Micro-vibration automatic evaluation and alarm system based on built-in VC criterion

Abstract

The invention relates to the technical field of vibration monitoring and diagnosis, and discloses a micro-vibration automatic evaluation and alarm system based on a built-in VC criterion, which comprises a data acquisition module, a storage module, a processing module and an alarm module. The system synchronously obtains a micro-vibration signal and a real-time working condition of a key vibration source through a data acquisition module; the processing module is internally provided with a vibration source characteristic spectrum decoupling module and an environment adaptive baseline evaluation module, and the vibration source characteristic spectrum decoupling module calls characteristic spectrum fingerprints from a preset working condition-vibration source correlation spectrum model library according to real-time working conditions and decouples real-time spectrums so as to realize attribution diagnosis of vibration sources. According to the method, the problems that in the prior art, only conformity judgment can be carried out, and diagnosis and prediction capabilities are lacked are solved, accurate source attribution can be carried out on complex mixed vibration, early warning of early abnormality and environmental deterioration trend is realized, and a set of integrated evaluation scheme from trend early warning to causal diagnosis is provided.

Description

Technical Field

[0001] This invention relates to the field of vibration monitoring and diagnostic technology, specifically to an automatic micro-vibration assessment and alarm system based on a built-in VC criterion. Background Technology

[0002] Micro-vibration control is crucial in high-precision fields such as semiconductor manufacturing, precision optical experiments, and biopharmaceuticals, as excessive vibration can directly affect product yield, experimental accuracy, and equipment operational stability. Currently, the common method for assessing environmental micro-vibration levels is to deploy vibration sensors at key monitoring points, collect vibration signals in real time, and compare their spectral characteristics with internationally accepted vibration standards to determine whether the environment meets the requirements of a specific process level.

[0003] However, existing technologies based on fixed thresholds have inherent limitations in practical applications. First, standards such as the VC criterion provide a set of static alarm thresholds that are independent of specific operating conditions. When the vibration at a monitoring point exceeds these thresholds, the system can only conclude that the vibration is "exceeding the limit," without diagnosing the cause. In complex industrial environments with multiple potential vibration sources (such as cooling pumps, air conditioning units, and process equipment), determining the specific source of vibration often requires manual investigation, which is time-consuming, labor-intensive, and inefficient. Furthermore, the system cannot automatically distinguish whether the excessive vibration is caused by a newly emerging equipment malfunction or by the normal operation of a group of equipment under a specific production formula, which can easily lead to false alarms or missed alarms.

[0004] Furthermore, this method is essentially a "post-event" alarm mechanism. It only responds when the vibration amplitude reaches or exceeds a preset limit, lacking sensitivity and early warning capabilities for long-term, trending changes caused by slow equipment performance degradation or gradual deterioration of environmental background vibrations that have not yet reached the alarm threshold. This lag prevents managers from conducting effective proactive maintenance, missing the opportunity to intervene in the early stages of a failure.

[0005] Therefore, how to provide a micro-vibration assessment method that can automatically identify vibration sources, adapt to different working conditions, and have trend warning capabilities is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides an automatic micro-vibration assessment and alarm system based on a built-in VC criterion. This system solves the problems of existing micro-vibration assessment technologies, which can only make judgments on exceeding limits based on fixed thresholds, but cannot automatically diagnose the source of vibration, and lack the ability to provide forward-looking early warning of vibration deterioration trends caused by equipment performance degradation or environmental changes.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an automatic micro-vibration assessment and alarm system based on a built-in VC criterion, comprising:

[0008] The data acquisition module is used to synchronously acquire the micro-vibration signals output from the monitoring points, as well as the real-time operating conditions associated with key vibration sources within the plant. Real-time operating conditions may include equipment start / stop status, load rate, or operating parameters such as rotational speed.

[0009] Storage module, which is used to store at least two of the following data models:

[0010] Operating Condition-Voltage Source Correlation Spectrum Model Library: This model library records the characteristic spectral fingerprints of one or more key vibration sources under specific operating conditions. The library is established through a model calibration module under controlled experiments. These controlled experiments involve independently running key vibration sources under specific operating conditions, acquiring their vibration signals, and extracting characteristic spectral fingerprints.

