Instrument fault diagnosis and health prediction method under complex working conditions

Through the diagnosis methods of noise power spectral density and dynamic characteristics, the problems of fault diagnosis and high cost of traditional instruments are solved, real-time fault detection and health assessment are realized, the diagnostic cost and false alarm rate of instruments are reduced, and diagnostic accuracy is improved. .

CN118913351APending Publication Date: 2024-11-08SHENYANG INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410946186.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-15
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

Traditional instruments have high costs, regular calibration requirements, high false alarm rates and false alarm rates in industrial automation and measurement technologies, and fault diagnosis is difficult to accurately carry out, especially due to the complex hardware design and lack of knowledge of failure mechanisms, which lead to system costs. High and low confidence in diagnostic results.

Method used

The noise power spectral density diagnosis method, dynamic feature extraction and fault evaluation method are used to detect fault types through data processing and signal processing, and fault detection is carried out in combination with the changes in dynamic feature response time, and a typical diagnostic platform is built for on-site verification to realize real-time fault diagnosis and health assessment.

Benefits of technology

It reduces the cost of instrument fault diagnosis, improves the accuracy and real-time diagnosis, solves the high cost and missed rate problems caused by redundancy of traditional instrument hardware, and realizes real-time fault diagnosis and health assessment in industrial environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118913351A_ABST
    Figure CN118913351A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of instrument detection, discloses an instrument fault diagnosis and health prediction method under complex working conditions, and relates to a noise power spectral density diagnosis method, an instrument dynamic feature extraction method, an instrument fault evaluation method and a data-driven instrument health evaluation method. The problem that typical faults of the instrument can only be diagnosed by external measures is solved, the diagnosis problem under the condition that the fault mechanism of the instrument is unknown is solved, and real-time state evaluation, field simulation of an industrial typical environment and instrument deployment are realized; real-time measurement of data of the industrial sensor network is realized, fault diagnosis and health assessment are carried out in real time, and the following problems of high cost, need of regular calibration and high missing report rate and false alarm rate caused by hardware redundancy in a traditional instrument are solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of instrument detection, and in particular relates to an instrument fault diagnosis and health prediction method under complex working conditions. Background Art

[0002] Instruments, as an indispensable part of industrial automation and measurement technology, refer to devices used to detect, display, record or control one or more physical or chemical quantities, which may include temperature, pressure, flow, speed, electricity, etc. Instruments play an important role in many fields such as modern industrial production, scientific research, and daily life.

[0003] Instruments are needed to measure and record data in many fields such as industrial automation and measurement technology. The problems with the above technologies are: traditional instruments have the following problems: high cost due to hardware redundancy, regular calibration required, high missed alarm rate and false alarm rate. For example, multiple instruments measure the same parameter and diagnose the instruments by cross-comparison, but the system is large in scale, complex in application, high in system cost and construction cost. Another example is: multi-channel and multi-core design methods are used inside the instrument to diagnose internal data and arbitrate data output; real-time performance is good and diagnostic capability is strong, but hardware design is complex, or: instrument fault diagnosis is performed based on the mathematical model established by the instrument; but the knowledge of failure mechanism is missing, accurate system modeling is difficult, monitoring parameters lack physical meaning, and the confidence level of the results is low. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a method for instrument fault diagnosis and health prediction under complex working conditions, which can overcome the above problems or at least partially solve the above problems.

[0005] The present invention is implemented as follows: a method for diagnosing instrument faults and predicting health in complex working conditions, the method comprising the following steps: S1: a method for diagnosing noise power spectrum density:

[0006] The instrument will generate process noise during use, and the output signal will generate output noise. Both process noise and output noise are processed through data processing algorithms to detect the fault type.

[0007] The high-frequency component of the instrument, that is, process noise, can be used to detect sudden faults through signal processing;

[0008] The low-frequency component of the instrument, that is, the output noise, can be used to detect slow-changing faults through signal processing;

[0009] S2: Methods for extracting instrument dynamic characteristics and evaluating instrument failures;

[0010] When a fault occurs, it is detected, and if the dynamic characteristic response time changes, it is a fault;

[0011] The dynamic characteristic value is the data obtained after filtering. The characteristic value is extracted and compared with the reference value to draw a fault conclusion;

[0012] S3: Collect the data used by the instrument, and then evaluate the health prediction of the instrument through data and feature data identification and extraction;

[0013] S4: Build a typical instrument diagnostic platform and conduct on-site verification on the platform.

[0014] Compared with the prior art, the present invention has the following beneficial effects:

[0015] The present invention solves the problem that typical instrument faults can only be diagnosed by external measures, solves the diagnosis problem when the instrument fault mechanism is unclear, realizes real-time status evaluation and realizes on-site simulation of typical industrial environments and instrument deployment through a noise power spectral density diagnosis method, an instrument dynamic feature extraction method, an instrument fault evaluation method and a data-driven instrument health assessment method; realizes real-time measurement of industrial sensor network data, and performs real-time fault diagnosis and health assessment, solving the following problems existing in traditional instruments: high cost due to hardware redundancy, the need for regular calibration, and high missed alarm rate and false alarm rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the principle of noise extraction provided by an embodiment of the present invention;

[0017] Figure 2 The embodiment of the present invention provides a fault diagnosis diagram of noise power spectrum density;

