Early warning method and system for detecting gas leakage and computing equipment

Abstract

The invention discloses an early warning method and system for detecting gas leakage and computing equipment, relates to the technical field of chemical safety management, and solves the technical problem of distortion of acquired monitoring data due to the fact that mutual influence of equipment operation temperature and vibration frequency is not considered in the prior art. A target device group is set according to an attention value; calculating a first influence value of the target equipment group on the equipment in the preset range through the first sensor; setting a grading value for the target equipment group through the first influence value, and grading to obtain an equipment grade; determining the number of second sensors of each level of equipment according to the grading value; acquiring a second influence value through a second sensor; preprocessing a second influence value acquired by a second sensor; setting a signal characteristic of gas leakage based on the preprocessed second influence value, and if the signal characteristic of gas leakage occurs in the second influence value, sending out an early warning signal of gas leakage; according to the invention, false alarms are reduced, and the sensitivity of leakage identification is improved.

Description

Technical Field

[0001] This invention belongs to the field of chemical safety management and relates to sensor detection technology, specifically an early warning method, system and computing device for detecting gas leaks. Background Technology

[0002] In coal-to-methanol chemical workshops, excessive vibration can cause temperature rise. For example, mechanical friction, such as bearing wear and poor gear meshing, can increase component friction and generate local high temperatures. Vibration energy is partially converted into heat energy. At the same time, increased temperature can also cause vibration to accelerate, such as component expansion, which changes the clearance and causes unbalanced vibration. High temperature can also cause lubricating oil viscosity to decrease or carbonize, exacerbating frictional vibration. In addition, gas leaks in methanol chemical workshops may simultaneously cause a sudden drop in temperature and abnormal vibration. Multi-parameter joint analysis is beneficial to reduce false alarms from single signals.

[0003] Existing technologies in coal-to-methanol chemical plants monitor the temperature and vibration of key equipment during operation. When the temperature and vibration exceed a set threshold, it is determined that the key equipment has failed. However, this technology does not take into account that the close proximity between equipment will cause the operating temperature and vibration frequency of the equipment to affect each other, making it impossible to capture the characteristics of complex faults. This results in distorted monitoring data and increases the risk of false alarms.

[0004] This invention provides an early warning method, system, and computing device for detecting gas leaks, in order to solve the above-mentioned technical problems. Summary of the Invention

[0005] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention proposes an early warning method, system, and computing device for detecting gas leaks. This addresses the technical problem in existing technologies in coal-to-methanol chemical plants where monitoring the temperature and vibration of key equipment during operation determines that a key equipment has malfunctioned when the temperature and vibration exceed a set threshold. However, this approach fails to consider that the close proximity of equipment can cause the operating temperature and vibration frequency to affect each other, making it impossible to capture complex fault characteristics. This results in distorted monitoring data and increases the risk of false alarms.

[0006] To achieve the above objectives, a first aspect of the present invention provides an early warning method for detecting gas leaks, comprising:

[0007] The system presets the importance value of equipment in the factory and sets a target equipment group based on the importance value. It calculates the first influence value of the target equipment group on equipment within a preset range using a first sensor. The first influence value includes a first temperature influence value and a first vibration frequency influence value.

[0008] The target equipment group is classified into levels by setting a first influence value and classifying the equipment. The number of second sensors for each level of equipment is determined according to the classification value. The second influence value is obtained through the second sensor. The second influence value collected by the second sensor is preprocessed. The second influence value includes: a second temperature influence value and a second vibration frequency influence value.

[0009] The signal characteristics of gas leakage are set based on the pre-processed second influence value. If the signal characteristics of gas leakage appear in the second influence value, a gas leakage warning signal is issued.

[0010] Preferably, the step of setting the target device group based on the importance value includes:

[0011] Retrieve the importance values, sort them in descending order, label the maximum importance value as Zmax, the minimum importance value as Zmin, and assign importance values ​​to intervals. All devices are marked as the target device group.

[0012] It should be noted that the importance value is set based on the value of the equipment, its necessity in use, and its service life. The higher the value of the equipment, the higher the importance value; the higher the necessity of the equipment in use, the higher the importance value; the shorter the service life of the equipment, the higher the importance value; the importance value ranges from [0, 100].

[0013] Preferably, the step of calculating the first influence value of the target device group on devices within a preset range using the first sensor includes:

[0014] Determine the number of devices in the target device group, and mark the device with the Nth importance value in the target device group as the Nth target device within a preset range; N is a positive integer.

