Chemical production mechanical equipment fault self-diagnosis management system based on intelligent inspection

Abstract

The invention discloses a chemical production mechanical equipment fault self-diagnosis management system based on intelligent inspection, and relates to the technical field of production mechanical equipment diagnosis management. A vibration signal self-adaptive acquisition module of the scheme carries out self-adaptive acquisition on a vibration signal of chemical equipment, and vibration spectrum data is obtained based on the vibration signal of the chemical equipment; the vibration spectrum data transmission analysis module is used for carrying out vibration spectrum data transmission analysis after the self-adaptive acquisition process is finished to obtain a transmission analysis result, and deciding whether to carry out vibration spectrum data dynamic transmission or not based on the transmission analysis result; the self-adaptive threshold value judgment module is used for carrying out self-adaptive threshold value judgment on the vibration spectrum data after the vibration spectrum data transmission analysis process is finished, and deciding whether to carry out chemical equipment fault self-diagnosis management or not based on an obtained judgment result; the problem that accurate recognition and advanced early warning of early faults of chemical production mechanical equipment are difficult to achieve through intelligent inspection is solved.

Description

Technical Field

[0001] This invention relates to the field of diagnostic management technology for production machinery and equipment, and in particular to a self-diagnostic management system for faults in chemical production machinery and equipment based on intelligent inspection. Background Technology

[0002] With the sweeping wave of industrial intelligence and the upgrading of the chemical industry towards large-scale and refined production, intelligent inspection technology is being applied more and more deeply in various industrial fields. It has become a core path to solve equipment operation and maintenance problems and strengthen the production safety line. Intelligent inspection of machinery in chemical production is particularly important. The core carriers of chemical production are various types of chemical production machinery, from reactors and compressors that withstand high temperatures and pressures for a long time, to pumps, valves and heat exchangers responsible for media transportation. The stable operation of these devices not only directly determines the continuity of production, but is also closely linked to the management and control of electrical equipment such as explosion-proof motors and control cabinets, as well as hazardous materials such as flammable and explosive solvents and reaction intermediates. Together, they form a closed loop for safety in chemical production. Currently, the operating conditions of chemical production machinery are becoming increasingly complex, requiring it to withstand multiple challenges such as corrosion, vibration, and media erosion. Even a minor malfunction in a single machine can trigger a chain of risks. Therefore, from early fixed-point inspections relying on manual experience, to mechanized tools capturing single operating parameters, and now to intelligent inspection achieving multi-dimensional perception, data analysis, and intelligent judgment, technological upgrades have consistently focused on fault prediction and risk prevention for chemical production machinery. This has allowed intelligent inspection and fault self-diagnosis to gradually transcend the scope of basic maintenance, becoming a key support for coordinating the stability of chemical machinery, electrical safety protection, and the management of explosives. Against this backdrop, the chemical industry has set stringent standards for core indicators such as the accuracy of anomaly identification, the speed of early warning response, and cross-equipment collaboration capabilities of intelligent inspection. Ensuring that faults in chemical production machinery and related processes are captured promptly and diagnosed accurately has become paramount in guaranteeing production safety.

[0003] However, traditional methods of fault diagnosis and inspection management for chemical production machinery rely heavily on manual experience or single-parameter monitoring, lacking systematic data fusion analysis and precise intelligent judgment capabilities. Faced with the complex demands of massive equipment operation and multiple overlapping risks in modern chemical production lines, this approach struggles to quickly adapt to fault characteristics under different operating conditions, easily leading to problems such as missed faults, misjudgments, and delayed early warnings. These issues not only reduce operation and maintenance efficiency and cause unplanned equipment downtime, but may also lead to major safety accidents such as leaks, explosions, and fires due to untimely fault handling. Therefore, improving the accuracy and timeliness of fault self-diagnosis under the intelligent inspection system has become a core issue that the chemical production industry urgently needs to address.

[0004] To achieve intelligent inspection of chemical production machinery and equipment (such as reaction vessels), methods include: using non-contact infrared thermal imagers to perform full-area temperature scanning and collect temperature data for key components; using ultrasonic leak detectors to detect gas leaks and partial electrical discharges during operation; and monitoring vibration using high-frequency piezoelectric accelerometers. For intelligent vibration inspection, firstly, vibration signals are continuously collected by the high-frequency piezoelectric accelerometer and converted into vibration spectrum data via Fast Fourier Transform (FFT). This vibration spectrum data refers to the conversion of the time-domain signal collected by the vibration sensor into a frequency domain spectrum showing the distribution of vibration energy at different frequencies, including vibration frequency and amplitude, through mathematical transformation (such as FFT). The obtained vibration spectrum data is transmitted to a central server or edge computing node via industrial Ethernet and wireless network. Then, the rule engine triggers a primary alarm for data exceeding the threshold or abnormal rate of change. The rule engine is a configurable logic processing core embedded in the diagnostic software. It encodes expert experience and industry standards into computer-executable judgment logic through "if-then" rules. For example, if the temperature of the stirring rod of the reactor is >85°C, a "high temperature alarm" is triggered. Based on this, the vibration spectrum data is matched with the built-in fault feature library to accurately identify typical mechanical faults such as reactor rotor imbalance, stirring rod wear, and abnormal gear meshing. In addition, the trend analysis of key performance indicators is carried out by combining historical data and real-time data, ultimately realizing a leapfrog transformation of chemical production machinery and equipment from traditional periodic inspection to adaptive intelligent operation and maintenance management.

[0005] During the intelligent inspection of chemical production machinery and equipment (such as reactors), when the reactor is handling high-viscosity materials, these materials tend to adhere to the surface of the stirring blades, creating a strong coupling effect. This results in the stirring rod of the reactor being subjected to non-uniform, highly variable periodic loads. These periodic loads excite the vibration of the stirring rod and stirring blades, ultimately generating vibration signals with complex modulation characteristics. This causes the vibration spectrum of the actual state of the chemical production machinery and equipment malfunction (such as loose stirring blades) to overlap with the broadband, low signal-to-noise ratio vibration spectrum, making it difficult to effectively capture using traditional time-domain indicators. Therefore, it is necessary to collect vibration spectrum data based on high sampling rates. This leads to a massive increase in vibration spectrum data transmitted through industrial Ethernet and wireless networks, crowding out limited network bandwidth and causing congestion and instability in industrial Ethernet and wireless networks. Consequently, it causes communication delays between edge computing nodes and the central server, resulting in damage to the temporal consistency and integrity of the vibration spectrum data. Consequently, the central server receives distorted or delayed vibration spectrum data.

[0006] At the same time, because the judgment logic of existing rule engines mainly relies on fixed thresholds and "if-then" rules, the rule engines are difficult to adapt to the individual differences in the operation of chemical production machinery and equipment, such as the changes in viscosity, temperature and speed of different batches of materials. These differences cause the uniform thresholds set based on historical or ideal operating conditions to frequently generate false alarms (judging normal operating condition fluctuations as abnormal) or false alarms (failing to identify faults that are abnormal under new operating conditions), and cannot adaptively obtain the correlation between the current real state and real-time operating conditions of chemical production machinery and equipment.

[0007] Ultimately, the combination of the above problems results in insufficient compatibility between the management strategy for self-diagnostic faults of chemical production machinery and equipment and intelligent inspection. It is difficult to extract accurate fault information from distorted vibration spectrum data, and intelligent inspection is unable to accurately identify and provide early warnings of early faults in chemical production machinery and equipment. Therefore, the fault warning of chemical production machinery and equipment suffers from low accuracy and timeliness. Summary of the Invention

[0008] This invention provides a self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection. The technical solution provided by this application is as follows: The vibration signal adaptive acquisition module is used to adaptively acquire vibration signals from chemical equipment, obtain corresponding vibration spectrum data based on the vibration signals, and determine whether to perform vibration spectrum data verification based on the vibration spectrum data. The vibration spectrum data transmission and analysis module is used to perform vibration spectrum data transmission and analysis on the transmitted vibration spectrum data after the adaptive acquisition process, obtain corresponding transmission analysis results, and decide whether to implement dynamic transmission of vibration spectrum data based on the transmission analysis results. The adaptive threshold determination module is used to perform adaptive threshold determination on the vibration spectrum data after the vibration spectrum data transmission and analysis process, and determine whether to implement chemical equipment fault self-diagnosis management based on the determination results.

