Dust collector filter bag breakage detection method and system, electronic device, and storage medium

Abstract

The application provides a dust collector filter bag breakage detection method and system, an electronic device and a storage medium, and belongs to the technical field of dust collectors. The method comprises the following steps: acquiring dust concentration data passing through the filter bag and working condition parameter data of the filter bag within a target time length; pre-processing the dust concentration data to obtain first dust concentration data, obtaining a dust concentration waveform based on the first dust concentration data, extracting features of the dust concentration waveform, and determining a dust concentration waveform type in combination with the working condition parameter data of the filter bag; positioning a fault position of the filter bag based on the dust concentration waveform type, and determining a breakage cause of the filter bag based on the fault position of the filter bag, a differential pressure of a chamber where the filter bag is located, a temperature and material level detection data of a hopper corresponding to the filter bag. The dust collector filter bag breakage detection method and system, the electronic device and the storage medium provided by the application can accurately position a filter bag breakage position and determine a filter bag breakage cause.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of dust collectors, and more particularly relates to a dust collector filter bag damage detection method and system, an electronic device, and a storage medium. BACKGROUND

[0002] In the industrial production process, dust collectors are widely used in various production scenarios. They filter dust-containing gas through filter bags to separate dust from gas, thereby purifying gas and reducing dust emissions, which plays a crucial role in environmental protection and production safety. However, filter bags are prone to damage during long-term operation. Once the filter bag is damaged, not only will the dust removal efficiency be reduced, leading to excessive dust emissions and environmental pollution, but it may also affect the normal operation of production equipment and even cause safety accidents. Currently, the detection methods for dust collector filter bag damage mostly rely on manual inspection or only on dust concentration, making it difficult to detect the damage location and cause of the filter bag. SUMMARY

[0003] The application aims to provide a dust collector filter bag damage detection method and system, an electronic device, and a storage medium to accurately locate the damage location of the filter bag and determine the cause of the damage.

[0004] The first aspect of the application provides a dust collector filter bag damage detection method, which includes:

[0005] Obtaining dust concentration data passing through the filter bag and working condition parameter data of the filter bag within a target time period;

[0006] Preprocessing the dust concentration data to obtain first dust concentration data, obtaining a dust concentration waveform based on the first dust concentration data, extracting features from the dust concentration waveform, and determining the dust concentration waveform type in combination with the working condition parameter data of the filter bag;

[0007] Determining whether the filter bag is damaged based on the dust concentration waveform type, positioning the fault location of the filter bag based on the dust concentration waveform type if the filter bag is damaged, and determining the cause of the filter bag damage based on the fault location of the filter bag, the differential pressure of the chamber where the filter bag is located, the temperature, and the material level detection data of the hopper corresponding to the filter bag.

[0008] The second aspect of the application provides a dust collector filter bag damage detection system, which includes:

[0009] A data acquisition module for obtaining dust concentration data passing through the filter bag and working condition parameter data of the filter bag within a target time period;

[0010] The data analysis module is configured to preprocess dust concentration data to obtain first dust concentration data, obtain a dust concentration waveform based on the first dust concentration data, extract features of the dust concentration waveform, and determine a dust concentration waveform type in combination with working condition parameter data of the filter bag.

[0011] The damage detection module is configured to determine whether the filter bag is damaged based on the dust concentration waveform type, locate a fault position of the filter bag based on the dust concentration waveform type if the filter bag is damaged, and determine a cause of the damage of the filter bag based on the fault position of the filter bag, a differential pressure of a chamber in which the filter bag is located, a temperature, and material level detection data of a hopper corresponding to the filter bag.

[0012] In a third aspect, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of the dust collector filter bag damage detection method described above are implemented.

[0013] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the steps of the dust collector filter bag damage detection method described above are implemented.

[0014] The dust collector filter bag damage detection method and system, the electronic device, and the storage medium provided by the embodiments of the present application have the following beneficial effects:

[0015] First, the dust concentration data and the working condition parameter data in a target time period are obtained, the dust concentration data is preprocessed and waveform analysis is performed, the fault position of the filter bag is located based on the dust concentration waveform type, and the specific position of the damaged filter bag can be quickly determined. This method of judging whether the filter bag is damaged and determining the fault position by comprehensively analyzing multiple aspects of information can reduce misjudgment and missed judgment, help the staff quickly find the problem filter bag, avoid blind investigation, and thus improve the accuracy of detection.