[0015] The system retrieves data from the first target device to the Nth target device and calculates several first influence values ​​of other devices on the Nth target device within a preset range using the first sensor.

[0016] It should be noted that the maximum value of N lies within the interval [0, 1]. The total number of devices included; the first sensor is the initial sensor installed for each target device; the initial sensors include: a temperature sensor and a vibration frequency sensor; other devices are: devices other than those that need to calculate the first influence value.

[0017] Preferably, the method for obtaining the first influence value includes:

[0018] Set a preset calculation range, turn off the target device within the calculation range, and turn on other devices except the target device;

[0019] The heat output Q of the i-th other device is obtained through a sensor.i Where i = 1, 2, 3, ..., n;

[0020] Through formula The first temperature influence value of all devices on the pre-selected target device is calculated; where A i Let d be the heat transfer area of ​​the i-th other device to the pre-selected target device. i The distance between the i-th other device and the target device is τ; τ is the preset thermal diffusivity.

[0021] Determine whether the real-time vibration frequency of the target device is 0; if yes, record the first vibration frequency influence value as 0; otherwise, mark the real-time vibration frequency of the target device as the first vibration frequency influence value of all non-target devices on the target device.

[0022] It should be noted that the thermal diffusivity τ in air is 0.026m. 2 / s.

[0023] Preferably, the step of setting a classification value for the target equipment group through a first influence value and classifying it to obtain the equipment level includes:

[0024] The first influence value is retrieved, and the influence value of the first vibration frequency is recorded as ΔZ, T. min T represents the minimum first temperature impact of all non-target devices on the target device. max T represents the maximum first temperature influence of all non-target devices on the target device. p This is the average of all values ​​affected by the first temperature.

[0025] Through formula The graded value FJ is calculated; where α and β are both weight coefficients greater than 0, and α+β=1;

[0026] Set classification thresholds F1 and F2. Target devices with classification values ​​greater than or equal to the first classification threshold F1 are marked as Level 1 devices; target devices with classification values ​​less than or equal to the second classification threshold F2 are marked as Level 3 devices; and target devices other than Level 1 and Level 3 devices are marked as Level 2 devices. Among them, classification threshold F1 is the median of all classification values; and classification threshold F2 is half of the median.

[0027] It should be noted that if T max =T mun Then Set to a fixed value of 0.5; α and β are set according to the sensitivity of the target device to temperature and vibration frequency. The greater the sensitivity of the target device to temperature, the greater α; the greater the sensitivity of the target device to vibration frequency, the greater β.

[0028] Preferably, determining the number of second sensors for each level of device based on the grading value includes:

[0029] Determine whether grading factor one is greater than grading factor two when calculating the grading value; if yes, increase the number of temperature sensors while keeping the number of vibration frequency sensors unchanged; if no, increase the number of vibration frequency sensors while keeping the number of temperature sensors unchanged.

[0030] The number of second sensors in each level of equipment includes: the number of first sensors is denoted as CS;

[0031] When F1 is greater than F2, the formula is used. The number of second sensors in each level of equipment is calculated; where k1, k2 and k3 are increment factors, and k1, k2 and k3 are all unequal positive integers, and k1 > k2 > k3;

[0032] When F1 is less than F2, the formula is used. The number of second sensors for each level of equipment was calculated.

[0033] It should be noted that, as can be seen from the formula for calculating S, the first sensor is equivalent to the basic sensor, and the second sensor is a sensor added on the basis of the first sensor; at the same time, both the first and second sensors include a vibration frequency sensor and a temperature sensor; when the grading factor one equals the grading factor two, the second sensor is not added, and the second influence value is the first influence value.

[0034] Preferably, the preprocessing of the second influence value collected by the second sensor includes:

[0035] The second influence value is retrieved, and the spectral data F(f) of the second temperature influence value and the second vibration frequency influence value in the target device is obtained through Fourier transform. The spectral data of the second temperature influence value and the second vibration frequency influence value in the i-th non-target device is denoted as F(f). i ;

[0036] The spectral data is denoised using wavelet thresholding; the window length range of the moving average filter is set; an adaptive filtering algorithm based on gradient descent is invoked, and the vibration signal of the non-target device is used as the reference input for adaptive filtering. The filter function is constructed to suppress spectral data other than F(f).

[0037] The target device signal is separated by dynamically adjusting the filter weights using an adaptive filtering algorithm based on gradient descent, and error signals are output separately. The error signals include the denoised target temperature signal and the target vibration frequency signal.