[0011] Environmental Gold Baseline: This baseline is a 1 / 3 octave band spectrum calibrated under ideal conditions where all key vibration sources are turned off, representing the optimal background vibration level of the monitoring environment.

[0012] The processing module, which is electrically connected to the data acquisition and storage modules, integrates a signal processing unit and the following two core evaluation and diagnostic modules:

[0013] Environmental Adaptive Baseline Assessment Module: This module assesses the stability of current environmental vibrations. It first establishes a dynamic health channel for the current real-time operating conditions based on collected historical vibration data. This dynamic health channel is determined by the statistical mean and standard deviation, with its upper bound U... rt (f k () is defined by the following formula:

[0014] U rt (f k )=μ rt (f k )+K σ ·σ rt (f k );

[0015] Among them, f k The center frequency of 1 / 3 octave band; U rt (f k () is the dynamic health channel at the center frequency f k The upper bound of the location; μ rt (f k (This refers to the real-time operating conditions at the center frequency f) k Historical vibration spectrum statistical mean at the location; σ rt (fk (This refers to the real-time operating conditions at the center frequency f) k The standard deviation of the historical vibration spectrum at K; σ This is a preset statistical confidence coefficient. When the real-time spectrum generated from the micro-vibration signal is at any center frequency f... k The amplitude on exceeds the upper limit U of the channel. rt (f k When this occurs, the module generates a fluctuation warning. In addition, the module also continuously compares the statistical mean μ. rt (f k The difference between the storage module and the environment's gold baseline is measured, and a drift warning is generated when the difference continues to exceed a preset drift threshold.

[0016] Vibration Source Feature Spectrum Decoupling Module: This module's function is to perform attribution diagnosis of vibration sources when needed. Based on the real-time operating conditions acquired by the data acquisition module, this module retrieves the feature spectrum fingerprints corresponding to all currently active key vibration sources from the operating condition-vibration source correlation spectrum model library stored in the storage module. Subsequently, this module calculates the contribution weight of each active feature spectrum fingerprint to the real-time spectrum by solving a linear decoupling model. When a key vibration source that is clearly off in the real-time operating conditions has a calculated contribution weight exceeding a preset ghost vibration threshold, this module determines that a ghost vibration anomaly exists. When there is a residual spectrum between the real-time spectrum and the weighted sum of the contributions of all active feature spectrum fingerprints, and the energy of this residual spectrum exceeds a preset decay threshold, this module determines that a performance degradation anomaly exists. Residual spectrum ε(f) k It is determined by the following formula:

[0017] ε(f k ) = V rt (f k )-∑ j w j ·F j (f k );

[0018] Wherein, ε(f k ) is at the center frequency f k The residual spectrum at V; rt (f k ) is at the center frequency f k Real-time spectrum at the location; w j The contribution weight of the key vibration source for the j-th activity; F j (f k The key vibration source of the j-th activity is at its center frequency f. k The module generates a characteristic spectral fingerprint of the location. Simultaneously, it can compare the real-time spectrum with a preset VC criterion alarm line to determine compliance.

[0019] The alarm module receives and presents diagnostic results and warning information generated by the processing module. Internally, this module includes an alarm priority arbitration unit that sorts different types of events (including but not limited to attribution diagnostic results, fluctuation warnings, drift warnings, ghost vibration alarms, performance degradation alarms, and VC criterion compliance judgment results) according to preset logic, and presents the highest priority event to the user interface.

[0020] A second aspect of this invention provides an automatic assessment and alarm method for micro-vibrations based on a built-in VC criterion, the method comprising the following steps:

[0021] S1. The micro-vibration signals of the monitoring points and the real-time operating conditions of key vibration sources are synchronously acquired through the data acquisition module.

[0022] S2. Generate a real-time spectrum based on the micro-vibration signal;

[0023] S3. Perform environmental adaptive baseline assessment: Establish a dynamic health channel for the current real-time operating conditions based on historical vibration data, and generate a fluctuation warning when the amplitude of the real-time spectrum deviates from the dynamic health channel;

[0024] S4. Perform vibration source characteristic spectrum decoupling analysis: Based on the real-time operating conditions, call the corresponding characteristic spectrum fingerprint from the preset operating condition-vibration source correlation spectrum model library, perform decoupling analysis on the real-time spectrum, calculate the contribution weight of each key vibration source, so as to achieve attribution diagnosis of vibration source.