[0016] Second, the cause of the damage of the filter bag is determined in combination with the fault position of the filter bag, the differential pressure of the chamber in which the filter bag is located, the temperature, and the material level detection data of the hopper, which can deeply analyze the root cause of the damage of the filter bag from multiple angles such as equipment operation conditions and environmental conditions, and is more accurate than the traditional single discrimination method. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort.

[0018] Figure 1 A flowchart of a filter bag breakage detection method provided by an embodiment of the present application is shown in FIG. 1.

[0019] Figure 2 A structural block diagram of a filter bag breakage detection system provided by an embodiment of the present application is shown in FIG. 2.

[0020] Figure 3 A schematic block diagram of an electronic device provided by an embodiment of the present application is shown in FIG. 3. DETAILED DESCRIPTION

[0021] In the following description, specific details are set forth, such as particular system configurations, techniques, etc., in order to provide a thorough understanding of the present application. However, persons skilled in the art will understand that the present application can be practiced without these specific details. In order to make the present application, technical solutions and advantages clearer, specific embodiments will be described in detail below with reference to the accompanying drawings.

[0022] Reference is made to Figure 1 , Figure 1 A flowchart of a filter bag breakage detection method provided by an embodiment of the present application can be executed by an electronic device, and the method can include the following steps.

[0023] S101: Obtain dust concentration data passing through the filter bag and working condition parameter data of the filter bag within a target time period.

[0024] In the present embodiment, the dust collector is a device used in industrial production to separate dust particles in dust-containing gas, realize gas purification and dust recovery, and is widely used in the fields of mining, metallurgy, chemical industry, electric power, etc. The dust collector can remove dust in the gas by filtering, electrostatic adsorption, centrifugal separation, etc., reduce dust emission, protect the environment and ensure the normal operation of production equipment.

[0025] The dust collector is generally composed of a clean gas layer, a filter layer, a dust conveying layer and a control unit, wherein the filter bag in the filter layer is the core component of the dust collector, usually made of fiber material (such as polyester, PTFE, etc.), in a bag-shaped structure, used to intercept dust particles in dust-containing gas.

[0026] The target time period refers to a specific time period for data collection, such as 1 hour, 8 hours or a production cycle, etc. By limiting the time range of data collection, the present embodiment facilitates the analysis of the running state of the filter bag within a specific time period, eliminates the interference of short-term fluctuations or abnormal data, and obtains regularity data with statistical significance.

[0027] Dust concentration data refers to the mass or number of dust particles in a unit volume of dust-containing gas, usually expressed in mg / m³ (mass concentration) or pieces / m³ (number concentration). Dust concentration is a key indicator of filter bag filtration effect. During normal operation, the dust concentration downstream of the filter bag should be maintained at a low level; if the filter bag is damaged, dust will directly penetrate the damaged area, causing the downstream dust concentration to rise significantly.

[0028] Filter bag working condition parameter data reflects various parameters of the filter bag operating environment and working state, including but not limited to: dust cleaning frequency, dust cleaning time, pulse blowing interval, vibration intensity, gas flow, etc. The working frequency or single dust cleaning duration of the dust cleaner dust cleaning system can reflect the running state of the dust cleaning system. Improper dust cleaning can exacerbate filter bag wear or cause dust to be re-adsorbed.

[0029] In this embodiment, during the operation of the dust collector, the dust concentration values passing through the filter bag within a certain period of time (i.e. target length) are collected to reflect the filtration effect of the filter bag; at the same time, various working condition parameter data of the filter bag during operation, such as dust cleaning time / frequency, gas flow, etc., are collected to characterize the working environment and state of the filter bag; by obtaining these two types of data, basic data support is provided for subsequent analysis of whether the filter bag is damaged and positioning the fault cause.

[0030] S102: Preprocess the dust concentration data to obtain first dust concentration data, obtain a dust concentration waveform based on the first dust concentration data, extract features from the dust concentration waveform, and determine the dust concentration waveform type in combination with the working condition parameter data of the filter bag.

[0031] In this embodiment, preprocessing the dust concentration data means the process of cleaning, denoising, and removing false data from the original dust concentration data. For example: removing outliers, filling missing data, smoothing fluctuation curves, etc. The purpose is to improve data accuracy and avoid interference in subsequent analysis. The first dust concentration data is standardized data after preprocessing, which can more truly reflect the actual filtration state of the filter bag.

[0032] In this embodiment, the dust concentration waveform is a waveform graph drawn by the first dust concentration data in time sequence, which directly shows the dynamic change trend of the dust concentration within the target length. The shape, peak value, fluctuation frequency, etc. of the waveform can be used to analyze the running condition of the filter bag. Feature extraction is the process of extracting key parameters or patterns that can reflect the running state of the filter bag from the dust concentration waveform. The dust concentration waveform type includes but is not limited to periodic fluctuation waveform, gradual rise waveform, random pulse waveform, stable platform waveform, etc.