[0009] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. By adaptively acquiring vibration signals from chemical equipment and performing vibration spectrum data verification, it helps to measure the stability of the equipment's operating status and the degree of potential fault risks, improves the reliability and effectiveness of vibration spectrum data, enhances the accuracy of fault feature identification, and strengthens the anti-interference ability against non-uniform loads, broadband noise, and other interference factors. Vibration spectrum data transmission analysis helps to assess network bandwidth adaptability and the timeliness and completeness of data transmission, improving the stability and anti-interference ability of vibration spectrum data transmission, thereby enhancing the authenticity and timing consistency of data received by the central server. Adaptive threshold determination of vibration spectrum data, and based on the acquired determination results, determining whether to implement self-diagnosis management of chemical equipment faults, can effectively assess the significance of chemical equipment fault characteristics and their adaptability to real-time operating conditions, improving the accuracy of fault identification and the timeliness of early warnings. It achieves dynamic matching of threshold standards and personalized operating conditions, thereby realizing accurate identification and early warning of early faults in chemical production machinery and equipment.

[0010] 2. Implementing wavelet packet decomposition of vibration spectrum signals helps to enhance the identification and extraction efficiency of fault features. Compared with existing technologies, when the reactor is processing high-viscosity materials, the high-viscosity materials tend to adhere to the surface of the stirring blades, which will generate a strong coupling effect with the stirring blades. This will cause the stirring rod of the reactor to be subjected to non-uniform and highly variable periodic loads. This solution helps to accurately locate abnormal frequency bands affected by non-uniform periodic loads. Then, the fault feature signal reconstruction further removes redundant noise and focuses on core fault features, reducing the risk of missed or misjudged faults in chemical equipment caused by interference.

[0011] 3. Fault feature signal reconstruction can improve the purity and recognizability of fault features in chemical equipment. Compared with existing technologies, periodic loads can excite the vibration of stirring rods and blades, generating vibration signals with complex modulation features. This causes the vibration spectrum of the actual state of chemical production machinery and equipment faults (such as loose stirring blades) to overlap with the vibration spectrum of wideband, low signal-to-noise ratio vibration. This solution helps to accurately retain effective signals that reflect the fault features of chemical equipment, eliminate the conventional noise generated by the operation of chemical equipment, improve the extraction accuracy of early fault features, and provide an accurate and reliable basis for the self-diagnosis of chemical equipment faults.

[0012] 4. Dynamic transmission of vibration spectrum data can effectively enhance the timing consistency and integrity of vibration spectrum data transmission. Compared with existing technologies, this leads to a massive increase in vibration spectrum data transmitted via industrial Ethernet and wireless networks, which consumes limited network bandwidth, causes congestion and instability in industrial Ethernet and wireless networks, and consequently causes communication delays between edge computing nodes and the central server. This results in damage to the timing consistency and integrity of vibration spectrum data, causing the central server to receive distorted or delayed vibration spectrum data. This solution is used to ensure the timeliness and quality of vibration spectrum data and guarantee the priority transmission efficiency of highly urgent vibration spectrum data. Attached Figure Description

[0013] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart of the self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection provided in an embodiment of the present invention; Figure 3 This is a diagram illustrating the vibration spectrum data verification operation architecture of the intelligent inspection-based fault self-diagnosis management system for chemical production machinery and equipment provided in this embodiment of the invention. Figure 4 This is an adaptive threshold determination architecture diagram of the self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection provided in an embodiment of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0016] Before providing a detailed explanation of the embodiments of this application, the application scenarios of these embodiments will be described first.

[0017] Example 1: This embodiment of the invention provides a self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection. For example... Figure 1 The flowchart shown is for a fault self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection. The system's processing flow may include the following steps: First, adaptive vibration signal acquisition is performed to determine whether the kurtosis of the chemical equipment vibration signal is within the preset kurtosis range and whether the amplitude ratio of the chemical equipment vibration signal is less than or equal to the preset amplitude ratio. If so, the corresponding vibration spectrum data is marked as Level 3 vibration spectrum data and stored in the edge computing node for vibration spectrum data transmission and analysis. If not, there are two possible outcomes: First, if the kurtosis of the chemical equipment vibration signal is not within the preset kurtosis range but the amplitude ratio is greater than the preset amplitude ratio, the corresponding vibration spectrum data is marked as Level 1 vibration spectrum data, directly transmitted to the central server, and a Level 1 warning is sent to the preset personnel to stop the operation of the chemical machinery. Second, if either of the following two situations exists: Situation 1: The kurtosis of the chemical equipment vibration signal is within the preset kurtosis range and the amplitude ratio is greater than the preset amplitude ratio; Situation 2: If the kurtosis of the vibration signal from the chemical equipment is not within the preset kurtosis range, and the amplitude ratio of the vibration signal is less than or equal to the preset amplitude ratio, the corresponding vibration spectrum data is marked as secondary vibration spectrum data. Vibration spectrum data transmission analysis is then performed to determine if the amount of vibration spectrum data is less than or equal to the preset amount. If so, it is determined whether the real-time network bandwidth occupancy rate is less than the real-time network bandwidth occupancy threshold. If not, the delay in vibration spectrum data transmission is assessed. Specifically, the vibration data transmission delay result is compared with the vibration data transmission threshold. If the comparison result is greater than or equal to the threshold, the vibration spectrum data is dynamically transmitted, and an adaptive threshold determination is performed. If the comparison result is less than the threshold, an adaptive threshold determination is performed, determining if the operating condition adaptation energy ratio benchmark value is within the preset energy ratio threshold range. If it is, it indicates that the current fault characteristics do not exceed normal operating condition fluctuations, and a self-diagnosis qualification prompt for the chemical production machinery and equipment is sent to the designated personnel. If it is not, it indicates that the fault characteristics of the chemical production machinery and equipment are significant, and self-diagnosis management of the chemical equipment fault is implemented.

[0018] It should be added that, prior to the design of the intelligent inspection-based self-diagnosis management system for chemical production machinery and equipment in this application, a database storing various preset data was established. The database includes, but is not limited to, preset vibration signal amplitude ratio, kurtosis fluctuation critical value, amplitude ratio stability critical value, preset vibration data transmission duration, and preset energy ratio threshold range.

[0019] The database's data sources are primarily based on the full lifecycle operation and maintenance data of chemical production machinery and equipment. This includes the equipment's rated operating parameters at the time of manufacture, normal operation test data under different working conditions, vibration characteristics and judgment criteria data corresponding to historical fault repairs, and effective operating condition correlation data collected through long-term inspections. It is further supplemented and calibrated by incorporating industry operation and maintenance standards and fault diagnosis experience values ​​for similar chemical equipment, eliminating invalid data caused by transient interference or human error. The database employs a structured data structure design, establishing data association indexes to precisely bind various set data with corresponding operating conditions and equipment models. For example, it associates benchmark threshold groups corresponding to different speed ranges and energy proportions adapted to different material viscosities. The data is categorized and stored according to threshold ranges to enable rapid data retrieval and access. Data extension fields are reserved to support flexible input of new operating parameters and judgment criteria. The storage method employs a dual-mode architecture combining local storage at edge nodes and cloud backup on the central server. Local storage at edge nodes stores frequently accessed core settings data to meet the need for rapid data retrieval during real-time diagnosis. The central server stores the complete database content in the cloud and updates it regularly. Data encryption algorithms ensure data transmission and storage security, while a data redundancy backup mechanism ensures a stable supply and reliable access to various settings data during the operation of the intelligent inspection-based self-diagnosis management system for chemical production machinery and equipment.

[0020] It should be added that, such as Figure 2 The diagram shown is a structural schematic of a fault self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection provided in an embodiment of the present invention, including: a vibration signal adaptive acquisition module, a vibration spectrum data transmission and analysis module, and an adaptive threshold determination module.