[0025] S5. Perform comprehensive alarm presentation: Arbitrate the warning and diagnostic results generated in S3 and S4 according to the preset priority logic, and present the event information with the highest priority.

[0026] This invention provides an automatic micro-vibration assessment and alarm system based on a built-in VC criterion. It has the following beneficial effects:

[0027] 1. This invention establishes a working condition-vibration source correlation spectrum model library and uses a vibration source characteristic spectrum decoupling module to decouple and analyze the real-time spectrum based on the real-time working conditions, calculating the contribution weight of each active vibration source. This achieves accurate source attribution for complex mixed vibrations. Compared to traditional methods that only determine whether vibration exceeds limits, this invention can clearly identify which one or more key vibration sources are causing the vibration anomaly. Furthermore, it can discover deep faults such as ghost vibrations or performance degradation through counterfactual reasoning, thus providing direct and specific target guidance for equipment maintenance and troubleshooting.

[0028] 2. This invention utilizes an environmental adaptive baseline assessment module to dynamically establish a health channel containing statistical mean and standard deviation based on historical data for the current real-time operating conditions. By comparing the real-time spectrum with this dynamic channel, it achieves proactive early warning of micro-vibration anomalies. This method can detect early abnormal fluctuations that have not yet reached the fixed VC criterion alarm line but have deviated from their normal statistical behavior under specific operating conditions.

[0029] 3. This invention integrates the causal diagnosis results of the vibration source characteristic spectrum decoupling module with the trend warning results of the environmental adaptive baseline assessment module, and makes decisions through the priority arbitration unit within the alarm module. This provides a multi-dimensional, integrated assessment conclusion from compliance determination and trend warning to source diagnosis, avoiding the information silo problem caused by single indicators. This allows users to obtain a complete and logically consistent diagnostic report, thereby significantly improving the interpretability of the assessment results and the effectiveness of guiding operations. Attached Figure Description

[0030] Figure 1 This is a system architecture diagram of the present invention;

[0031] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] Example:

[0034] Please see the appendix Figure 1 - Appendix Figure 2 This invention provides an automatic micro-vibration assessment and alarm system based on a built-in VC criterion, comprising:

[0035] The data acquisition module is used to simultaneously acquire the micro-vibration signals of the monitoring points and the real-time operating conditions of key vibration sources;

[0036] In this embodiment, the data acquisition module is a multi-source data synchronous acquisition unit. Its core function is to provide accurate, complete and strictly time-synchronized input data for the subsequent processing module. This input data is the technical basis for realizing the vibration source attribution diagnosis and environmental trend prospective early warning of the present invention.

[0037] The structure and functions of the data acquisition module can be explained in the following aspects:

[0038] Acquisition of micro-vibration signals

[0039] In this embodiment, the data acquisition module connects to one or more vibration sensors deployed at monitoring points via multiple physical channels. The monitoring points are typically located on equipment foundations, factory floors, or precision instrument stands that are sensitive to micro-vibrations.

[0040] Preferably, the vibration sensor can be a piezoelectric accelerometer, a MEMS accelerometer, or an optical vibration sensor to capture high-bandwidth, high-sensitivity vibration information. This module performs necessary signal conditioning on the analog signal from the sensor, including charge amplification (as with piezoelectric sensors) and anti-aliasing filtering, and then converts it into a discrete digital time-series signal, i.e., the micro-vibration signal x, via a high-precision analog-to-digital converter (ADC). rt (t). This signal forms the basis of the raw data for all subsequent spectral analyses.

[0041] Parallel acquisition of real-time operating conditions

[0042] Unlike conventional systems that only collect vibration signals, a key feature of the data acquisition module in this embodiment is its ability to acquire real-time operating conditions C associated with the key vibration source in parallel. rt .

[0043] To this end, the module has one or more standardized data interfaces, such as an industrial Ethernet interface, an RS-485 serial interface, or an OPCUA (Open Platform Communication Unified Architecture) client. Through these interfaces, the module can communicate with the factory's Manufacturing Execution System (MES), Programmable Logic Controller (PLC), Supervisory Control System (SCADA), or the control unit of the equipment itself.