[0033] In this embodiment, first, the original dust concentration data is preprocessed to remove noise, correct outliers, etc., to obtain more accurate first dust concentration data; based on this data, the dust concentration waveform is plotted in time sequence, and the dynamic change of the dust concentration is intuitively presented. Secondly, the peak value, fluctuation frequency and other key features are extracted from the dust concentration waveform, and at the same time, the filter bag cleaning time / frequency, pulse blowing interval, gas flow and other working condition parameter data are combined to comprehensively judge which type the dust concentration waveform belongs to. Different types of waveforms have different causes and different formation positions, so this method helps to judge whether the filter bag is damaged and locate the leakage position, so as to realize effective monitoring of the running state of the filter bag.

[0034] S103: judging whether the filter bag is damaged based on the dust concentration waveform type, if the filter bag is damaged, positioning the fault position of the filter bag based on the dust concentration waveform type, and determining the cause of the filter bag damage based on the fault position of the filter bag, the differential pressure of the chamber where the filter bag is located, the temperature and the material level detection data of the hopper corresponding to the filter bag.

[0035] In this embodiment, filter bag damage refers to the phenomenon that the filter bag is damaged due to reasons such as material aging, mechanical wear, external force impact, etc., cracks, holes or sealing failure occur, and dust leaks directly without effective filtration. By analyzing the dust concentration waveform type and the structure of the dust collector, the specific area where the damaged filter bag is located (such as a certain chamber, a certain group of filter bags or a single filter bag) can be determined.

[0036] The chamber in the dust collector refers to the space where the filter bags are grouped. The differential pressure of the chamber where the filter bag is located refers to the gas pressure difference between different chambers (independent spaces where filter bags are grouped) or different filter bag inlets and outlets in the dust collector, reflecting the degree of dust accumulation on the surface of the filter bag and the airflow resistance. Excessive pressure difference may be caused by excessive dust accumulation on the surface of the filter bag (insufficient cleaning) or decreased filter bag permeability, and low pressure difference may be caused by filter bag damage.

[0037] The chamber temperature refers to the gas temperature in the chamber where the filter bag is located, which is affected by process production heat source, environmental temperature and dust properties (such as endothermic and exothermic reactions). Excessive temperature may cause carbonization and melting of the filter bag material, and sudden temperature change may cause thermal expansion and contraction fatigue damage of the filter bag.

[0038] The hopper material level detection data is the real-time monitoring data of the dust accumulation height or weight in the hopper in the dust removal layer at the bottom of the dust collector, which can be obtained by a level meter. Abnormal increase of the material level will cause poor cleaning effect of the filter bag.

[0039] In the embodiment, firstly, according to the dust concentration waveform type, the normal operation characteristics (such as regular fluctuation, concentration threshold) are compared to judge whether the filter bag is damaged, and the preliminary identification of the fault is realized. When the damage is confirmed, the time node, amplitude change of the waveform and the dust remover sub-bin structure are used to lock the bin or group where the damaged filter bag is located, and the positioning problem of "where is the damage" is solved. Finally, combined with the bin differential pressure corresponding to the fault position (judging whether the damage is caused by dust accumulation blockage or air flow short circuit), the temperature (analyzing whether the material damage is caused by insufficient high temperature resistance), the hopper level (checking the failure of the dust cleaning system or the dust leakage), the damage root cause is comprehensively inferred, such as mechanical wear, thermal aging or abnormal dust cleaning.

[0040] From the above, first of all, the dust concentration data and the working condition parameter data in the target time length are acquired, the dust concentration data is preprocessed and waveform analysis is performed, the fault position of the filter bag is positioned based on the dust concentration waveform type, and the specific position of the damaged filter bag can be quickly determined. The method of judging whether the filter bag is damaged and determining the fault position by comprehensively considering multiple aspects of information can reduce the misjudgment and omission, help the staff to quickly find the problem filter bag, avoid blind investigation, and thus improve the detection accuracy.

[0041] Secondly, the filter bag damage reason is determined by combining the filter bag fault position, the differential pressure of the bin where the filter bag is located, the temperature and the hopper level detection data, which can deeply analyze the root cause of the filter bag damage from multiple angles such as equipment operation condition and environmental condition, and is more accurate than the traditional single discrimination method.

[0042] In an embodiment of the present application, the working condition parameter data of the filter bag includes a pulse blowing interval.