[0038] The filter function is defined by the formula: The filter function H is calculated. k (f); where δ is the suppression bandwidth;

[0039] The suppression bandwidth δ is obtained as follows: F(f) i The bandwidth corresponding to the maximum value in the middle.

[0040] Preferably, the method of setting the signal characteristics of gas leakage based on the preprocessed second influence value includes:

[0041] The target temperature signal and the target vibration frequency signal are retrieved, and the timestamps of the two signals are aligned through dynamic time adjustment. The two signals are then integrated into the same spectrum using the short-time Fourier transform method.

[0042] The nonlinear coupling relationship between the target temperature signal and the target vibration frequency signal is identified by PCA analysis algorithm, and the signal characteristics of gas leakage are extracted.

[0043] The signal characteristics of the gas leak include: signal characteristic one of the gas leak, signal characteristic two of the gas leak, and signal characteristic three of the gas leak;

[0044] Among them, one of the signal characteristics of gas leakage is that the phase difference between the harmonics of the target temperature signal fluctuation and the harmonics of the target vibration frequency is greater than a preset threshold.

[0045] The second signal characteristic of gas leakage is that a signal with a frequency band of 50-500Hz appears in the spectrum of the target vibration frequency signal within three consecutive cycles, and the slope of the target temperature signal spectrum appears negative within the corresponding cycle.

[0046] The third signal characteristic of gas leakage is that, during the increase of the target vibration frequency signal, the slopes of the target temperature signals collected by different temperature sensors are opposite.

[0047] To achieve the above objectives, a second aspect of the present invention provides an early warning system for detecting gas leaks, comprising: a data acquisition module, a data processing module, and a monitoring and early warning module;

[0048] Data acquisition module: presets the importance value of equipment in the factory, sets the target equipment group according to the importance value; calculates the first influence value of the target equipment group on equipment within the preset range through the first sensor; wherein, the first influence value includes: first temperature influence value and first vibration frequency influence value;

[0049] Data processing module: The module sets a classification value for the target equipment group based on the first influence value and classifies the equipment to obtain the equipment level; determines the number of second sensors for each level of equipment based on the classification value; acquires the second influence value through the second sensor; and preprocesses the second influence value collected by the second sensor; wherein the second influence value includes: a second temperature influence value and a second vibration frequency influence value.

[0050] Monitoring and early warning module: Based on the pre-processed second influence value, the signal characteristics of gas leakage are set. If the signal characteristics of gas leakage appear in the second influence value, an early warning signal of gas leakage is issued.

[0051] To achieve the above objectives, a third aspect of the present invention provides a gas leak detection early warning computing device, comprising: a memory and a processor, wherein the memory stores executable instructions of the processor; wherein the processor is configured to execute the executable instructions to implement a gas leak detection early warning method provided in the first aspect.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. This invention analyzes the differences in local vibration and temperature patterns caused by leaks at different locations; for example, leaks at pipe welds may cause high-frequency vibrations and local temperature drops; leaks at the bottom of storage tanks may be accompanied by low-frequency vibrations and regional temperature stratification; spatial correlation analysis through a multi-sensor network can narrow down the fault range; in the early stages of chemical gas leaks, the leakage amount is small and may only cause slight temperature gradient changes or local vibration modulation; the weak correlation between vibration and temperature can be captured, allowing for earlier detection of potential hazards than simply relying on pressure or concentration sensors.

[0054] 2. This invention utilizes a first sensor to calculate the first impact value of the target equipment group on equipment within a preset range, including temperature impact value and vibration frequency impact value. This helps to comprehensively understand the impact of the target equipment on the surrounding environment and other equipment, providing a scientific basis for subsequent equipment classification and sensor placement. Preprocessing the second impact value collected by the second sensor can remove interference factors such as noise and outliers, improving the accuracy and reliability of the data and providing a solid foundation for subsequent gas leak judgment. Based on the preprocessed second impact value, the signal characteristics of gas leaks are set, and the occurrence of these characteristics is monitored in real time. Once the signal characteristics of gas leaks are detected, an early warning signal is immediately issued, which helps to take timely measures to prevent the accident from escalating. Attached Figure Description

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0056] Figure 1 This is a schematic diagram of the principle structure of the present invention;

[0057] Figure 2 A schematic diagram illustrating the specific steps for determining the number of sensors in this invention;

[0058] Figure 3 This is a schematic diagram illustrating the specific steps of the monitoring and early warning system of the present invention;

[0059] Figure 4 This is a schematic diagram illustrating the module relationships included in this invention. Detailed Implementation

[0060] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.