[0044] Through the communication connection, this module can query and obtain the operating parameters of each monitored key vibration source in real time and periodically. These parameters constitute the real-time operating conditions, which may include, but are not limited to: the start / stop status of the equipment, the current load percentage, the actual speed of the motor, the working pressure of the vacuum pump, the name or batch number of a specific processing program, etc.

[0045] The purpose of acquiring these real-time operating conditions is to provide the necessary contextual information for subsequent processing modules to conduct in-depth analysis. For example, the construction and invocation of the operating condition-vibration source correlation spectrum model library, as well as the accurate establishment of the dynamic health channel, both rely on binding vibration data with the exact operating conditions at the time the vibration occurred.

[0046] Data stream timestamp synchronization mechanism

[0047] A key technical feature of this embodiment is that the data acquisition module integrates a high-precision timestamp synchronization mechanism. This mechanism ensures strict alignment of the micro-vibration signal with the real-time operating conditions on the time reference.

[0048] Specifically, the module employs a unified internal or external clock source (e.g., synchronized via the Network Time Protocol NTP). When the module samples the micro-vibration signal, it appends a high-resolution timestamp to each sampling point or each data frame. Simultaneously, when the module receives any frame of real-time operating data through the data interface, it immediately appends a timestamp of the same reference to it.

[0049] This rigorous time synchronization mechanism ensures that any micro-vibration data output by this module at any given time point corresponds precisely to the operating status of all key vibration sources at that point in time. This characteristic is crucial for subsequent processing modules to perform causal analysis, enabling the vibration source characteristic spectrum decoupling module to accurately decouple a mixed real-time spectrum V rt (f k It is decomposed into a linear superposition of contributions from different vibration sources under specific working conditions, thereby ensuring the effectiveness of attribution diagnosis.

[0050] Output of composite data stream

[0051] Finally, the data acquisition module outputs the collected and processed data to the processing module as a composite data stream. Each data unit in this composite data stream contains a set of synchronization information, such as a triple (timestamp, x...). rt (t),C rt These represent the acquisition time, the sampled value of the micro-vibration signal at that time, and the set of real-time operating conditions of the equipment at that time, respectively. This structured data output provides a regular and reliable data input for the implementation of all subsequent advanced analysis functions.

[0052] The storage module is used to store a preset working condition-vibration source correlation spectrum model library. The working condition-vibration source correlation spectrum model library records the characteristic spectral fingerprint of at least one key vibration source under a specific working condition.

[0053] In this embodiment, the storage module is a structured, persistent data storage unit. Its core function is to provide the necessary data support for the system's model calibration, real-time analysis, and historical tracing. This module is not merely a simple information recording unit, but also the core knowledge base and data foundation required for achieving causal diagnosis and proactive early warning in this invention.

[0054] The contents stored in the storage module can be divided into the following main parts:

[0055] Storage of the core model library

[0056] This section stores a relatively static prior knowledge model established through offline calibration or system configuration, which serves as the foundation for advanced analysis by subsequent processing modules.

[0057] First, this storage module stores the Operating Condition-Source Correlation Spectrum Model Library (OSCS). This library exists as a structured database or associated file, used to accurately record the vibration characteristics of each key vibration source under different operating conditions. Specifically, each record in the library is a triplet (s...). i ,c i j,F i j). Wherein, s i It is the unique identifier of the critical vibration source; c i j is an identifier for a specific operating condition of the vibration source; while F i j is the corresponding feature spectral fingerprint, pre-extracted through the model calibration module. Feature spectral fingerprint F i j itself is a multidimensional vector, which at least contains the vector at the center frequency f. k The velocity spectrum amplitude. The purpose of this storage structure is to provide the vibration source characteristic spectrum decoupling module with a value that can be adjusted according to real-time operating conditions during real-time diagnosis. rt A reference base with clear physical meaning that can be quickly queried and invoked.

[0058] Secondly, this storage module is also used to store the environmental gold baseline. This baseline is stored as a one-dimensional spectrum vector V. gb (f k The baseline is obtained through long-term data collection and statistical averaging under ideal, quiet operating conditions without critical vibration sources. The purpose of storing this baseline is to provide the environmental adaptive baseline assessment module with a long-term, stable, and absolute reference representing the optimal environmental state, thereby effectively determining the overall, slow drift trend of the environmental background vibration.