[0043] The dust concentration waveform is feature extracted, and the dust concentration waveform type is determined in combination with the working condition parameter data of the filter bag, including:

[0044] The first feature vector is obtained by performing time domain feature extraction on the dust concentration waveform, and the second feature vector is obtained by performing frequency domain feature extraction on the dust concentration waveform;

[0045] The feature weights corresponding to the first feature vector and the second feature vector are adjusted based on the pulse blowing interval, and the target feature vector is obtained by weighted fusion based on the first feature vector, the second feature vector and the feature weights corresponding thereto respectively;

[0046] The dust concentration waveform type is obtained based on the target feature vector and the classification model.

[0047] In the embodiment, feature extraction is performed on the dust concentration waveform, including feature extraction from two dimensions of time domain and frequency domain. The time domain dimension can directly observe the shape of the waveform on the time axis, and the frequency domain dimension can convert the waveform into frequency components. Through multi-dimensional feature extraction, misjudgment caused by insufficient information in a single dimension is avoided.

[0048] In the embodiment, time domain feature extraction refers to analyzing waveform features from the time dimension and extracting parameters directly related to time. Time domain feature extraction on the dust concentration waveform obtains a first feature vector, including extracting the peak value, valley value, peak value number, valley value number, rise time, fall time, pulse width, and average dust concentration between peak values of the dust concentration waveform.

[0049] The peak value represents the maximum value of the dust concentration in a period of time, and the valley value represents the minimum value of the dust concentration in a period of time. The peak value number and the valley value number respectively represent the number of peak values and the number of valley values in a target time length. The rise time represents the time from an adjacent valley value point to a peak value point for any peak value point. The fall time represents the time from a peak value point to an adjacent valley value point for any peak value point. The pulse width represents the duration for which the dust concentration is greater than a target concentration threshold.

[0050] The peak value, valley value, peak value number, valley value number, rise time, fall time, pulse width, and average dust concentration between peak values are taken as the first feature vector. In the embodiment, the first feature vector quantifies time domain feature parameters into a numerical vector, which is a mathematical expression of the time domain features of the waveform.

[0051] In the embodiment, frequency domain feature extraction is performed on the dust concentration waveform to obtain a second feature vector, including extracting frequency, amplitude, phase, and other frequency domain feature parameters. The second feature vector quantifies frequency domain feature parameters into a numerical vector, which is a mathematical expression of the frequency domain features of the waveform.

[0052] In the embodiment, the working condition parameter data of the filter bag includes pulse blowing interval. The pulse blowing interval refers to the time interval between adjacent two pulse blowings when the dust collector filter bag is cleaned, which can reflect the working rhythm of the filter layer cleaning structure. Adjusting the feature weights corresponding to the first feature vector and the second feature vector based on the pulse blowing interval can be achieved by the following formula:

[0053] .

[0054] wherein, represents the feature weight corresponding to the first feature vector, represents the feature weight corresponding to the second feature vector, represents the pulse blowing interval, represents the time domain feature sensitive interval reference value, denotes a time domain feature sensitive interval width, denotes a pulse jet intensity correction factor, denotes a frequency domain attenuation coefficient, denotes a frequency domain gain constant, denotes a prevention zero constant, which can be .

[0055] The target feature vector is obtained by weighted fusion based on the first feature vector, the second feature vector and the respective feature weights corresponding thereto, which can be realized according to the following formula:

[0056] .

[0057] wherein, denotes a channel splicing operation, is the target feature vector, is the first feature vector, is the second feature vector.

[0058] In the present embodiment, when ≈ , , the basic weight is 0.5, denotes that the longer the jet interval time is, the more important the time domain feature is. denotes that the shorter the jet interval time is, the more the frequency domain analysis needs to be strengthened, and the shorter the interval is, the closer to 1, and the longer the interval is, the closer to 0. denotes that the shorter the jet interval is, the greater the frequency domain feature weight is.

[0059] In the present embodiment, each feature in the target feature vector is input into a classification model, and the classification model can judge the probability value of each type of dust concentration waveform type. When the probability value is greater than a preset probability threshold, the model outputs the dust concentration waveform type corresponding to the probability value; when the probability value is less than or equal to the preset probability threshold, the model does not output the dust concentration waveform corresponding to the probability value. The preset probability threshold in the present embodiment is set when the classification model is initialized, and the purpose of setting the preset probability threshold is to enable the model to judge whether to output the corresponding dust concentration waveform type according to the calculated probability value. The classification model can be a decision tree model, a convolutional neural network model, etc.