[0021] The vibration signal adaptive acquisition module is used in the self-diagnosis management of chemical production machinery and equipment to adaptively acquire vibration signals of chemical equipment to reflect the accuracy of the acquisition of the equipment's operating status. Based on the vibration signals, corresponding vibration spectrum data is obtained. The module determines whether to perform vibration spectrum data verification to enhance the reliability of fault feature identification. The vibration signal of chemical equipment refers to the time-domain signal that changes over time and is continuously acquired by detection equipment such as high-frequency piezoelectric accelerometers during the operation of chemical production machinery and equipment (such as reactor agitators). The vibration spectrum data refers to the frequency domain spectrum that is converted into a distribution of the vibration signal at different frequencies through mathematical transformations such as fast Fourier transform. It covers data such as the vibration frequency, vibration amplitude, and modulation sideband of the chemical equipment. The adaptive acquisition of vibration signals helps to improve the reliability and effectiveness of vibration spectrum data and enhance the accuracy of fault feature identification.

[0022] The vibration spectrum data transmission analysis module is used to perform vibration spectrum data transmission analysis on the transmitted vibration spectrum data after the adaptive acquisition process is completed. This analysis reflects the timeliness and network adaptability of vibration spectrum data transmission, obtains the corresponding transmission analysis results, and decides whether to implement dynamic transmission of vibration spectrum data based on the transmission analysis results to improve the integrity and timing consistency of vibration spectrum data transmission. Vibration spectrum data transmission analysis helps to improve the stability and anti-interference capability of vibration spectrum data transmission, thereby improving the authenticity and timing consistency of the data received by the central server.

[0023] The adaptive threshold determination module is used to perform adaptive threshold determination on the vibration spectrum data after the vibration spectrum data transmission and analysis process is completed. This determination is used to measure the significance of fault characteristics of chemical equipment. Based on the determination results, it determines whether to carry out self-diagnosis management of chemical equipment faults to assess whether there are early faults in chemical production machinery and equipment. Adaptive threshold determination helps to achieve dynamic matching between threshold standards and personalized working conditions, thereby realizing accurate identification and early warning of early faults in chemical production machinery and equipment.

[0024] Furthermore, the specific process for adaptive acquisition of vibration signals from chemical equipment is as follows: Vibration signals of chemical equipment are collected within a preset time period during the adaptive acquisition period. The ratio of the fourth-order central moment of the vibration signal to the square of the amplitude variance of the vibration signal is expressed as the kurtosis of the chemical equipment vibration signal, which characterizes the degree of impact of non-uniform load.

[0025] The fourth-order central moment of the vibration signal is represented by the average value of the difference between the vibration signal amplitude at each moment monitored by the vibration sensor within a preset time period set by the preset personnel and the average value of the vibration signal amplitude. This value is used to amplify the non-uniform load in the vibration signal of chemical equipment. The ratio of the total amplitude of the modulation sideband to the fundamental amplitude in the vibration signal of chemical equipment is expressed as the amplitude ratio of the vibration signal of chemical equipment, which characterizes the overlap between the fault spectrum and broadband noise.

[0026] The total amplitude of the modulation sideband refers to the amplitude of the modulation sideband of the vibration signal of chemical equipment (such as the modulation sideband generated by abnormal gear meshing in a reactor). The fundamental amplitude refers to the amplitude of the vibration signal of chemical equipment corresponding to the idling of chemical production machinery (such as the rated speed of a reactor). While judging whether the kurtosis of the chemical equipment vibration signal is within the preset kurtosis range, it is also judged whether the amplitude ratio of the chemical equipment vibration signal is less than or equal to the preset amplitude ratio. The preset kurtosis range is represented by the range between the maximum and minimum values ​​of the kurtosis of the chemical equipment vibration signal over a historical period, including both the maximum and minimum values. The preset amplitude ratio is represented by the average value of the amplitude ratio of the chemical equipment vibration signal over a historical period.

[0027] If the kurtosis of the vibration signal of the chemical equipment is within the preset range, and the amplitude ratio of the vibration signal of the chemical equipment is less than or equal to the preset amplitude ratio of the vibration signal of the chemical equipment, it indicates that the chemical machinery and equipment is operating normally. The corresponding vibration spectrum data is marked as level three vibration spectrum data, stored in the edge computing node, and uploaded to the central server at the preset level three frequency (generally once per hour).

[0028] If either of the following two situations exists: In the first scenario, the kurtosis of the chemical equipment vibration signal is within the preset range, and the amplitude ratio of the chemical equipment vibration signal is greater than the preset amplitude ratio. In the second scenario, the kurtosis of the chemical equipment vibration signal is not within the preset range, and the amplitude ratio of the chemical equipment vibration signal is less than or equal to the preset amplitude ratio. This indicates a slight abnormality in the operating status of the chemical production machinery and equipment. The corresponding vibration spectrum data is marked as secondary vibration spectrum data, and a vibration spectrum data verification operation is performed at the edge computing node.

[0029] If the kurtosis of the vibration signal of the chemical equipment is not within the preset range, and the amplitude ratio of the vibration signal is greater than the preset amplitude ratio, it indicates that the operating status of the chemical machinery and equipment is abnormal. The corresponding vibration spectrum data is marked as Level 1 vibration spectrum data and directly transmitted to the central server. A Level 1 warning prompt is also sent to the preset personnel to stop the operation of the chemical machinery and equipment. The urgency of Level 1 vibration spectrum data is greater than that of Level 2 vibration spectrum data, which is greater than that of Level 3 vibration spectrum data.

[0030] In this embodiment, the process can effectively capture potential equipment malfunctions and obvious anomalies, clearly distinguish between three states: normal operation, minor anomalies, and severe anomalies, ensure differentiated handling of operating condition information with different levels of urgency, guarantee efficient archiving and low-frequency transmission of vibration spectrum data, ensure consistency and reliability of self-diagnosis management of chemical production machinery and equipment faults, strengthen real-time control of the operating status of chemical production machinery and equipment, and ensure stable operation of chemical production machinery and equipment.

[0031] Furthermore, the specific process for verifying the vibration spectrum data is as follows: The vibration signal of the chemical equipment corresponding to the secondary vibration spectrum data is divided into a preset number of segments (generally 3 to 20 segments) set in advance by pre-set personnel based on historical experience. Randomly select segments of the chemical equipment vibration signal (generally from the preset number of segments). ); Obtain the kurtosis and amplitude ratio of the vibration signals of different chemical equipment segments; Use the ratio of the standard deviation of the obtained vibration signal kurtosis to the average value of the vibration signal kurtosis as the vibration kurtosis fluctuation value to reflect the degree of kurtosis fluctuation of the vibration signal of the chemical equipment.

[0032] The acquired amplitude ratios of the chemical equipment vibration signals are sorted according to their numerical values. The ratio of the difference between the maximum and minimum amplitude ratios of the chemical equipment vibration signals to the average amplitude ratio is used as the stability value of the amplitude ratio, reflecting the stability of the amplitude ratio. The maximum amplitude ratio refers to the maximum amplitude ratio of the chemical equipment vibration signals acquired during the vibration spectrum data verification operation. The minimum amplitude ratio refers to the minimum amplitude ratio of the chemical equipment vibration signals acquired during the vibration spectrum data verification operation. The average amplitude ratio of the chemical equipment vibration signals is the average value obtained by summing all randomly selected amplitude ratios of the chemical equipment vibration signals at preset times during the vibration spectrum data verification operation and dividing it by the total number of sampled amplitude ratios.

[0033] If either of the following two results occurs: The first result is that while the vibration kurtosis fluctuation value exceeds the kurtosis fluctuation critical value, the vibration amplitude ratio stability value does not exceed the amplitude ratio stability critical value.

[0034] The second result is that while the vibration amplitude ratio stability value exceeds the amplitude ratio stability threshold, the vibration kurtosis fluctuation value does not exceed the kurtosis fluctuation threshold. This indicates that the vibration signal of the chemical equipment acquired by the adaptive acquisition has random fluctuations, and wavelet packet decomposition processing of the vibration spectrum signal needs to be performed. The kurtosis fluctuation threshold and the amplitude ratio stability threshold are both set in advance by the preset personnel based on historical experience.