[0061] Please see Figure 1 The first aspect of the present invention provides an early warning method for detecting gas leaks, comprising:

[0062] The system presets the importance value of equipment in the factory and sets a target equipment group based on the importance value. It calculates the first influence value of the target equipment group on equipment within a preset range using a first sensor. The first influence value includes a first temperature influence value and a first vibration frequency influence value.

[0063] The target equipment group is classified into levels by setting a first influence value and classifying the equipment. The number of second sensors for each level of equipment is determined according to the classification value. The second influence value is obtained through the second sensor. The second influence value collected by the second sensor is preprocessed. The second influence value includes: a second temperature influence value and a second vibration frequency influence value.

[0064] The signal characteristics of gas leakage are set based on the pre-processed second influence value. If the signal characteristics of gas leakage appear in the second influence value, a gas leakage warning signal is issued.

[0065] Please see Figure 2The specific steps for determining the number of sensors are as follows: Preset the importance values ​​for equipment in the factory, sort the importance values ​​in descending order, mark the maximum importance value as Zmax, the minimum importance value as Zmin, and assign importance values ​​within the specified intervals. All devices are marked as the target device group; the number of devices in the target device group is determined, and the device with the Nth importance value in the target device group is marked as the Nth target device within a preset range; N is a positive integer;

[0066] Retrieve data from the first target device to the Nth target device, and calculate several first influence values ​​of other devices on the Nth target device using the first sensor;

[0067] Set a preset calculation range, turn off the target device within the calculation range, and turn on other devices except the target device;

[0068] The heat output Q of the i-th other device is obtained through a sensor. i Where i = 1, 2, 3, ..., n;

[0069] Through formula The first temperature influence value of all devices on the pre-selected target device is calculated; where A i Let d be the heat transfer area of ​​the i-th other device to the pre-selected target device. i Let be the distance between the i-th other device and the target device; τ be the thermal diffusivity; where the heat transfer area is the heat exchange area along the airflow path between the two devices where the heat is not zero;

[0070] Determine whether the real-time vibration frequency of the target device is 0; if yes, record the first vibration frequency influence value as 0; otherwise, mark the real-time vibration frequency of the target device as the first vibration frequency influence value of all non-target devices on the target device.

[0071] The first influence value is retrieved, and the influence value of the first vibration frequency is recorded as ΔZ, T. min T represents the minimum first temperature impact of all non-target devices on the target device. max T represents the maximum first temperature influence of all non-target devices on the target device. p This is the average of all values ​​affected by the first temperature.

[0072] Through formula The graded value FJ is calculated; where α and β are both weight coefficients greater than 0, and α+β=1;

[0073] Set classification thresholds F1 and F2. Target devices with classification values ​​greater than or equal to the first classification threshold F1 are marked as Level 1 devices; target devices with classification values ​​less than or equal to the second classification threshold F2 are marked as Level 3 devices; and target devices other than Level 1 and Level 3 devices are marked as Level 2 devices. Among them, classification threshold F1 is the median of all classification values; and classification threshold F2 is half of the median.

[0074] Determine whether grading factor one is greater than grading factor two when calculating the grading value; if yes, increase the number of temperature sensors while keeping the number of vibration frequency sensors unchanged; if no, increase the number of vibration frequency sensors while keeping the number of temperature sensors unchanged.

[0075] The methods for obtaining the number of second sensors for each level of equipment include: denoting the number of first sensors as CS; and using the formula... The number of second sensors in each level of equipment is calculated; where k1, k2 and k3 are increment factors, and k1, k2 and k3 are all unequal positive integers, and k1 > k2 > k3.

[0076] For example, an engine assembly line in a chemical workshop contains 10 key pieces of equipment (numbered AJ). Based on the criticality of each piece of equipment to the production line (e.g., failure rate, maintenance cost, impact on production capacity), the 10 pieces of equipment are assigned a importance value (0-100 points), arranged in descending order as follows: Equipment A is an engine assembly robot, importance value 90; Equipment B is a welding machine, importance value 85; Equipment C is a spraying machine, importance value 80; Equipment D is a cooling system, importance value 75; Equipment E is a conveyor belt, importance value 70; Equipment F is a testing machine, importance value 65; Equipment G is a warehouse robot, importance value 60; Equipment H is an air compressor, importance value 55; Equipment I is an air conditioning system, importance value 50; Equipment J is a lighting system, importance value 45.