[0018] Figure 3 is a schematic diagram of a time response algorithm for providing dynamic features according to an embodiment of the present invention;

[0019] Figure 4 It is a schematic diagram of time response fault judgment of dynamic characteristics provided by an embodiment of the present invention;

[0020] Figure 5 It is a schematic diagram of a dynamic feature experimental environment provided by an embodiment of the present invention;

[0021] Figure 6 It is a schematic diagram of a normal situation of dynamic feature detection provided by an embodiment of the present invention;

[0022] Figure 7 This is a schematic diagram of a normal fault of dynamic feature detection provided by an embodiment of the present invention;

[0023] Figure 8It is a schematic diagram of a process for obtaining health-related data provided by an embodiment of the present invention;

[0024] Fig. 9 This is a schematic diagram of sensitive health characteristic data identification provided by an embodiment of the present invention;

[0025] Fig.10 The embodiment of the present invention provides a health assessment process of the instrument and a schematic diagram of the assessment reality;

[0026] Fig.11 The present invention provides a schematic diagram of quantitative rating of instrument health index. DETAILED DESCRIPTION

[0027] In order to further understand the content, features and effects of the present invention, the following embodiments are given as examples and described in detail with reference to the accompanying drawings.

[0028] The structure of the present invention is described in detail below in conjunction with the accompanying drawings.

[0029] like Figures 1 to 11As shown, an embodiment of the present invention provides a method for diagnosing instrument faults and predicting health under complex working conditions, and the method includes the following steps: S1: a method for diagnosing noise power spectrum density: the instrument will generate process noise during use, and the output signal will generate output noise. Both the process noise and the output noise are processed by a data processing algorithm to detect the type of fault; the high-frequency component of the instrument, that is, the process noise, can be used to detect sudden faults through signal processing; the low-frequency component of the instrument, that is, the output noise, can be used to detect slow-changing faults through signal processing; the characteristics of the instrument acquisition data are studied, the instrument acquisition data is decoupled and decomposed, and an effective data acquisition mechanism based on noise is established to solve the problem that typical instrument faults can only be diagnosed by external measures, S2: a method for extracting dynamic characteristics of the instrument and a method for evaluating instrument faults; when a fault occurs, it is detected, and if the dynamic characteristic response time changes, it is a fault; the dynamic characteristic value is the data obtained after filtering, the characteristic value is extracted, and then compared with the reference value to draw a fault conclusion; the fault will change the response time of the instrument S3: Collect the data used by the instrument, and then identify and extract the data and feature data to evaluate the health prediction of the instrument. The instrument collects a lot of data, and effectively identifying sensitive features is the key to health assessment; Study the extraction of sensitive health feature data based on the Spearman rank correlation coefficient; Study the data-driven instrument feature vector extraction and MD-based health indicator quantification method; Study the quantitative evaluation method of the instrument based on the health index; Study the online health assessment method under the operating conditions of the instrument to achieve real-time status assessment; S4: Build a typical instrument diagnosis platform and conduct field verification on the platform, realize field simulation of typical industrial environments and deployment of instruments; Realize real-time measurement of industrial sensor network data, and perform real-time fault diagnosis and health assessment.

[0030] Working principle of the present invention:

[0031] When in use, the method includes the following steps: a diagnostic method for noise power spectrum density: the instrument will generate process noise during use, and the output signal will generate output noise. Both the process noise and the output noise are processed through a data processing algorithm to detect the type of fault; the high-frequency component of the instrument, that is, the process noise, can be used to detect sudden faults through signal processing; the low-frequency component of the instrument, that is, the output noise, can be used to detect slow-changing faults through signal processing; a method for extracting dynamic characteristics of the instrument and a method for evaluating instrument faults; when a fault occurs, it is detected, and if the dynamic characteristic response time changes, it is a fault; the dynamic characteristic value is the data obtained after filtering, the characteristic value is extracted, and then compared with the benchmark value to draw a fault conclusion; the data used by the instrument is collected, and then the health prediction of the instrument is evaluated through data and characteristic data identification and extraction; a typical instrument diagnostic platform is built and field verification is carried out on the platform.

[0032] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0033] The above description is only a preferred embodiment of the present invention, and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this patent will not depart from the scope of the technical solution of the present invention.

Claims

1. A method for instrument fault diagnosis and health prediction under complex working conditions, characterized by: The method comprises the following steps: S1: Noise power spectrum density diagnostic method: The instrument will generate process noise during use, and the output signal will generate output noise. Both process noise and output noise are processed through data processing algorithms to detect the fault type. The high-frequency component of the instrument, that is, process noise, can be used to detect sudden faults through signal processing; The low-frequency component of the instrument, that is, the output noise, can be used to detect slow-changing faults through signal processing; S2: Methods for extracting instrument dynamic characteristics and evaluating instrument failures; When a fault occurs, it is detected, and if the dynamic characteristic response time changes, it is a fault; The dynamic characteristic value is the data obtained after filtering. The characteristic value is extracted and compared with the reference value to draw a fault conclusion; S3: Data-driven instrument health assessment method: Collect the data used by the instrument, and then identify and extract the data and feature data to evaluate the health prediction of the instrument; S4: Build a typical instrument diagnostic platform and conduct on-site verification on the platform.