σ $\genfrac{}{}{0ex}{}{\text{2}}{\text{m}}$ (t _{k + n} ) σ $\genfrac{}{}{0ex}{}{\text{2}}{\text{m}}$ _{, k} +

(t _{k + n} -t _k ), n 1,2
The sequence of samples of the approximating function at moments i> k plays the role of the corresponding predicted value for the vibration parameter of the mechanism σ _m ² . After that, determine the corresponding finite difference of the first order
Δ _m (t _i ) σ ² _m (t _i ) σ ² _m (t _i-1 ),
i = k, k + 1. as a numerical characteristic of the rate of increase of the vibration parameter σ ² _m , which is compared in magnitude with the corresponding threshold level α _m > 0. A fully functional mechanism corresponds to the case of values close to zero Δ _m (t _i ) for all m

As a result, the system of inequalities
Δ _m (t _i ) <α _m ∀t _i ∈ [t _k ; t _{k + n} ]
When a defect occurs at the current time t _k in the mechanism, the integrated intensity σ _m ^{2 of the} spectral components from the corresponding m-th region of the vibration spectrum increases and the corresponding first-order difference Δ _m (t _i ) also increases. Under the condition Δ _m (t _i ) ≥ α _{m, a} defect that has arisen is automatically recorded for further assessment of its development according to the corresponding future instants of time t _i > t _{k to the} values of the approximating function σ ² _m (t _i ). At the moment t _{k + n} , when the equality Δ _m (t _{k + n} ) α _m ^{* is} achieved, where α _m ^* (2-3) α _{m the} maximum permissible value of the rate of increase of the vibration parameter σ _m ² determines the predicted residual life of the mechanism.

 The results of such forecasting make it possible for each controlled mechanism to effectively plan preventive and repair work at a time convenient from the point of view of production costs until the mechanism fails.

 The method can be implemented, for example, using a device, a diagram of which is shown in the drawing.

 The device comprises a series-connected vibration transducer 1, a broadband amplifier 2, a block of 3 bandpass filters, a block of 4 quadratic detectors, a block of 5 integrators, a multi-channel tape recorder 6, an analog-to-digital converter 7, an arithmetic block 8, a threshold block 9, a logical block 10, and a display block 11 as well as a switching unit 12, the information inputs of which are connected to the corresponding outputs of the arithmetic unit 8 and the inputs of the threshold unit 9, control inputs with the corresponding outputs of the logical unit 10 and first the input inputs of the display unit 11, and the outputs with the corresponding second inputs of the display unit 11.

 The device operates as follows.

в арифметическом блоке 8 определяется соответствующая аппроксимирующая функция σ_m ²(t_i) в дискретном времени t_i, i 1,2. и после этого формируется соответствующая конечная разность первого порядка Δ_m (t_i). Ее текущий отсчет, равный Δ_m(t_k), подается на соответствующий вход порогового блока 9, где он сравнивается по величине с заданным пороговым уровнем α_m > 0. При выполнении условия Δ_m(t_k) ≥ α_m на соответствующем m-м выходе порогового блока 9 формируется сигнал логической единицы. Совокупность логических единиц и нулей одновременно на всех М выходах порогового блока 9 представляет собой М-разрядный двоичный код соответствующего дефекта в работе механизма. Ее дешифрация и определение конкретного типа дефекта осуществляется в логическом блоке 10, логика работы которого определяется в процессе предварительной настройки устройства на группу механизмов с заранее известными дефектами. Сигнал с каждого выхода логического блока 10 как признак соответствующего дефекта в работе механизма регистрируется в блоке 11 отображения диагностической информации и одновременно служит "ключом" для передачи на отображение через блок 12 коммутации последовательностей значений, соответствующих аппроксимирующей функции σ_m ²(t_i) и ее разности первого порядка Δ_m(t_i), как численных характеристик текущего и прогнозируемого технического состояния механизма.The signal of the vibration transducer 1 connected to a controlled mechanism (not shown) is amplified in a broadband amplifier 2 and fed simultaneously to the N inputs of the bandpass filter unit 3, where separate groups of its spectral components are allocated that carry information about the technical condition of the mechanism with respect to each specific defect in his work. The signals from the outputs of N bandpass filters are detected by amplitude in block 4 of quadratic detectors and then separately averaged in block 5 of integrators to obtain current estimates of the vibration parameter of the mechanism σ _m ² } Signals of estimates of the vibration parameter from the interval [t ₁ ; t _k ] for each frequency domain are stored in a multi-channel tape recorder 6 and after that they are successively fed to the input of the amplitude-to-digital converter 7 in an accelerated time scale. Based on the arrays of digital codes obtained at the output of the latter, σ _{m, 1} ² , σ _{m, 2} ² , σ _{m, k} ^-2 for the vibration parameter σ _m ² , m

in the arithmetic unit 8, the corresponding approximating function σ _m ² (t _i ) is determined in the discrete time t _i , i 1,2. and then a corresponding finite first-order difference Δ _m (t _i ) is formed. Its current count, equal to Δ _m (t _k ), is fed to the corresponding input of the threshold block 9, where it is compared in magnitude with a given threshold level α _m > 0. When the condition Δ _m (t _k ) ≥ α _m is met on the corresponding m- m output of the threshold block 9 is formed by a signal of a logical unit. The set of logical units and zeros at the same time on all M outputs of the threshold block 9 is an M-bit binary code of the corresponding defect in the mechanism. Its decryption and determination of a specific type of defect is carried out in a logical block 10, the logic of which is determined in the process of pre-setting the device to a group of mechanisms with previously known defects. The signal from each output of the logical unit 10 as a sign of a corresponding defect in the mechanism is registered in the diagnostic information display unit 11 and at the same time serves as a “key” for transmitting to the display through the switching unit 12 sequences of values corresponding to the approximating function σ _m ² (t _i ) and its the first order difference Δ _m (t _i ) as the numerical characteristics of the current and predicted technical condition of the mechanism.

 The use of information on the rate of development of each individual defect in time expands the information content and functionality of the vibration diagnostics of mechanisms, which together contributes to a more complete use of their working resource and to minimize maintenance costs.

 The analysis indicates the novelty and inventive step of the solution, and the technical study performed confirms the possibility of its industrial applicability.

Claims (1)

METHOD FOR VIBRODIAGNOSTICS OF THE TECHNICAL STATE OF THE MECHANISM, which consists in the fact that at successive moments of time the current values of the vibration parameter are measured, the corresponding approximating function is determined from the measured sequences of the vibration parameter and their state of the mechanism is evaluated, characterized in that the vibration spectrum of the mechanism is divided into N frequency regions vibration parameter and the determination of approximating functions are carried out in each of the N frequency domains, as a vibration parameter of the mechanism you irayut total intensity components of the vibration spectrum after determining respective approximating functions define finite difference of first order Δ _m (t _i) at discrete time function, compared its value with a threshold level and the comparison results determine the presence of a defect and its development in the mechanism, and technical the state of the mechanism is estimated from the ratio

for every current and future points in time,
where σ $\genfrac{}{}{0ex}{}{\text{2}}{\text{m}}$ (t _i ) -total intensity of the components of the spectrum of vibrations from the m-th frequency domain in a discrete time t _i ,
i 1, 2,

Images