[0077] Maximum importance value Zmax = 90 (equipment A); Minimum importance value Zmin = 45 (equipment J);

[0078] The target equipment group is: the importance value is in the range That is, the equipment within (67.5, 90] (i.e., all equipment except F, G, H, I, J, a total of 5 units);

[0079] Suppose that the current target equipment to be analyzed and repaired is the N=3rd target equipment (arranged in descending order of importance value), namely equipment C (spraying equipment); stop the target equipment (i.e. equipment C) within the calculation range and start the non-target equipment.

[0080] The first sensor (temperature sensor) measures several first influence values ​​of the non-target device on the target device C;

[0081] Real-time heat generation (in W) of non-target devices is obtained using a temperature sensor.

[0082] Equipment A: 200W, Equipment B: 500W, Equipment D: 1200W, Equipment E: 800W, Equipment F: 500W, Equipment G: 300W, Equipment H: 700W, Equipment I: 400W, Equipment J: 400W;

[0083] Where A1 represents the heat transfer area of ​​the Ath other device to the target device C, which is 2.6m². 2 A2 represents the heat transfer area of ​​the target device C from the other device B, which is 1.56 m². 2 A3 represents the heat transfer area of ​​the Dth other device to the target device C, which is 0.89 m². 2 A4 represents the heat transfer area of ​​the Eth other device to the target device C, which is 1.1 m². 2 A5 represents the heat transfer area of ​​the Fth other device to the target device C, which is 2.89 m². 2 A6 represents the heat transfer area of ​​the Gth other device to the target device C, which is 0.53 m². 2 A7 represents the heat transfer area of ​​the Hth other device to the target device C, which is 2.64 m². 2 A8 represents the heat transfer area of ​​the Fth other device to the target device I, which is 3.65m². 2 A9 represents the heat transfer area of ​​the Fth other device to the target device J, which is 0.98 m². 2 .

[0084] Through formula The first temperature influence values ​​of all non-target devices on target device C are calculated; where d is set. i =1m;

[0085] Taking devices A, B, D, and E as target devices, the calculated first temperature influence of all non-target devices on target device A is: 9.26 × 10⁻⁶. -13 The calculated first temperature influence of all non-target devices on target device B is: 1.98 × 10⁻⁶. -13 The calculated first temperature influence of all non-target devices on target device D is: 2.65 × 10⁻⁶. -13 The calculated first temperature influence of all non-target devices on target device E is 6.37 × 10⁻⁶. -13 .

[0086] If the real-time vibration frequency of the target device C is obtained by the vibration sensor as 15kHz, then the influence value ΔZ of the first vibration frequency of all non-target devices on the target device is 15kHz.

[0087] T max 1.98×10 -13 , T min It is 9.26×10 -13 , T p =5.63×10 -13 ;

[0088] Through formula The classification value FJ of device C is calculated. c The value is 13.15; where α and β are determined according to the grading factor table (as shown in Table 1);

[0089] Since the target equipment's preset temperature sensitivity level is Level 1 (the preset temperature sensitivity level of the target equipment is set by those skilled in the art based on actual working conditions), then α = 0.2, β = 0.8. (Table 1: Grading Factor Table)

[0090] The target device's preset temperature sensitivity level Grading Factor 1 Grading Factor Two Level 1 0.2 0.8 Level 2 0.4 0.6 Level 3 0.6 0.4 Level 4 0.8 0.2

[0091] Based on the above steps, the classification value FJ of device A is obtained. A The rating value FJ for device B is 5.69. B The classification value FJ of equipment D is 10.15. D The classification value FJ for equipment E is 19.22. E It is 9.14;

[0092] Set the classification threshold F1 to the median of the target device group, 10.15, and the classification threshold F2: F1 / 2 = 5.08;

[0093] Since the grading factor 1 (0.2) is less than the grading factor 2 (0.8), the number of vibration frequency sensors is increased while the number of temperature sensors remains unchanged.

[0094] The method to increase the number of vibration frequency sensors is as follows: the number of the first sensor (the sum of the number of vibration frequency sensors 2 and the number of temperature sensors 4) is 6;

[0095] Through formula The number of second sensors for each level of equipment is calculated; where k1, k2 and k3 are incremental factors and are obtained through the incremental factor acquisition table (as shown in Table 2).