[0060] As can be seen from the above, the present embodiment adjusts the time domain and frequency domain feature weights according to the pulse jet interval, so that the model can sensitively capture the influence of the jetting period on the dust concentration (for example, a short jet interval can result in more significant periodic fluctuation characteristics), and avoids ignoring the differences in working conditions due to a single fixed weight. The present embodiment also fuses the time domain feature and the frequency domain feature to form a target feature vector in combination with the working condition parameters of the filter bag, which can comprehensively depict the amplitude variation, frequency characteristics and relevance to the jetting period of the waveform, and improve the classification accuracy under complex working conditions.

[0061] In an embodiment of the present application, the training process of the classification model comprises:

[0062] grouping the historical dust concentration waveforms based on the pulse injection intervals to obtain a plurality of groups of analysis training sets;

[0063] calculating a main loss function and a similarity constraint loss function based on the plurality of groups of analysis training sets, and obtaining a target loss function by weighted fusion of the main loss function and the similarity constraint loss function;

[0064] obtaining the trained classification model in response to the target loss function satisfying a preset condition.

[0065] In the embodiment, the formula of the similarity constraint loss function is:

[0066] .

[0067] wherein i and j represent the numbers of two samples in a selected sample pair in the plurality of groups of analysis training sets, represents the similarity constraint loss value, represents a feature extraction function, represents the dust concentration waveform of the i-th sample, represents the dust concentration waveform of the j-th sample, represents a 2-norm, 、 respectively represent the pulse injection interval of the i-th sample and the pulse injection interval of the j-th sample, represents a working condition attenuation coefficient, represents a pulse injection interval difference penalty function.

[0068] In the embodiment, the mechanism of the similarity constraint is that the samples in the same pulse interval time group are gathered in the feature space, while the samples in different pulse interval time groups are automatically released from the constraint. The purpose of this is to strengthen the feature gathering, and thus improve the working condition robustness of the classification model.

[0069] The training process of a traditional classification model is to input all the training samples to the classification model for training to obtain a trained model, but such training is prone to the same class damage, misjudgment of different dust concentration waveform types under different injection intervals, and misclassification of different class damage under the same injection interval. Based on this, the embodiment adds a similarity constraint loss function to the training process of the classification model, which is only for the selected sample pairs in the plurality of groups of analysis training sets, and can greatly improve the robustness of the model.

[0070] In this embodiment, the main loss function and the similarity constraint loss function are calculated based on multiple sets of analysis training sets, wherein the main loss function is responsible for the basic classification task, ensuring the classification accuracy of the model for waveform types such as periodic fluctuations, random pulses, etc. The similarity constraint loss function constrains the feature space distribution by the following logic:

[0071] Sample pair similarity constraint: If two samples come from the same working condition group (pulse blowing interval is similar), the feature vectors of the two samples are forced to be close in space; if they come from different working condition groups, the sample pairs with larger interval differences are penalized to increase the feature distance. Working condition decay coefficient The constraint strength can be adjusted to avoid excessive dependence on the blowing interval, which may lead to a decrease in the model's generalization ability.

[0072] The target loss function is a weighted sum of the main loss function and the similarity constraint loss function, so that the model balances between classification accuracy and working condition feature decoupling. In this embodiment, the model parameters are updated by optimizing the target loss function, and the training is stopped when the loss value meets the preset threshold or the convergence condition, thereby obtaining a classification model that can capture both waveform features and working condition correlations. The preset threshold in this embodiment refers to a value set in advance, which is a constant. For example, the preset threshold is P, and the model stops training when the loss value is less than or equal to P. The convergence condition can be a user-defined condition, for example, the model stops training when the number of training times is greater than R.

[0073] As can be seen from the above, in this embodiment, the model focuses on waveform features under different blowing intervals based on working condition grouping. The similarity constraint loss function forces the features of samples in the same working condition to cluster and the features of samples in different working conditions to separate, thereby strengthening the correlation between working conditions and waveforms. The fusion of the main loss function and the constraint loss function improves the classification accuracy while making the feature space distribution more consistent with the physical logic, avoiding cross-working condition feature confusion, enhancing the model's adaptability and generalization ability to complex working conditions, and providing more reliable model support for accurate classification and fault diagnosis of filter bag operation states.

[0074] In an embodiment of the present application, the dust concentration data is preprocessed to obtain first dust concentration data, including:

[0075] A dust concentration dynamic baseline model is established, which is used to determine a dust concentration reference baseline;

[0076] The dust concentration data is input into the dust concentration dynamic baseline model, and the dynamic standard deviation algorithm is used to identify outliers. The outliers that exceed ±a times the standard deviation in the continuous N sampling periods are replaced by the average value of the dust concentration data in the N+1 sampling period, to obtain the first dust concentration data.