[0035] If the vibration kurtosis fluctuation value does not exceed the kurtosis fluctuation threshold and the vibration amplitude ratio stability value does not exceed the amplitude ratio stability threshold, it indicates that the vibration spectrum data is stable and can be directly entered into vibration spectrum data transmission and analysis. If the vibration kurtosis fluctuation value exceeds the kurtosis fluctuation threshold and the vibration amplitude ratio stability value exceeds the amplitude ratio stability threshold, it indicates that the chemical machinery equipment is in an abnormal operating state. The corresponding vibration spectrum data is marked as Level 1 vibration spectrum data, directly transmitted to the central server, and a Level 1 warning prompt is sent to the preset personnel to stop the operation of the chemical machinery equipment.

[0036] It should be added that, such as Figure 3The diagram shown illustrates the vibration spectrum data verification operation architecture of the intelligent inspection-based fault self-diagnosis management system for chemical production machinery provided in this embodiment of the invention. During vibration spectrum data verification, the system determines whether the vibration kurtosis fluctuation value exceeds the kurtosis fluctuation threshold and whether the vibration amplitude ratio stability value exceeds the amplitude ratio stability threshold. If either of the following two results applies: First, the vibration kurtosis fluctuation value exceeds the kurtosis fluctuation threshold, but the vibration amplitude ratio stability value does not exceed the amplitude ratio stability threshold; second, the vibration amplitude ratio stability value exceeds the amplitude ratio stability threshold, but the vibration kurtosis fluctuation value does not exceed the kurtosis fluctuation threshold, then wavelet packet decomposition processing of the vibration spectrum signal is performed to determine whether the vibration signal frequency ratio is within the vibration frequency band reference range. If not, the corresponding vibration frequency band is identified as a fault-sensitive frequency band; if it is, the corresponding vibration frequency band is identified as a fault-sensitive frequency band. The vibration frequency band is determined to be a normal operating frequency band. The system checks if the number of fault-sensitive frequency bands is greater than or equal to the number of normal operating frequency bands. If so, the corresponding chemical equipment vibration signal is marked as a fault characteristic signal; otherwise, it is marked as a normal operating characteristic signal, and the corresponding vibration spectrum data is directly marked as secondary vibration spectrum data for vibration spectrum data transmission and analysis. If the vibration kurtosis fluctuation value does not exceed the kurtosis fluctuation threshold and the vibration amplitude ratio stability value does not exceed the amplitude ratio stability threshold, vibration spectrum data transmission and analysis are performed. If the vibration kurtosis fluctuation value exceeds the kurtosis fluctuation threshold and the vibration amplitude ratio stability value exceeds the amplitude ratio stability threshold, the corresponding vibration spectrum data is marked as primary vibration spectrum data, directly transmitted to the central server, and a primary warning is sent to designated personnel to stop the operation of the chemical machinery equipment.

[0037] In this embodiment, the process is used to ensure the reliability of fault warning and status determination, accurately quantify the degree of vibration signal kurtosis fluctuation and amplitude ratio stability, ensure the pertinence of fault feature extraction, avoid missing potential faults due to accidental interference, strengthen the rapid control of fault risks of chemical production machinery and equipment, and ensure the safe and stable operation of chemical production machinery and equipment.

[0038] Furthermore, the specific process of wavelet packet decomposition processing of vibration spectrum signals is as follows: Based on the preset frequency range of chemical equipment vibration signals divided in advance by preset personnel, the vibration signals of chemical equipment are decomposed into vibration frequency bands of a preset number of signal divisions; the frequency of each vibration frequency band is obtained and monitored by vibration detection sensors; the ratio of each vibration frequency band to the frequency of the chemical equipment vibration signal is expressed as the vibration signal frequency ratio used to quantify the degree of influence of the vibration frequency band on the vibration signals of chemical equipment, wherein the preset number of signal divisions is set in advance by preset personnel.

[0039] Determine whether the vibration signal frequency ratio is within the vibration frequency band reference range; the vibration frequency band reference range is represented by the range between the maximum and minimum values ​​of the vibration frequency ratio of chemical machinery equipment during normal operation obtained from historical time periods; if the vibration signal frequency ratio is not within the vibration frequency band reference range, the corresponding vibration frequency band is determined to be a fault-sensitive frequency band.

[0040] If the frequency ratio of the vibration signal is within the reference range of the vibration frequency band, the corresponding vibration frequency band is determined as the normal operation frequency band. The number of fault-sensitive frequency bands and the number of normal operation frequency bands are counted. If the number of fault-sensitive frequency bands is greater than or equal to the number of normal operation frequency bands, it indicates that the current chemical equipment vibration signal contains fault characteristics, and the corresponding chemical equipment vibration signal is marked as a fault characteristic signal for fault characteristic signal reconstruction. The fault characteristic signal refers to the chemical equipment vibration signal that can accurately reflect the equipment fault state during the operation of chemical production machinery and equipment (such as reactor agitators). It is a time-domain signal containing the core characteristics of equipment faults after removing redundant noise. If the number of fault-sensitive frequency bands is less than the number of normal operation frequency bands, it indicates that the chemical equipment vibration signal is mainly composed of normal operation vibration and conventional noise, with no obvious fault characteristics. The corresponding chemical equipment vibration signal is marked as a normal operation characteristic signal, and the corresponding vibration spectrum data is directly marked as secondary vibration spectrum data for vibration spectrum data transmission and analysis.

[0041] In this embodiment, the process realizes multi-band splitting and feature classification of vibration signals, enhances the accuracy of distinguishing between fault-sensitive frequency bands and normal operating frequency bands, effectively removes the interference of conventional noise and redundant signals on fault diagnosis, ensures the accuracy and reliability of fault-sensitive frequency band determination, and improves the identification and purity of fault characteristics of chemical production machinery and equipment.

[0042] Furthermore, the specific process of reconstructing the fault characteristic signals is as follows: First, the amplitude of the time-domain signal in the fault-sensitive frequency band is filtered by amplitude threshold. Time-domain signals with amplitudes greater than or equal to the amplitude threshold are retained to reflect fault characteristics, while time-domain signals with amplitudes less than the amplitude threshold are discarded. The retained time-domain signals of the fault-sensitive frequency band are sorted from largest to smallest according to the frequency of each vibration band and then superimposed to form the time-domain signal of the fault-sensitive frequency band that reflects the fault characteristics of the complete chemical equipment. A fast Fourier transform is then performed to convert the time-domain signal of the fault-sensitive frequency band into vibration spectrum data that reflects the fault characteristics, and it is updated to secondary vibration spectrum data for vibration spectrum data transmission and analysis.

[0043] The specific process of converting the time-domain signal of the fault-sensitive frequency band into vibration spectrum data reflecting fault characteristics using Fast Fourier Transform (FFT) is as follows: The time-domain signal of the fault-sensitive frequency band is substituted into the FFT algorithm for calculation, converting the vibration displacement, velocity, or acceleration signal in the time domain into complex-form spectrum data in the frequency domain, obtaining the real and imaginary part values ​​corresponding to each frequency point; Based on the complex spectrum data obtained from the FFT, the vibration amplitude corresponding to each frequency point is solved through modulus calculation, forming a frequency domain data set with frequency as the abscissa and amplitude as the ordinate; Amplitude correction is performed on the transformed spectrum amplitude to correct the amplitude deviation caused by windowing processing and algorithm calculation; From the corrected frequency domain data set, all spectrum data within the fault-sensitive frequency band is extracted, and invalid data outside the frequency band is removed, finally obtaining vibration spectrum data that can be directly used for fault characteristic identification and analysis of chemical equipment.

[0044] In this embodiment, the process is used to purify and optimize the time-domain signal of the fault-sensitive frequency band, reduce the interference of redundant noise and invalid signals on the fault diagnosis of chemical production machinery and equipment, enhance the aggregation and integrity of the core fault features, and integrate the scattered fault-sensitive signals into unified and clear vibration spectrum data through signal filtering, sorting and superposition and frequency domain conversion, thereby improving the identifiability of fault features.