[0096] Table 2. Increment Factor Acquisition Table

[0097] The range of values ​​for CS k1 k2 k3 [0，4] 2 3 4 [5，9] 1 2 3 [10，∞] 0 1 2

[0098] This invention constructs a three-dimensional spatial heat conduction model by introducing the parameters of heat transfer area Ai and equipment spacing di, which can accurately quantify the impact of equipment layout on heat conduction efficiency, providing a quantitative optimization basis for scenarios such as data center rack arrangement and factory equipment layout. The dual-logic judgment mechanism of vibration frequency effectively solves the coupling problem between self-excited vibration and externally transmitted vibration, making it particularly suitable for vibration source tracing in precision manufacturing scenarios. By setting a classification value for the target equipment group based on the first influence value and then classifying it to obtain the equipment level, it is equivalent to further classifying the target equipment group based on the importance value classification, thus ensuring that important equipment with high importance value and susceptible to influence from other equipment is continuously monitored, which is beneficial for timely maintenance or other treatment of important equipment. Through spatial redundant sensor deployment, a multi-dimensional spatial monitoring network is formed to obtain more refined local data.

[0099] Please see Figure 3 The specific steps for monitoring and early warning are as follows: retrieve the second influence value, obtain the spectrum data F(f) of the second temperature influence value and the second vibration frequency influence value in the target device through Fourier transform, and record the spectrum data of the second temperature influence value and the second vibration frequency influence value in the i-th non-target device as F(f). i ;

[0100] The spectral data is denoised using wavelet thresholding; the window length range of the moving average filter is set; an adaptive filtering algorithm based on gradient descent is invoked, and the vibration signal of the non-target device is used as the reference input for adaptive filtering. The filter function is constructed to suppress spectral data other than F(f).

[0101] The target device signal is separated by dynamically adjusting the filter weights using an adaptive filtering algorithm based on gradient descent, and error signals are output separately. The error signals include the denoised target temperature signal and the target vibration frequency signal.

[0102] The target temperature signal and the target vibration frequency signal are retrieved, and the timestamps of the two signals are aligned through dynamic time adjustment. The two signals are then integrated into the same spectrum using the short-time Fourier transform method.

[0103] The nonlinear coupling relationship between the target temperature signal and the target vibration frequency signal is identified by PCA analysis algorithm, and the signal characteristics of gas leakage are extracted.

[0104] The signal characteristics of gas leaks include: signal characteristic one of gas leaks, signal characteristic two of gas leaks, and signal characteristic three of gas leaks;

[0105] Among them, the signal characteristics of gas leakage include: the phase difference between the harmonics of the target temperature signal fluctuation and the harmonics of the target vibration frequency is greater than a preset threshold.

[0106] The second signal characteristic of gas leakage is that a signal with a frequency band of 50-500Hz appears in the spectrum of the target vibration frequency signal within three consecutive cycles, and the slope of the target temperature signal spectrum appears negative within the corresponding cycle.

[0107] The third signal characteristic of gas leakage is that, during the increase of the target vibration frequency signal, the slopes of the target temperature signals collected by different temperature sensors are opposite.

[0108] For example, retrieve the second influence value, and obtain the spectral data F(f) of the second temperature influence value and the second vibration frequency influence value in the target device through Fourier transform. Record the spectral data of the second temperature influence value and the second vibration frequency influence value in the i-th non-target device as F(f). i ;

[0109] The spectral data is denoised using the Sym4 wavelet basis; the window length of the moving average filter is set to 5-10 seconds; an adaptive filtering algorithm based on gradient descent is invoked, and the vibration signal of the non-target device is used as the reference input for the adaptive filtering. The filter function is constructed to suppress spectral data other than F(f).

[0110] The target device signal is separated by dynamically adjusting the filter weights using an adaptive filtering algorithm based on gradient descent, and error signals are output separately. The error signals include the denoised target temperature signal and the target vibration frequency signal.

[0111] This invention employs spectral analysis to separate fault characteristic frequencies from complex vibration signals. By using spectral data, it analyzes and obtains spectral data of abnormal and normal vibrations, thereby reducing the generation of interference frequencies and decreasing the possibility of misjudgment.

[0112] Please see Figure 4 A second aspect of the present invention provides an early warning system for detecting gas leaks, comprising: a data acquisition module, a data processing module, and a monitoring and early warning module;

[0113] Data acquisition module: presets the importance value of equipment in the factory, sets the target equipment group according to the importance value; calculates the first influence value of the target equipment group on equipment within the preset range through the first sensor; wherein, the first influence value includes: first temperature influence value and first vibration frequency influence value;

[0114] Data processing module: The module sets a classification value for the target equipment group based on the first influence value and classifies the equipment to obtain the equipment level; determines the number of second sensors for each level of equipment based on the classification value; acquires the second influence value through the second sensor; and preprocesses the second influence value collected by the second sensor; wherein the second influence value includes: a second temperature influence value and a second vibration frequency influence value.