[0077] In the embodiment, the dust concentration dynamic baseline model can provide a real-time updated reference baseline for the dust concentration data. The baseline is not a fixed value, but is dynamically adjusted according to historical data over time, reflecting the fluctuation trend of the dust concentration under normal working conditions. The dynamic standard deviation algorithm is a statistical method for real-time analysis of data volatility. The core is to calculate the standard deviation based on the dynamic baseline of the data, so as to identify the abnormal fluctuations in the data. The continuous N sampling periods are the time window unit of data collection, which is used to avoid false judgments caused by single random noise points. For example, when N=3 is set, only when the data of the continuous 3 periods all exceed the threshold value, it is determined as a real anomaly, not an accidental fluctuation. The ±a times standard deviation is the threshold range for determining abnormal values. a is an adjustable coefficient (such as a=2 or 3), which represents the fluctuation amplitude in units of dynamic standard deviation. The ±a times standard deviation corresponds to the confidence interval of the data (such as a=2 corresponds to 95% confidence interval), and the data exceeding the range is regarded as “abnormal”.

[0078] In the embodiment, when an abnormal value is identified, the abnormal value exceeding ±a times the standard deviation in the continuous N sampling periods can be replaced by the average value of the dust concentration data in the sampling period before and after the abnormal value, for example, the average value of the dust concentration data in the N+1 sampling period. The first dust concentration data obtained in this way is smooth data.

[0079] From the above, it can be concluded that the embodiment can adapt to working condition changes through a dynamic baseline model, accurately identify abnormal values by combining a dynamic standard deviation algorithm, and repair the abnormal values using the average value of the lag period after identifying the abnormal values. The above method can effectively eliminate noise and sudden interference in the dust concentration data, retain the true signal characteristics, provide high-quality data for subsequent feature extraction and classification model training, and enhance the accuracy and robustness of working condition analysis.

[0080] In an embodiment of the present application, based on the fault position of the filter bag, the differential pressure, temperature of the chamber where the filter bag is located, and the material level detection data of the hopper corresponding to the filter bag are used to determine the cause of the filter bag damage, including:

[0081] When the fault position is the bag opening, the differential pressure of the bag opening area is compared with the preset differential pressure threshold to obtain a first comparison result, and the temperature at the bag opening is compared with the temperature at other positions of the filter bag to obtain a second comparison result. Based on the first comparison result and the second comparison result, the cause of the filter bag damage is determined;

[0082] When the fault position is the bag bottom, a first proportion of the material level detection data greater than the height threshold in the target time period is calculated, and a second proportion of the synchronous change time length of the differential pressure and the temperature in the target time period is calculated. Based on the first proportion and the second proportion, the cause of the filter bag damage is determined.

[0083] In the embodiment, when the fault position is the bag opening, firstly, the pressure difference of the bag opening area is compared with the preset threshold value. If the pressure difference exceeds the threshold value, it can be that the bag opening is not tightly sealed, causing airflow short circuit, or the filter bag is blocked, causing increased resistance. If the pressure difference is lower than the threshold value, it can be that the bag opening is damaged, causing air leakage, and the airflow resistance decreases. Secondly, the temperature of the bag opening is compared with the temperature of other positions. If the temperature of the bag opening is significantly higher than the temperature of other positions, it can be that the installation is improper, causing local friction heating, or high-temperature flue gas directly washes the bag opening. If the temperature of the bag opening is significantly lower than the temperature of other positions, it can be that external cold air leaks in (for example, the bag opening is not tightly sealed).

[0084] Therefore, if the pressure difference of the bag opening area in the target time period is less than the preset pressure difference threshold value, and the temperature of the bag opening is significantly lower than the temperature of other positions, the cause of the filter bag damage can be sealing failure. If the pressure difference of the bag opening area in the target time period is greater than the preset pressure difference threshold value, and the temperature of the bag opening is the same as the temperature of other positions of the filter bag, the cause of the filter bag damage can be airflow erosion.