[0045] Furthermore, the specific process of vibration spectrum data transmission and analysis is as follows: The vibration spectrum data volume is obtained by an industrial data acquisition instrument, and the real-time network bandwidth utilization rate is obtained by a network bandwidth analyzer. The vibration spectrum data volume refers to the total amount of vibration spectrum data transmitted to the central server. The real-time network bandwidth utilization rate is represented by the ratio of the network bandwidth occupied by vibration spectrum data transmission at each moment during the vibration spectrum data transmission analysis period to the total rated available network bandwidth of the transmission link. It is used to reflect the degree of network link occupancy of vibration spectrum data transmission in real time.

[0046] The system determines whether the amount of vibration spectrum data is less than or equal to a preset amount, where the preset amount is represented by the maximum data capacity of the network bandwidth. If the amount of vibration spectrum data is less than or equal to the preset amount, it further determines whether the real-time network bandwidth occupancy rate is less than a pre-set network bandwidth occupancy threshold. If not, a network bandwidth anomaly alert is sent to the pre-set personnel. If so, based on the urgency of the vibration spectrum data, it transmits the vibration spectrum data to the central server at the corresponding preset transmission frequency and performs an adaptive threshold determination. Specifically: If the vibration spectrum data is a Level 1 vibration spectrum data (marked as abnormal operating status of chemical production machinery and equipment), it is transmitted to the central server in real time at the preset Level 1 transmission frequency (highest priority transmission frequency), and an adaptive threshold judgment is triggered immediately after the transmission is completed.

[0047] If the vibration spectrum data is a level 2 vibration spectrum data (marked as a slight abnormality in the operating status of chemical production machinery and equipment), it is transmitted to the central server at the edge computing node according to the preset level 2 transmission frequency (medium priority transmission frequency).

[0048] If the vibration spectrum data is level three vibration spectrum data (marked as normal equipment operation status), it is uploaded from the edge computing node to the central server at the preset level three transmission frequency, where the preset level one transmission frequency is greater than the preset level two transmission frequency, which is greater than the preset level three transmission frequency.

[0049] If the amount of vibration spectrum data exceeds the preset amount of vibration spectrum data, the delay in vibration spectrum data transmission will be assessed to evaluate the timeliness of data transmission. Specifically, this means: The time interval between the start time of the edge computing node sending low-level vibration spectrum data and the end time of the central server successfully receiving the batch of data is represented as the vibration data transmission duration.

[0050] Low-level vibration spectrum data includes: Level 2 vibration spectrum data and Level 3 vibration spectrum data; the ratio of the vibration data transmission duration to the preset vibration data transmission duration is used to characterize the degree of vibration spectrum data delay, and the preset vibration data transmission duration is represented by the average vibration data transmission duration over a historical period; the vibration data transmission delay result is compared with the vibration data transmission threshold to obtain the corresponding comparison result; if the comparison result is that the vibration data transmission delay result is greater than or equal to the vibration data transmission threshold, then dynamic transmission of vibration spectrum data is implemented to reduce network bandwidth usage and avoid vibration spectrum data transmission delay and distortion, and adaptive threshold determination is performed after the dynamic transmission of vibration spectrum data is completed; if the comparison result is that the vibration data transmission delay result is less than the vibration data transmission threshold, then adaptive threshold determination is performed; dynamic transmission of vibration spectrum data refers to the transmission operation of the vibration spectrum data corresponding to the next vibration spectrum data transmission analysis.

[0051] In this embodiment, the process helps ensure that vibration spectrum data is transmitted in an orderly manner according to operating conditions, maintaining the timeliness, integrity, and reliability of data transmission. By accurately matching data volume, network status, and data urgency, differentiated transmission strategies are implemented, thereby optimizing the efficient allocation of network resources and avoiding network resource waste and transmission bottlenecks.

[0052] Furthermore, the specific process of dynamic transmission of vibration spectrum data is as follows: The amount of low-level vibration spectrum data is divided into a preset number of data fragments set in advance by pre-defined personnel, and the network bandwidth occupancy fluctuation rate is monitored in real time. Among them, the data fragments corresponding to the second-level vibration spectrum data have a higher priority than the data fragments corresponding to the third-level vibration spectrum data. The amount of low-level vibration spectrum data refers to the total amount of second-level and third-level vibration spectrum data. The network bandwidth occupancy fluctuation rate refers to the ratio of the difference between the maximum and minimum actual network bandwidth occupancy during the preset monitoring period to the average network bandwidth occupancy during the preset monitoring period, which is used to quantify the stability of network bandwidth occupancy. The average network bandwidth occupancy refers to the average value of the actual bandwidth occupied by the network during the dynamic transmission of vibration spectrum data, which is used to quantify the overall network bandwidth occupancy level during the monitoring period.

[0053] If the network bandwidth usage fluctuation rate is greater than the preset network bandwidth usage fluctuation rate, the data fragmentation transmission rate is adjusted; otherwise, data fragmentation continues to be transmitted to the central server. Adjusting the data fragmentation transmission rate is used to reduce the degree of network bandwidth usage fluctuation. The specific process is as follows: The data fragment transmission rate is monitored in real time by an industrial network monitoring instrument. If the data fragment transmission rate is greater than the first transmission rate standard value obtained by the preset personnel based on past experience, the data fragment transmission rate is reduced to the first transmission rate standard value, and the transmission of the third-level vibration spectrum data fragment is suspended. If the data fragment transmission rate is lower than the first transmission rate standard value, the backup network link is used to transmit the second-level vibration spectrum data fragment. The bandwidth usage fluctuation rate is checked again to see if it is still greater than the preset network bandwidth usage fluctuation rate. If it is, a bandwidth usage abnormality prompt is sent to the preset personnel; otherwise, an adaptive threshold judgment is performed.

[0054] In this embodiment, the process helps ensure the stable and orderly transmission of vibration spectrum data under network status fluctuation scenarios, maintaining the temporal consistency and integrity of data transmission. By data fragmentation and priority sorting, high-urgency fault-related data fragments are transmitted first, effectively reducing the fluctuation of network bandwidth usage, alleviating network load pressure, and avoiding data transmission delays, distortions, or interruptions.

[0055] Furthermore, the specific process of adaptive threshold determination is as follows: The ratio of the frequency band energy corresponding to the fault feature to the total energy of the vibration signal is expressed as the fault feature frequency band energy ratio, used to quantify the fault feature identification capability; the ratio of real-time material viscosity to commonly used material viscosity is expressed as the viscosity condition adaptation coefficient, characterizing the degree of difference between the current operating condition and the standard operating condition; real-time material viscosity is monitored in real time by a vibratory online viscometer; commonly used material viscosity is expressed as the viscosity of homogeneous materials at standard temperature and pressure over a historical period, without special external interference; the viscosity condition adaptation coefficient and other ranges are divided into a preset number of continuous intervals, and for each interval, the operating condition adaptation energy ratio benchmark value of the chemical production machinery and equipment during operation is obtained as the corresponding value for that interval. The operating condition adaptability energy ratio benchmark value; if the operating condition adaptability energy ratio benchmark value is within the preset energy ratio threshold range, then the current fault characteristics do not exceed the normal operating condition fluctuations, and a self-diagnosis qualified prompt for chemical production machinery and equipment is sent to the preset personnel. Otherwise, it indicates that the fault characteristics of chemical production machinery and equipment are significant, and chemical equipment fault self-diagnosis management is carried out. This process can quickly locate the fault type, fault location and fault development degree, and at the same time, it links the parameters of equipment operation to carry out comprehensive analysis, output targeted fault warning information and handling, and realize accurate identification and rapid response to abnormal equipment faults. Among them, the power spectral density corresponding to the frequency point of the fault characteristic is integrated to obtain the frequency band energy corresponding to the fault characteristic.