[0115] Monitoring and early warning module: Based on the pre-processed second influence value, the signal characteristics of gas leakage are set. If the signal characteristics of gas leakage appear in the second influence value, an early warning signal of gas leakage is issued.

[0116] A third aspect of the present invention provides an early warning computing device for detecting gas leaks, comprising: a memory and a processor, wherein the memory stores executable instructions of the processor; wherein the processor is configured to execute the executable instructions to implement an early warning method for detecting gas leaks provided in the first aspect.

[0117] Some of the data in the above formula are calculated by removing dimensions and taking their numerical values. The formula is the closest to the real situation obtained by software simulation of a large amount of collected data. The preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

[0118] The working principle of this invention is as follows: This invention sets a target equipment group based on a preset importance value for equipment in the factory; calculates a first influence value of the target equipment group on equipment within a preset range using a first sensor; sets a grading value for the target equipment group based on the first influence value and grades it to obtain equipment levels; determines the number of second sensors for each level of equipment based on the grading value; acquires a second influence value using the second sensor; preprocesses the second influence value acquired by the second sensor; sets a gas leak signal characteristic based on the preprocessed second influence value; if a gas leak signal characteristic appears in the second influence value, a gas leak warning signal is issued.

[0119] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A method for early warning of gas leaks, characterized in that, include: Preset the importance value of equipment in the factory, and set the target equipment group based on the importance value; The first sensor calculates the first impact value of the target equipment group on the equipment within a preset range; wherein the first impact value includes: a first temperature impact value and a first vibration frequency impact value; The target equipment group is classified into levels by setting a first influence value and classifying the equipment. The number of second sensors for each level of equipment is determined according to the classification value. The second influence value is obtained through the second sensor. The second influence value collected by the second sensor is preprocessed. The second influence value includes: a second temperature influence value and a second vibration frequency influence value. The signal characteristics of gas leakage are set based on the pre-processed second influence value. If the signal characteristics of gas leakage appear in the second influence value, a gas leakage warning signal is issued.

2. The method for early warning of gas leak detection according to claim 1, characterized in that, The step of setting the target device group based on the importance value includes: Retrieve the importance values, sort them in descending order, label the maximum importance value as Zmax, the minimum importance value as Zmin, and assign importance values ​​to intervals. All devices are marked as the target device group.

3. The method for early warning of gas leak detection according to claim 1, characterized in that, The calculation of the first impact value of the target device group on devices within a preset range using the first sensor includes: Determine the number of devices in the target device group, and mark the device with the Nth importance value in the target device group as the Nth target device within a preset range; N is a positive integer. The system retrieves data from the first target device to the Nth target device and calculates several first influence values ​​of other devices on the Nth target device within a preset range using the first sensor.

4. The method for early warning of gas leak detection according to claim 3, characterized in that, The methods for obtaining the first influence value include: Set a preset calculation range, turn off the target device within the calculation range, and turn on other devices except the target device; The heat output Q of the i-th other device is obtained through a sensor. i Where i = 1, 2, 3, ..., n; Through formula The first temperature influence value of all devices on the pre-selected target device is calculated; where A i Let d be the heat transfer area of ​​the i-th other device to the pre-selected target device. i The distance between the i-th other device and the target device is τ; τ is the thermal diffusivity. Determine whether the real-time vibration frequency of the target device is 0; if yes, record the first vibration frequency influence value as 0; otherwise, mark the real-time vibration frequency of the target device as the first vibration frequency influence value of all non-target devices on the target device.

5. The method for early warning of gas leak detection according to claim 1, characterized in that, The step of setting a classification value for the target equipment group based on a first influence value and then classifying it to obtain the equipment level includes: The first influence value is retrieved, and the influence value of the first vibration frequency is recorded as ΔZ, T. min T represents the minimum first temperature impact of all non-target devices on the target device. max T represents the maximum first temperature influence of all non-target devices on the target device. p This is the average of all values ​​affected by the first temperature. Through formula The graded value FJ is calculated; where α and β are both weight coefficients greater than 0, and α+β=1; Set classification thresholds F1 and F2. Target devices with classification values ​​greater than or equal to the first classification threshold F1 are marked as Level 1 devices; target devices with classification values ​​less than or equal to the second classification threshold F2 are marked as Level 3 devices; and target devices other than Level 1 and Level 3 devices are marked as Level 2 devices. Among them, classification threshold F1 is the median of all classification values; and classification threshold F2 is half of the median.