[0085] In the embodiment, when the fault position is the bag bottom, firstly, a first proportion of the material level detection data greater than a height threshold value in a target time period is calculated. If the first proportion is greater than m1, it indicates that the hopper is operated with high material for a long time, and the bag bottom can be damaged due to dust accumulation, friction, or material impact. If the first proportion is less than or equal to m1, the damage cause is irrelevant to the material level. Secondly, a second proportion of the time when the pressure difference and the temperature change synchronously is calculated. If the second proportion is greater than m2, it indicates that the pressure difference change affects the temperature change, or the temperature change affects the pressure difference change, and the cause of the filter bag damage is irrelevant to the pressure difference and the temperature. If the second proportion is less than or equal to m2, it indicates that the pressure difference and the temperature do not change synchronously. At this time, if the pressure difference is abnormal and the first proportion is greater than m1, the cause of the filter bag damage can be dust accumulation at the bag bottom. If the pressure difference and the temperature abnormally fluctuate, and the first proportion is less than or equal to m1, it is determined that the cause of the damage is that the returned dust airflow carries sparks to burn through the bag bottom. In the embodiment, m1 can be any constant between 0.4 and 0.8, and m2 can be any constant between 0.6 and 1.0.

[0086] From the above, it can be concluded that the embodiment designs different analysis indexes for different fault positions of the bag opening and the bag bottom, which can improve the accuracy of damage cause positioning.

[0087] Corresponding to the dust collector filter bag damage detection method of the above embodiment, Figure 2 A structural block diagram of a dust collector filter bag damage detection system according to an embodiment of the present application is provided. For ease of illustration, only parts related to the embodiments of the present application are shown. For reference Figure 2 The dust collector filter bag damage detection system 20 includes a data acquisition module 21, a data analysis module 22, and a damage detection module 23.

[0088] Among them, the data acquisition module 21 is used to acquire dust concentration data and filter bag operating parameter data within the target time period;

[0089] Data analysis module 22 is used to preprocess dust concentration data to obtain first dust concentration data, obtain dust concentration waveform based on first dust concentration data, extract features from dust concentration waveform and determine dust concentration waveform type by combining filter bag operating parameter data;

[0090] The damage detection module 23 is used to determine whether the filter bag is damaged based on the dust concentration waveform type. If the filter bag is damaged, the fault location of the filter bag is located based on the dust concentration waveform type. Based on the fault location of the filter bag, the pressure difference and temperature of the chamber where the filter bag is located, and the material level detection data of the hopper corresponding to the filter bag, the cause of the filter bag damage is determined.

[0091] In one embodiment of this application, the operating parameter data of the filter bag includes the pulse jet interval.

[0092] Data analysis module 22 is specifically used for:

[0093] Feature extraction is performed on the dust concentration waveform, and the dust concentration waveform type is determined by combining it with the filter bag's operating parameter data, including:

[0094] The first feature vector is obtained by performing time-domain feature extraction on the dust concentration waveform, and the second feature vector is obtained by performing frequency-domain feature extraction on the dust concentration waveform.

[0095] The feature weights corresponding to the first feature vector and the second feature vector are adjusted based on the pulse jet interval, and the target feature vector is obtained by weighted fusion based on the first feature vector, the second feature vector and their corresponding feature weights.

[0096] The dust concentration waveform type is obtained based on the target feature vector and classification model.

[0097] In one embodiment of this application, the data analysis module 22 is specifically used for:

[0098] In one embodiment of this application, the peak value, valley value, number of peak values, number of valley values, rise time, fall time, pulse width, and average dust concentration between peak values ​​are extracted from the dust concentration waveform.

[0099] The peak value, valley value, number of peak values, number of valley values, rise time, fall time, pulse width, and average dust concentration between peak values ​​are used as the first feature vector.

[0100] wherein the rise time represents, for any peak point, the time from the adjacent valley point to the peak point; the fall time represents, for any peak point, the time from the peak point to the adjacent valley point; and the pulse width represents the duration of the dust concentration being greater than the target concentration threshold.

[0101] In an embodiment of the present application, the data analysis module 22 is further configured to:

[0102] group the historical dust concentration waveforms based on the pulse injection intervals to obtain a plurality of sets of analysis training sets;

[0103] calculate a main loss function and a similarity constraint loss function based on the plurality of sets of analysis training sets, and fuse the main loss function and the similarity constraint loss function by weighting to obtain a target loss function;

[0104] obtain a trained classification model in response to the target loss function satisfying a preset condition.

[0105] In an embodiment of the present application, the formula of the similarity constraint loss function is:

[0106] .

[0107] wherein i and j represent the numbers of two samples in a selected sample pair in the plurality of sets of analysis training sets, represents the similarity constraint loss value, represents a feature extraction function, represents the dust concentration waveform of the i-th sample, represents the dust concentration waveform of the j-th sample, represents a 2-norm, 、 respectively represent the pulse injection interval of the i-th sample and the pulse injection interval of the j-th sample, represents a working condition attenuation coefficient, represents a pulse injection interval difference penalty function.