[0056] It should be added that, such as Figure 4 The diagram shows the adaptive threshold determination architecture of the chemical production machinery equipment fault self-diagnosis management system based on intelligent inspection provided in this embodiment of the invention. Adaptive threshold determination is performed to determine if the energy ratio benchmark value for the operating condition is within the preset energy ratio threshold range. If it is, it indicates that the current fault characteristics do not exceed normal operating condition fluctuations, and chemical equipment fault self-diagnosis management is initiated. If it is not, it indicates that the chemical production machinery equipment fault characteristics are significant, and chemical equipment fault self-diagnosis management is initiated. If the feature frequency similarity is greater than or equal to the first-level fault similarity threshold, a first-level warning is immediately triggered, prompting preset personnel to stop the operation of the chemical machinery equipment. If the feature frequency similarity is greater than or equal to the second-level fault similarity threshold but less than the first-level fault similarity threshold, a second-level warning is triggered, prompting preset maintenance personnel to push chemical machinery equipment fault warning information. If the similarity is less than the second-level fault similarity threshold, it is determined whether the similarity change rate is greater than the similarity change threshold. If not, a chemical production machinery equipment self-diagnosis qualified prompt is sent to preset personnel. If yes, a first-level warning is immediately triggered, prompting preset personnel to stop the operation of the chemical machinery equipment.

[0057] In this embodiment, the process is used to achieve precise binding between fault feature identification and real-time material viscosity conditions, eliminate the interference of operating condition differences on fault determination, ensure that the self-diagnosis of faults in chemical production machinery and equipment is consistent with real-time operating conditions, and improve the accuracy and reliability of fault identification.

[0058] Furthermore, the specific process of self-diagnosis management of chemical equipment faults is as follows: real-time vibration spectrum is obtained based on vibration spectrum data.

[0059] Specifically, the vibration spectrum data is amplified, filtered, and impedance matched by a signal conditioning circuit. Then, an analog-to-digital converter transforms the continuous analog signal into a discrete digital signal. A real-time Fourier transform is then performed on this discrete digital signal to convert it from the time domain to the frequency domain, ultimately outputting a segmentless real-time vibration spectrum. This results in the acquisition of the real-time vibration spectrum and the standard spectrum for each fault type in the dynamic fault feature library. The dynamic fault feature library is a structured, scalable, and real-time updated digital information set for chemical production machinery and equipment, with the reactor stirring system as the core research object. It integrates typical fault types, fault mechanisms, fault evolution laws, and standard vibration spectrum data under corresponding fault states. The number of fault feature frequency bands that are identical between the real-time vibration spectrum and the standard spectrum for each fault type in the dynamic fault feature library is counted. The ratio of this number of fault feature frequency bands to the total number of reference fault feature frequency bands is used as the characteristic frequency similarity characterizing the consistency of their frequency distributions.

[0060] Set feature frequency similarity thresholds: a first-level fault similarity threshold and a second-level fault similarity threshold, wherein the first-level fault similarity threshold is higher than the second-level fault similarity threshold.

[0061] Trigger corresponding level warnings based on feature frequency similarity thresholds: If the similarity of the characteristic frequency is greater than or equal to the first-level fault similarity threshold, it is judged as a serious fault, and a first-level warning is immediately triggered, prompting the preset personnel to stop the operation of the chemical machinery and equipment.

[0062] If the feature frequency similarity is greater than or equal to the secondary fault similarity threshold but less than the primary fault similarity threshold, it is determined to be an early fault, triggering a secondary warning and prompting the preset maintenance personnel to push fault warning information for chemical machinery equipment. This includes: the judgment result (e.g., an early fault exists in the stirring blades of the reactor), a preliminary indication of the fault type triggering the warning (e.g., early wear of reactor bearings, slight gear meshing defects, etc.), and the real-time operating status of the equipment at the time of the fault, etc.; if the feature frequency similarity is less than the secondary fault similarity threshold, the feature frequency similarity of the preset similarity division is obtained. Both the primary and secondary fault similarity thresholds are preset by the preset personnel. During the self-diagnosis management of chemical equipment faults, the ratio of the difference between randomly selected feature frequency similarities and the average feature frequency similarities to the average feature frequency similarities is used as the similarity change rate to reflect the changing trend of chemical equipment fault characteristics. The preset number of similarity divisions is also preset by the preset personnel.

[0063] The average value of the feature frequency similarity refers to the arithmetic mean of the feature frequency similarity within the self-diagnosis management of chemical equipment faults. If the similarity change rate is greater than the similarity change threshold, it indicates that the fault characteristics of the chemical production machinery and equipment are continuously enhanced, triggering a first-level warning prompt to the preset personnel to stop the operation of the chemical machinery and equipment. Conversely, it indicates that the chemical production machinery and equipment is operating normally, and a self-diagnosis qualification prompt for the chemical production machinery and equipment is sent to the preset personnel.

[0064] In this embodiment, the process is used to ensure the accuracy of fault identification, ensure that different levels of faults are responded to and dealt with in a targeted manner, accurately capture the potential changing trends of fault characteristics of chemical production machinery and equipment, realize early prediction and early warning of early weak faults and fault initiation states, and avoid the escalation of faults.

[0065] Example 2: Building upon Example 1, in chemical production, the rotational speed of the reactor agitator may fluctuate dynamically due to factors such as changes in material viscosity and motor load adjustments. These speed variations directly alter the frequency characteristics of the vibration signal, leading to deviations in judgment methods relying solely on fixed thresholds (for example, vibration characteristics normal at low speeds may be misjudged as abnormal at high speeds, and vice versa). Therefore, Example 2 employs speed-related correction to achieve dynamic calibration of vibration signal characteristic parameters, thereby improving the accuracy and stability of monitoring the operating status of chemical equipment. The specific process is as follows: The operating speed of chemical production machinery and equipment (such as reactor agitators) is continuously monitored by a speed sensor. The average speed within a preset collection period is used as the real-time speed parameter of the current operating condition of the chemical production machinery and equipment. The ratio of the average speed of the chemical production machinery and equipment to the rated speed of the chemical production machinery and equipment is defined as the speed difference coefficient, which is used to quantify the degree of deviation between the current speed condition and the standard speed condition.

[0066] When the speed difference coefficient is greater than 1, it indicates that the current speed of the chemical production machinery is higher than the rated speed of the chemical production machinery. The interval corresponding to the current speed of the chemical production machinery is marked as the high rated speed interval.

[0067] When the speed difference coefficient is less than 1, it indicates that the current speed of the chemical production machinery is lower than the rated speed of the chemical production machinery. The range corresponding to the current speed of the chemical production machinery is marked as the low rated speed range.

[0068] When the speed difference coefficient is equal to 1, it means that the current speed of the chemical production machinery and equipment is consistent with the rated speed of the chemical production machinery and equipment. The interval corresponding to the current speed of the chemical production machinery and equipment is marked as the equal speed interval.

[0069] Key performance indicator data for chemical production machinery and equipment were statistically analyzed across three consecutive speed ranges (low rated speed range, equal rated speed range, and high rated speed range), including the baseline thresholds for the energy proportion of fault characteristic frequency bands, vibration signal kurtosis, and vibration signal amplitude ratio. These key performance indicator data were then used as baseline threshold sets for each type of speed operating condition to construct a dynamically updatable speed-related baseline threshold library. Specifically: The baseline threshold groups corresponding to the three consecutive intervals are stored in the database with a mapping relationship between the speed interval and the baseline threshold group. The unique identifier of the speed interval is obtained, which includes: interval identifier ID, key indicator data, calibration time and data source. Based on the preset table structure requirements, the unique identifier of the speed interval is bound to its corresponding baseline threshold group one by one to obtain the corresponding mapping pair. Then, the mapping pair is written to the preset database table one by one using the database write interface. During the writing process, the data rationality is checked simultaneously to verify whether the value of the baseline threshold group meets the normal range of the equipment operation indicators. If it meets the requirements, the writing continues. If it does not meet the requirements, an alarm is triggered and the writing is paused until manual review and correction before continuing the writing.

[0070] A dual mechanism of regular and triggered updates is implemented to ensure that the threshold library adapts to long-term needs such as equipment aging and changes in operating conditions. Specifically: The system automatically collects key indicator data for each speed range of the equipment within a preset cycle (usually 3 months), recalculates the baseline thresholds of the key indicators, and compares them with the baseline thresholds of the original key indicator data in the threshold library. If the difference between the baseline threshold of the key indicator and the baseline threshold of the original key indicator data is within a preset fluctuation range (e.g., 1±5%), the corresponding baseline threshold group is updated, and self-diagnosis management of chemical equipment faults is carried out. Otherwise, an adaptive threshold judgment anomaly prompt is sent to the preset personnel.