6. The method for early warning of gas leak detection according to claim 1, characterized in that, The determination of the number of second sensors for each level of equipment based on the grading value includes: Determine whether grading factor one is greater than grading factor two when calculating the grading value; if yes, increase the number of temperature sensors while keeping the number of vibration frequency sensors unchanged; if no, increase the number of vibration frequency sensors while keeping the number of temperature sensors unchanged. The number of second sensors in each level of equipment includes: the number of first sensors is denoted as CS; When F1 is greater than F2, the formula is used. The number of second sensors in each level of equipment is calculated; where k1, k2 and k3 are increment factors, and k1, k2 and k3 are all unequal positive integers, and k1 > k2 > k3; When F1 is less than F2, the formula is used. The number of second sensors for each level of equipment was calculated.

7. The method for early warning of gas leak detection according to claim 1, characterized in that, The preprocessing of the second influence value collected by the second sensor includes: The second influence value is retrieved, and the spectral data F(f) of the second temperature influence value and the second vibration frequency influence value in the target device is obtained through Fourier transform. The spectral data of the second temperature influence value and the second vibration frequency influence value in the i-th non-target device is denoted as F(f). i ; The spectral data is denoised using wavelet thresholding; the window length range of the moving average filter is set; an adaptive filtering algorithm based on gradient descent is invoked, and the vibration signal of the non-target device is used as the reference input for adaptive filtering. The filter function is constructed to suppress spectral data other than F(f). An adaptive filtering algorithm based on gradient descent dynamically adjusts the filter weights to separate the target device signal and output error signals. The error signals include the denoised target temperature signal and the target vibration frequency signal. The filter function is defined by the formula: The filter function H is calculated. k (f); where δ is the suppression bandwidth; The suppression bandwidth δ is obtained as follows: F(f) i The bandwidth corresponding to the maximum value in the middle.

8. The method for early warning of gas leak detection according to claim 7, characterized in that, The signal characteristics of gas leakage based on the preprocessed second influence value include: The target temperature signal and the target vibration frequency signal are retrieved, and the timestamps of the two signals are aligned through dynamic time adjustment. The two signals are then integrated into the same spectrum using the short-time Fourier transform method. The nonlinear coupling relationship between the target temperature signal and the target vibration frequency signal is identified by PCA analysis algorithm, and the signal characteristics of gas leakage are extracted. The signal characteristics of the gas leak include: signal characteristic one of the gas leak, signal characteristic two of the gas leak, and signal characteristic three of the gas leak; Among them, one of the signal characteristics of gas leakage is that the phase difference between the harmonics of the target temperature signal fluctuation and the harmonics of the target vibration frequency is greater than a preset threshold. The second signal characteristic of gas leakage is that a signal with a frequency band of 50-500Hz appears in the spectrum of the target vibration frequency signal within three consecutive cycles, and the slope of the target temperature signal spectrum appears negative within the corresponding cycle. The third signal characteristic of gas leakage is that, during the increase of the target vibration frequency signal, the slopes of the target temperature signals collected by different temperature sensors are opposite.

9. A gas leak detection early warning system, adapted to the gas leak detection early warning method according to any one of claims 1-8, characterized in that, include: Data acquisition module, data processing module, and monitoring and early warning module; Data acquisition module: Presets the importance value of equipment in the factory, and sets the target equipment group based on the importance value; The first sensor calculates the first impact value of the target equipment group on the equipment within a preset range; wherein the first impact value includes: a first temperature impact value and a first vibration frequency impact value; Data processing module: The module sets a classification value for the target equipment group based on the first influence value and classifies the equipment to obtain the equipment level; determines the number of second sensors for each level of equipment based on the classification value; acquires the second influence value through the second sensor; and preprocesses the second influence value collected by the second sensor; wherein the second influence value includes: a second temperature influence value and a second vibration frequency influence value. Monitoring and early warning module: Based on the pre-processed second influence value, the signal characteristics of gas leakage are set. If the signal characteristics of gas leakage appear in the second influence value, an early warning signal of gas leakage is issued.

10. A gas leak detection and early warning computing device, characterized in that, include: A memory and a processor, wherein the memory stores executable instructions of the processor; wherein the processor is configured to perform an early warning method for detecting gas leaks according to any one of claims 1-8 by executing the executable instructions.

Images