[0059] Storage of historical and dynamic data

[0060] To achieve the adaptive evaluation capability of this invention, the storage module is also configured with a region for storing massive amounts of time-series data.

[0061] This area is used to continuously and sequentially record the composite data stream output by the data acquisition module. Each stored record is strictly time-stamped and contains the real-time spectrum of micro-vibrations at that moment. rt (f k ) and the real-time operating condition C at that moment. rt Perform association and binding.

[0062] The purpose of storing this historical data is to provide the necessary dataset for the dynamic health channel established by the environmental adaptive baseline assessment module. When the processing module needs to assess the vibration health under the current real-time operating conditions, it retrieves all historical spectral data that are the same as or similar to the current operating conditions from this historical data area, and dynamically calculates the statistical mean μ based on these data. rt (f k ) and standard deviation σ rt (f k Therefore, this historical data storage area is the source of data for the system to achieve its "adaptive" capability.

[0063] System configuration and log storage

[0064] In addition, the storage module is also used to persist various configuration parameters and operation logs of the storage system.

[0065] Configuration parameters include, but are not limited to: user-defined VC criterion alarm thresholds and statistical confidence coefficient K used to define the dynamic health channel width. σ The system includes parameters such as a drift threshold for drift warning, a ghost vibration threshold for ghost vibration detection, and a degradation threshold for performance degradation. Storing these parameters ensures that the system can recover to a preset operating state after a restart or power failure.

[0066] The operation log records every alarm or warning event generated by the system in detail, including the event type, occurrence time, related diagnostic results (such as a list of contributing sources), and the real-time spectrum and operating conditions at that time. The purpose of storing these logs is to provide a complete and reliable record for subsequent troubleshooting, equipment maintenance history tracing, and system performance evaluation.

[0067] In summary, the storage module in this embodiment constitutes the memory center and knowledge center of the entire system through the structured storage of core models, dynamic historical data, system configurations, and operation logs. The completeness and accuracy of its content are prerequisites for ensuring the realization of advanced analysis functions of the system.

[0068] The processing module, connected to the data acquisition module and the storage module, is adapted to perform the following operations:

[0069] Real-time spectrum is generated based on micro-vibration signals;

[0070] The environmental adaptive baseline assessment module establishes a dynamic health channel based on historical vibration data and generates a fluctuation warning when the real-time spectrum deviates from the dynamic health channel.

[0071] The vibration source characteristic spectrum decoupling module calls the corresponding characteristic spectrum fingerprint from the working condition-vibration source correlation spectrum model library according to the real-time working conditions, and performs decoupling analysis on the real-time spectrum to achieve attribution diagnosis of vibration source.

[0072] In this embodiment, the processing module is the central computing and analysis core of the system. It interacts with the data acquisition module and the storage module to execute a series of complex signal processing, model evaluation and diagnostic reasoning algorithms, thereby transforming the original vibration and operating condition data into evaluation conclusions and early warning information with clear physical meaning.

[0073] The internal functionality of this processing module can be explained by dividing it into the following logically interconnected units:

[0074] Signal preprocessing and real-time spectrum generation unit

[0075] This unit forms the basis for all subsequent analyses. It receives the micro-vibration time-domain signal x, with precise timestamps, from the data acquisition module. rt (t).

[0076] To perform frequency domain analysis, this unit first normalizes the time-domain signal. Preferably, this process includes applying a window function (e.g., a Hanning window) to suppress spectral leakage, followed by performing a Fast Fourier Transform (FFT) to transform it from the time domain to the frequency domain. To align with common microvibration assessment criteria (such as the VC criterion), this unit further passes the linear spectrum obtained from the FFT through a set of 1 / 3 octave band filters, ultimately generating the real-time spectrum V. rt (f k ), where f k This represents the center frequency of each 1 / 3 octave band. This real-time spectrum is the direct input for all subsequent evaluation and diagnostic modules.

[0077] Environmental Adaptive Baseline (AEB) Assessment Module

[0078] This module aims to provide a dynamic and forward-looking assessment of the background vibration health of the monitored environment. Its core feature is that it is not based on a fixed absolute threshold, but on a data-driven statistical model.