[0071] In summary, in this embodiment, by adaptively acquiring vibration signals from chemical equipment and performing vibration spectrum data verification, it helps to measure the stability of the operating status of chemical equipment and the degree of potential fault risks, improves the reliability and effectiveness of vibration spectrum data, enhances the accuracy of fault feature identification, and strengthens the anti-interference ability against interference factors such as non-uniform loads and broadband noise. Vibration spectrum data transmission analysis helps to assess network bandwidth adaptability and the timeliness and integrity of data transmission, improving the stability and anti-interference ability of vibration spectrum data transmission, thereby improving the authenticity and timing consistency of data received by the central server. Adaptive threshold determination of vibration spectrum data, and based on the obtained determination results, determining whether to implement self-diagnosis management of chemical equipment faults, can effectively assess the significance of chemical equipment fault characteristics and their adaptability to real-time operating conditions, improve the accuracy of chemical equipment fault identification and the timeliness of early warning, and achieve dynamic matching of threshold standards and personalized operating conditions, thereby realizing accurate identification and early warning of early faults in chemical production machinery and equipment.

[0072] The above-disclosed embodiments are merely some examples of the present invention and should not be construed as limiting the scope of the present invention. Therefore, any equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection, characterized in that, The system includes: an adaptive vibration signal acquisition module, a vibration spectrum data transmission and analysis module, and an adaptive threshold determination module; The vibration signal adaptive acquisition module is used to adaptively acquire vibration signals of chemical equipment, obtain corresponding vibration spectrum data based on the vibration signals of chemical equipment, and determine whether to perform vibration spectrum data verification operation based on the vibration spectrum data. The vibration spectrum data transmission and analysis module is used to perform vibration spectrum data transmission and analysis on the transmitted vibration spectrum data after the adaptive acquisition process is completed, obtain the corresponding transmission analysis results, and decide whether to implement dynamic transmission of vibration spectrum data based on the transmission analysis results. The adaptive threshold determination module is used to perform adaptive threshold determination on the vibration spectrum data after the vibration spectrum data transmission and analysis process is completed, and to determine whether to carry out chemical equipment fault self-diagnosis management based on the determination result.

2. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 1, characterized in that, The specific process for adaptive acquisition of vibration signals from chemical equipment is as follows: Vibration signals of chemical equipment were collected within a preset time period during the adaptive acquisition period. The ratio of the fourth-order central moment of the vibration signal to the square of the amplitude variance of the vibration signal is expressed as the kurtosis of the vibration signal of chemical equipment. The fourth-order central moment of the vibration signal is represented by the average value of the difference between the vibration signal amplitude and the average vibration signal amplitude at each moment within a preset time period, calculated by power of the fourth power. The ratio of the total amplitude of the modulation sideband to the fundamental amplitude in the vibration signal of chemical equipment is expressed as the amplitude ratio of the vibration signal of chemical equipment. While determining whether the kurtosis of the vibration signal of the chemical equipment is within the preset range of the kurtosis of the vibration signal of the chemical equipment, it is also determined whether the amplitude ratio of the vibration signal of the chemical equipment is less than or equal to the preset amplitude ratio of the vibration signal of the chemical equipment. If the kurtosis of the vibration signal of the chemical equipment is within the preset range of the kurtosis of the vibration signal of the chemical equipment, and the amplitude ratio of the vibration signal of the chemical equipment is less than or equal to the preset amplitude ratio of the vibration signal of the chemical equipment, then the corresponding vibration spectrum data is marked as third-level vibration spectrum data, stored in the edge computing node, and uploaded to the central server at the preset third-level frequency. If either of the following two situations exists: In the first case, the kurtosis of the chemical equipment vibration signal is within the preset range of the chemical equipment vibration signal kurtosis, and the amplitude ratio of the chemical equipment vibration signal is greater than the preset amplitude ratio of the chemical equipment vibration signal. In the second case, the kurtosis of the chemical equipment vibration signal is not within the preset range of the chemical equipment vibration signal kurtosis, and the amplitude ratio of the chemical equipment vibration signal is less than or equal to the preset amplitude ratio of the chemical equipment vibration signal. The corresponding vibration spectrum data is marked as secondary vibration spectrum data, and a vibration spectrum data verification operation is performed on the edge computing node. If the kurtosis of the vibration signal of the chemical equipment is not within the preset range, and the amplitude ratio of the vibration signal of the chemical equipment is greater than the preset amplitude ratio of the vibration signal of the chemical equipment, then the corresponding vibration spectrum data will be marked as first-level vibration spectrum data, directly transmitted to the central server, and a first-level warning will be sent to the preset personnel to stop the operation of the chemical machinery and equipment.

3. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 2, characterized in that, The specific process of the vibration spectrum data verification operation is as follows: The vibration signal of the chemical equipment corresponding to the secondary vibration spectrum data is divided into a preset number of chemical equipment vibration segment signals, and the chemical equipment vibration segment signals are randomly selected. Obtain the kurtosis and amplitude ratio of the vibration signals of different chemical equipment segments; The ratio of the standard deviation of the kurtosis of the acquired chemical equipment vibration signal to the average kurtosis of the chemical equipment vibration signal is taken as the vibration kurtosis fluctuation value. The amplitude ratios of the acquired vibration signals from the chemical equipment are sorted according to their numerical values. The ratio of the difference between the maximum and minimum amplitude ratios of the vibration signals of chemical equipment to the average amplitude ratio of the vibration signals of chemical equipment is taken as the stable value of the vibration amplitude ratio. The maximum amplitude ratio of the vibration signal of the chemical equipment refers to the maximum amplitude ratio of the vibration signal of the chemical equipment obtained in the vibration spectrum data verification operation. The minimum amplitude ratio of the vibration signal of the chemical equipment refers to the minimum amplitude ratio of the vibration signal of the chemical equipment obtained in the vibration spectrum data verification operation. If either of the following two results occurs: The first result is that while the vibration kurtosis fluctuation value exceeds the critical value of kurtosis fluctuation, the vibration amplitude ratio stability value does not exceed the critical value of amplitude ratio stability. The second result is that while the amplitude ratio to the stable value exceeds the critical value for the amplitude ratio to the stable value, the kurtosis fluctuation value does not exceed the critical value for the kurtosis fluctuation. Wavelet packet decomposition of the vibration spectrum signal needs to be performed; If the vibration kurtosis fluctuation value does not exceed the kurtosis fluctuation critical value and the vibration amplitude ratio stability value does not exceed the amplitude ratio stability critical value, then proceed directly to vibration spectrum data transmission analysis. If the vibration kurtosis fluctuation value exceeds the kurtosis fluctuation threshold, or the vibration amplitude ratio to stability value exceeds the amplitude ratio to stability threshold, the corresponding vibration spectrum data will be marked as Level 1 vibration spectrum data, directly transmitted to the central server, and a Level 1 warning will be sent to the designated personnel to stop the operation of the chemical machinery equipment.

4. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 3, characterized in that, The specific process of wavelet packet decomposition of the vibration spectrum signal is as follows: Based on the frequency range of vibration signals from chemical equipment, the vibration signals of chemical equipment are decomposed into a preset number of vibration frequency bands. Obtain the frequency of each vibration band; The ratio of each vibration frequency band to the vibration signal frequency of chemical equipment is expressed as the vibration signal frequency ratio. Determine whether the frequency ratio of the vibration signal is within the reference range of the vibration frequency band; If the vibration signal frequency ratio is not within the vibration frequency band reference range, the corresponding vibration frequency band will be identified as a fault-sensitive frequency band. If the vibration signal frequency ratio is within the vibration frequency band reference range, then the corresponding vibration frequency band is determined to be the normal operating frequency band. Count the number of fault-sensitive frequency bands and the number of normally operating frequency bands; If the number of fault-sensitive frequency bands is greater than or equal to the number of normal operating frequency bands, the corresponding chemical equipment vibration signal is marked as a fault characteristic signal and the fault characteristic signal is reconstructed. If the number of fault-sensitive frequency bands is less than the number of normal operating frequency bands, the corresponding chemical equipment vibration signal will be marked as a normal operating characteristic signal, and the corresponding vibration spectrum data will be directly marked as secondary vibration spectrum data, and vibration spectrum data transmission and analysis will be performed.

5. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 4, characterized in that, The specific process for reconstructing the fault characteristic signal is as follows: First, the amplitude of the time-domain signal in the fault-sensitive frequency band is filtered by amplitude threshold; The time-domain signals with amplitudes greater than or equal to the amplitude threshold in the fault-sensitive frequency band are retained, while the time-domain signals with amplitudes less than the amplitude threshold in the fault-sensitive frequency band are discarded. The time-domain signals of the retained fault-sensitive frequency bands are sorted and superimposed according to the frequency of each vibration frequency band from largest to smallest to form the time-domain signal of the fault-sensitive frequency band. Then, a fast Fourier transform is performed to convert the time-domain signal of the fault-sensitive frequency band into vibration spectrum data, which is then updated to secondary vibration spectrum data for vibration spectrum data transmission and analysis.

6. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 5, characterized in that, The specific process of vibration spectrum data transmission and analysis is as follows: Acquire vibration spectrum data volume and real-time network bandwidth utilization; Determine whether the amount of vibration spectrum data is less than or equal to the preset amount of vibration spectrum data. If the amount of vibration spectrum data is less than or equal to the preset amount of vibration spectrum data, then it is further determined whether the real-time network bandwidth occupancy rate is less than the real-time network bandwidth occupancy threshold. If not, a network bandwidth error message will be sent to the designated personnel. If so, based on the urgency of the vibration spectrum data, the vibration spectrum data is transmitted to the central server at the corresponding preset transmission frequency, and an adaptive threshold determination is performed. Specifically: If the vibration spectrum data is a first-level vibration spectrum data, it is transmitted to the central server in real time at the preset first-level transmission frequency, and an adaptive threshold determination is triggered immediately after the transmission is completed. If the vibration spectrum data is secondary vibration spectrum data, it is transmitted from the edge computing node to the central server at the preset secondary transmission frequency. If the vibration spectrum data is a level three vibration spectrum data, it is uploaded from the edge computing node to the central server according to the preset level three transmission frequency; If not, a network bandwidth error message will be sent to the designated personnel. If the amount of vibration spectrum data exceeds the preset amount of vibration spectrum data, then the delay in vibration spectrum data transmission is determined, specifically: The time interval between the start time of the edge computing node sending low-level vibration spectrum data and the end time of the central server successfully receiving the batch of data is expressed as the vibration data transmission duration. The low-level vibration spectrum data includes: secondary vibration spectrum data and tertiary vibration spectrum data; The ratio of the vibration data transmission duration to the preset vibration data transmission duration is expressed as the vibration data transmission delay result. The vibration data transmission delay result is compared with the vibration data transmission threshold to obtain the corresponding comparison result; If the comparison result shows that the vibration data transmission delay is greater than or equal to the vibration data transmission threshold, then dynamic transmission of vibration spectrum data is implemented, and adaptive threshold determination is performed after the dynamic transmission of vibration spectrum data is completed. If the comparison result shows that the vibration data transmission delay is less than the vibration data transmission threshold, then an adaptive threshold determination is performed.

7. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 6, characterized in that, The specific process of dynamically transmitting the vibration spectrum data is as follows: The low-level vibration spectrum data is divided into a preset number of data segments, and the network bandwidth usage fluctuation rate is monitored in real time. The amount of low-level vibration spectrum data refers to the total amount of data from the second-level and third-level vibration spectrum data. If the network bandwidth usage fluctuation rate is greater than the preset network bandwidth usage fluctuation rate, the data fragmentation transmission rate will be adjusted; otherwise, the data fragmentation transmission to the central server will continue. The specific process for adjusting the data fragmentation transmission rate is as follows: Real-time monitoring of data fragmentation transmission rate; If the data fragment transmission rate is greater than the first transmission rate standard value, then reduce the data fragment transmission rate to the first transmission rate standard value, and at the same time suspend the transmission of the third-level vibration spectrum data fragments. If the data fragment transmission rate is lower than the first transmission rate standard value, the backup network link is activated to transmit the secondary vibration spectrum data fragments. Reassess whether the bandwidth usage volatility is still greater than the preset network bandwidth usage volatility. If yes, a bandwidth usage anomaly alert will be sent to the designated personnel; otherwise, an adaptive threshold determination will be performed.

8. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 7, characterized in that, The specific process of adaptive threshold determination is as follows: The ratio of the frequency band energy corresponding to the fault feature to the total energy of the vibration signal is expressed as the fault feature frequency band energy percentage. The ratio of real-time material viscosity to commonly used material viscosity is expressed as the viscosity-condition compatibility factor. The viscosity operating condition adaptation coefficient is divided into a preset number of continuous intervals. For each interval, the operating condition adaptation energy ratio benchmark value of the chemical production machinery and equipment during operation is obtained as the corresponding operating condition adaptation energy ratio benchmark value for that interval. If the energy ratio benchmark value for the working condition adaptation is within the preset energy ratio threshold range, a self-diagnosis qualification prompt for chemical production machinery and equipment will be sent to the preset personnel. Conversely, a self-diagnosis management process for chemical equipment malfunctions should be implemented.

9. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 8, characterized in that, The specific process of the self-diagnosis management of chemical equipment faults is as follows: Real-time vibration spectrum is obtained based on vibration spectrum data; The number of fault characteristic frequency bands whose energy proportion falls within the reference range of fault characteristic frequency band energy proportion is counted, and the ratio of the number of such fault characteristic frequency bands to the total number of reference fault characteristic frequency bands is used as the characteristic frequency similarity. If the feature frequency similarity is greater than or equal to the first-level fault similarity threshold, a first-level early warning will be immediately triggered, prompting the preset personnel to stop the operation of the chemical machinery and equipment. If the feature frequency similarity is greater than or equal to the second-level fault similarity threshold and less than the first-level fault similarity threshold, a second-level early warning will be triggered, prompting the preset maintenance personnel to push fault warning information for chemical machinery and equipment. If the feature frequency similarity is less than the second-level fault similarity threshold, obtain the feature frequency similarity of the preset number of similarity divisions; In the process of implementing self-diagnosis management of chemical equipment faults, the ratio of the difference between randomly selected characteristic frequency similarities and the average characteristic frequency similarities to the average characteristic frequency similarities is used as the similarity change rate. If the similarity change rate is greater than the similarity change threshold, it indicates that the fault characteristics of chemical production machinery and equipment are continuously enhanced, triggering a first-level warning prompt to stop the operation of chemical machinery and equipment by preset personnel; Conversely, if the chemical production machinery and equipment are operating normally, a self-diagnosis qualification message will be sent to the designated personnel.

10. The self-diagnosis management system for chemical production machinery and equipment based on intelligent inspection as described in claim 7, characterized in that, The specific process of adaptive threshold determination is as follows: The operating speed of chemical production machinery and equipment is continuously monitored by a speed sensor, and the average speed within a preset collection period is used as the real-time speed parameter of the current operating condition of the chemical production machinery and equipment. The ratio of the average rotational speed of chemical production machinery and equipment to the rated rotational speed of the chemical production machinery and equipment is defined as the rotational speed difference coefficient; When the speed difference coefficient is greater than 1, the range corresponding to the current speed of the chemical production machinery and equipment is marked as the high rated speed range; When the speed difference coefficient is less than 1, the range corresponding to the current speed of the chemical production machinery and equipment is marked as the low rated speed range; When the speed difference coefficient is equal to 1, the interval corresponding to the current speed of the chemical production machinery and equipment is marked as the equal speed interval; The key indicator data of chemical production machinery and equipment were statistically analyzed in three continuous intervals, including the baseline thresholds for the energy ratio of fault characteristic frequency bands, vibration signal kurtosis, and vibration signal amplitude ratio.

Images