[0079] First, this module uses real-time operating conditions C obtained from the data acquisition module. rt The system queries the storage module to retrieve all historical vibration spectrum data recorded under this operating condition. Based on this historical dataset, the module dynamically constructs a statistically significant dynamic health channel for the current real-time operating condition. This channel is composed of the statistical mean μ. rt (f k ) and standard deviation σ rt (f kThe upper bound of its channel, U, is determined. rt (f k It can be defined by the following formula:

[0080] U rt (f k )=μ rt (f k )+K σ ·σ rt (f k );

[0081] Among them, U rt (f k () is the dynamic health channel at the center frequency f k The upper bound at μ represents the statistically normal fluctuation range of the vibration amplitude under that specific operating condition; rt (f k () is the real-time operating condition at the center frequency f k The historical vibration spectrum mean at the location; σ rt (f k K represents the standard deviation of the corresponding historical vibration spectrum. σ This is a preset statistical confidence coefficient used to adjust the channel width.

[0082] Based on this dynamic health channel, this module performs dual-mode early warning judgment. The first is a fluctuation warning: when the real-time spectrum V... rt (f k At any center frequency f k The amplitude on the upper bound of the corresponding channel U exceeded the upper bound of the upper bound. rt (f k When this occurs, the module generates a fluctuation warning. The significance of this warning lies in the fact that it indicates a momentary, abnormal vibration event that deviates from the normal statistical behavior under the current operating conditions.

[0083] Secondly, there is a drift warning: this module continuously updates the statistical mean μ rt (f k ) and the environmental gold baseline V stored in the storage module gb (f k The two are compared. If the difference between the two shows a continuous, monotonous increasing trend, and the cumulative amount of the difference exceeds a preset drift threshold, the module generates a drift warning. The purpose of this warning is to detect long-term background vibration deterioration trends caused by slow, collective degradation of equipment or structural changes in the environment.

[0084] Vibration Source Characteristic Spectrum Decoupling (VSSD) Module

[0085] This module is the core of the causal diagnostic function of this invention. It is activated when the system receives a warning signal or at the request of the user to analyze the specific physical source behind complex mixed vibrations.

[0086] The analysis in this module begins with a context-sensitive query. It is based on real-time operating conditions C. rt The system queries the operating condition-vibration source correlation spectrum model library in the storage module and filters out all key vibration sources that are currently in the "on" state and their corresponding health characteristic spectrum fingerprints F. j (f k These constitute a candidate fingerprint set.

[0087] Subsequently, this module constructs a linear decoupling model, whose basic assumption is that the real-time spectrum is the sum of the linear superposition of the spectra generated by all active vibration sources and a residual spectrum. This model can be expressed as:

[0088] V rt (f k )=∑ j w j ·F j (f k )+ε(f k );

[0089] Among them, w j ε(f) represents the contribution weight of the j-th active vibration source to the current total vibration, and is the unknown to be solved; k The residual spectrum (w) cannot be explained by known fingerprints. To determine the contribution weights w... j This module solves a non-negative constrained least squares (NNLS) optimization problem to minimize the sum of squares of the fitting error at all center frequencies.

[0090] After solving for the contribution weights w, j Then, this module performs in-depth attribution diagnosis and counterfactual reasoning. The most direct result is that the weight w j The magnitude of the value directly reflects the contribution of each active vibration source to the current real-time spectrum, thus realizing the quantitative attribution of vibration sources.

[0091] More importantly, this module also performs two counterfactual inferences. One is ghost vibration diagnosis: this module checks a vibration in real-time under condition C. rt The key vibration source is clearly identified as being in a "closed" state, and its corresponding contribution weight w k Does it significantly exceed a preset ghost vibration threshold? If so, it is determined that there is an abnormal ghost vibration, such as equipment malfunction or serious electrical fault.

[0092] Secondly, performance degradation diagnosis: This module calculates the residual spectrum ε(f kThe energy of the device is considered as follows: If the energy exceeds a preset degradation threshold, it indicates that at least one critical vibration source is operating and its actual vibration characteristics have deviated from its healthy fingerprint stored in the model library. This suggests that the device may be showing signs of early performance degradation, such as bearing wear or loose components.

[0093] Finally, as a basic function, this processing module will also process the real-time spectrum V. rt (f k The alarm is directly compared with the VC criterion alarm line configured by the user in the storage module to determine compliance with the standard. All analysis results, including warnings, diagnostic conclusions, and compliance determination results, are output to the alarm module in a structured manner.

[0094] The alarm module is used to present attribution diagnosis results and fluctuation warnings based on the operation results of the processing module.

[0095] In this embodiment, the alarm module is the final human-computer interaction interface of the system. Its core function is to transform the complex and multi-dimensional analysis results output by the processing module into structured alarm and diagnostic information with clear priorities and easy user understanding. This module is not a simple audible and visual alarm unit, but a comprehensive presentation terminal integrating decision-making logic and information visualization functions.

[0096] The internal structure and workflow of the alarm module can be explained in the following aspects:

[0097] Alarm Priority Arbitration Unit

[0098] This unit is the decision-making center of the alarm module. It receives all event streams from the processing module, which may include: compliance determination results based on the VC criterion, fluctuation and drift warnings generated by the environmental adaptive baseline assessment module, and a list of contributing sources, ghost vibration alarms, and performance degradation alarms generated by the vibration source characteristic spectrum decoupling module.

[0099] To avoid information confusion caused by multiple concurrent events and to ensure that operators can focus on the most critical issues first, the arbitration unit has a pre-set fixed alarm priority queue. Preferably, this queue is sorted according to the urgency of the event, its potential hazard, and the certainty of the diagnosis. A typical priority setting could be: Ghost vibration alarm > Performance degradation alarm > High VC alarm > Fluctuation warning > High VC alarm > Drift warning.

[0100] At any given time, the arbitration unit compares all concurrent events and selects only the event with the highest current priority as the final system state for output. The purpose of this mechanism is to ensure that the user is presented with the core issues that most need attention.

[0101] Integrated diagnostic information presentation

[0102] After the arbitration unit determines the highest priority event, the alarm module generates an integrated, multimodal diagnostic report, which is presented through a graphical user interface.

[0103] A core component of this interface is a comprehensive spectrum. This spectrum uses 1 / 3 octave bands as the horizontal axis and vibrational velocity as the vertical axis, overlaying multiple layers of information simultaneously to facilitate intuitive comparison and analysis. These information layers include at least:

[0104] Real-time spectrum curve V representing the current vibration state rt (f k ).

[0105] The VC criterion alarm line represents the static, absolute threshold.

[0106] The dynamic health channel represents the normal statistical range under the current operating conditions, i.e., the upper bound U of the channel. rt (f k ) and lower bound.

[0107] When attribution diagnosis is triggered, the spectrum will also display the characteristic spectral fingerprint F of the identified major contributing source using curves of different colors or styles. j (f k Furthermore, its magnitude can be determined based on its contribution weight w. j Scaling is possible. This multi-information overlay visualization method aims to correlate abstract diagnostic conclusions with raw spectral data, providing technicians with a data view that allows for in-depth analysis and verification.

[0108] In addition to graphical information, the interface also provides diagnostic conclusions in explicit text form. These conclusions directly state the highest priority event type currently triggered and provide crucial contextual information. For example, when a performance degradation alarm is triggered, the text will indicate that residual spectral energy exceeds limits; when a contributor attribution alarm is triggered, the text will list the names of the primary contributors and their quantified percentage contribution.

[0109] Maintaining and recording alarm status

[0110] To ensure that important events are not overlooked, the alarm module also includes an alarm status maintenance mechanism. Once a high-level alarm is triggered (e.g., a ghost vibration alarm), the overall alarm status of the system will be "locked" at that level. Even if the triggering conditions for the event subsequently disappear, the alarm status will remain on the interface until it is explicitly acknowledged or reset by the user.

[0111] Simultaneously, for each event with the highest priority determined by the arbitration unit, its complete diagnostic information (including event type, occurrence timestamp, spectrum snapshot, diagnostic conclusion, etc.) is packaged into a structured log record by the alarm module and sent to the storage module for permanent storage. The purpose of this is to provide a traceable and complete historical event archive for equipment maintenance and process optimization.

[0112] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A micro-vibration automatic assessment and alarm system based on a built-in VC criterion, characterized in that, include: The data acquisition module is used to simultaneously acquire the micro-vibration signals of the monitoring points and the real-time operating conditions of key vibration sources; The storage module is used to store a preset working condition-vibration source correlation spectrum model library. The working condition-vibration source correlation spectrum model library records the characteristic spectral fingerprint of at least one key vibration source under a specific working condition. The processing module, connected to the data acquisition module and the storage module, is adapted to perform the following operations: Real-time spectrum is generated based on micro-vibration signals; The environmental adaptive baseline assessment module establishes a dynamic health channel based on historical vibration data and generates a fluctuation warning when the real-time spectrum deviates from the dynamic health channel. The vibration source characteristic spectrum decoupling module calls the corresponding characteristic spectrum fingerprint from the working condition-vibration source correlation spectrum model library according to the real-time working conditions, and performs decoupling analysis on the real-time spectrum to achieve attribution diagnosis of vibration source. The alarm module is used to present attribution diagnosis results and fluctuation warnings based on the operation results of the processing module.

2. The system according to claim 1, characterized in that, The vibration source characteristic spectrum decoupling module is specifically used for: By solving a linear decoupling model, the contribution weights of key vibration sources of each activity to the real-time spectrum are determined, thereby enabling attribution diagnosis.

3. The system according to claim 1, characterized in that, The dynamic health channel is determined by the statistical mean and standard deviation calculated based on historical data, and its upper bound U is... rt (f k () is defined by the following formula: U rt (f k )=μ rt (f k )+K σ ·s rt (f k ); Among them, U rt (f k () is the dynamic health channel at the center frequency f k The upper bound of μ rt (f k () is the real-time operating condition at the center frequency f k The historical vibration spectrum mean at the location, σ rt (f k () is the real-time operating condition at the center frequency f k The standard deviation of the historical vibration spectrum at K σ This is the preset statistical confidence coefficient.

4. The system according to claim 1, characterized in that, The storage module is also used to store a preset environmental baseline; The environment adaptive baseline assessment module is also used for: By comparing the statistical mean of the dynamic health channel with the environmental golden baseline, a drift warning is generated when the difference continues to exceed a preset drift threshold.

5. The system according to claim 1, characterized in that, The system also includes a model calibration module, used for: Under controlled experiments, the characteristic spectral fingerprints of key vibration sources under specific working conditions were calibrated and extracted to construct a working condition-vibration source correlation spectrum model library.

6. The system according to claim 2, characterized in that, The vibration source characteristic spectrum decoupling module is also used for: When a critical vibration source that is in a closed state during real-time operation has a calculated contribution weight that exceeds a preset ghost vibration threshold, a ghost vibration alarm is generated.

7. The system according to claim 2, characterized in that, The vibration source characteristic spectrum decoupling module is also used to generate a performance degradation alarm when a residual spectrum exists between the real-time spectrum and the weighted sum of contributions from all active key vibration sources, and the energy of the residual spectrum exceeds a preset degradation threshold; wherein, the residual spectrum ε(f k It is determined by the following formula: ε(f k )=V rt (f k )-∑ j w j ·F j (f k ); Wherein, ε(f k ) is at the center frequency f k The residual spectrum at V rt (f k ) is at the center frequency f k Real-time spectrum at w j F represents the contribution weight of the key vibration source for the j-th activity. j (f k The key vibration source of the j-th activity is at its center frequency f. k The characteristic spectral fingerprint at the location.

8. The system according to claim 1, characterized in that, The processing module is also used for: The real-time spectrum is compared with the preset VC criterion alarm line to determine compliance.

9. The system according to claim 1, characterized in that, The alarm module also includes an alarm priority arbitration unit, which is used to prioritize attribution diagnosis results, fluctuation warnings, ghost vibration alarms or performance degradation alarms according to preset logic, and present the highest priority event.

10. A method for automatic assessment and alarm of micro-vibrations based on a built-in VC criterion, characterized in that, Includes the following steps: S1. Simultaneously acquire the micro-vibration signals of the monitoring points and the real-time operating conditions of key vibration sources; S2. Generate a real-time spectrum based on the micro-vibration signal; S3. Establish a dynamic health channel based on historical vibration data, and generate a fluctuation warning when the real-time spectrum deviates from the dynamic health channel; S4. Based on the real-time operating conditions, call the corresponding characteristic spectrum fingerprint from the preset operating condition-vibration source correlation spectrum model library to perform decoupling analysis on the real-time spectrum in order to achieve attribution diagnosis of vibration source. S5. Based on the results of the attribution diagnosis and the fluctuation warning, an alarm will be displayed.

Images