[0108] In an embodiment of the present application, the data analysis module 22 is configured to:

[0109] establish a dust concentration dynamic baseline model, the dust concentration dynamic baseline model being configured to determine a dust concentration reference baseline;

[0110] input the dust concentration data into the dust concentration dynamic baseline model, identify abnormal values by using a dynamic standard deviation algorithm, replace the abnormal values exceeding ±a times of the standard deviation in the continuous N sampling periods with the average value of the dust concentration data in the N+1 sampling period, and obtain first dust concentration data.

[0111] In an embodiment of the present application, the damage detection module 23 is specifically configured to:

[0112] When the fault location is at the bag opening, the pressure difference in the bag opening area is compared with a preset pressure difference threshold to obtain a first comparison result, and the temperature at the bag opening is compared with the temperature at other locations on the filter bag to obtain a second comparison result. The cause of the filter bag damage is determined based on the first and second comparison results.

[0113] When the fault location is at the bottom of the bag, calculate the first proportion of the material level detection data exceeding the height threshold within the target time period, and calculate the second proportion of the synchronous change time of pressure difference and temperature within the target time period. Based on the first and second proportions, determine the cause of the filter bag damage.

[0114] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of the modules in the aforementioned system embodiments, for example... Figure 2 The functions of the data acquisition module 21, data analysis module 22, and damage detection module 23 are shown.

[0115] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0116] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0117] The memory 304 can include read-only memory and random access memory, and provide instructions and data to the processor 301. A portion of the memory 304 can also include non-volatile random access memory.

[0118] In a particular implementation, the processor 301, the input device 302, and the output device 303 described in the embodiments of the present application can perform the implementation manners described in the dust collector filter bag breakage detection method provided by the embodiments of the present application, and can also perform the implementation manners of the electronic device described in the embodiments of the present application, which will not be described here.

[0119] In another embodiment of the present application, a computer readable storage medium is provided, which stores a computer program. The computer program includes program instructions, which are executed by a processor to implement all or part of the processes of the above-mentioned embodiments. The computer program can also be used to instruct related hardware to complete the implementation. The computer program can be stored in a computer readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiments can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer readable medium can include any entity or system that can carry the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0120] The computer readable storage medium can be an internal storage unit of the electronic device of any of the above-mentioned embodiments, such as a hard disk or a memory of the electronic device. The computer readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the computer readable storage medium can include both the internal storage unit and the external storage device of the electronic device. The computer readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer readable storage medium can also be used to temporarily store data that has been output or will be output.

[0121] Those skilled in the art can clearly understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware, computer software or a combination of both. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been described in general terms in the above description. Whether the functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0122] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the electronic device and the unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0123] In several embodiments provided in the present application, it should be understood that the disclosed electronic device and method can be implemented in other ways. For example, the system embodiments described above are only schematic, for example, the division of units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be indirect coupling or communication connection through some interfaces or units, and can also be electrical, mechanical or other form of connection.

[0124] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0125] In addition, the functional modules in each embodiment of the present application can be integrated in one processing module, or each module can exist physically, or two or more modules can be integrated in one module. The above integrated module can be realized in the form of hardware or software functional module.

[0126] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for detecting filter bag damage in a dust collector, characterized in that, include: Acquire dust concentration data and filter bag operating parameter data within the target time period; The dust concentration data is preprocessed to obtain the first dust concentration data. Based on the first dust concentration data, the dust concentration waveform is obtained. The dust concentration waveform is feature extracted and combined with the working condition parameter data of the filter bag to determine the dust concentration waveform type. The filter bag is determined to be damaged based on the dust concentration waveform type. If the filter bag is damaged, the fault location of the filter bag is located based on the dust concentration waveform type. The cause of the filter bag damage is determined based on the fault location of the filter bag, the pressure difference and temperature of the chamber where the filter bag is located, and the material level detection data of the hopper corresponding to the filter bag. The operating parameters of the filter bag include the pulse jet interval; The step of extracting features from the dust concentration waveform and determining the dust concentration waveform type by combining it with the filter bag's operating parameter data includes: A first feature vector is obtained by performing time-domain feature extraction on the dust concentration waveform, and a second feature vector is obtained by performing frequency-domain feature extraction on the dust concentration waveform. The feature weights corresponding to the first feature vector and the second feature vector are adjusted based on the pulse jet interval, and the target feature vector is obtained by weighted fusion based on the first feature vector, the second feature vector and their corresponding feature weights. The dust concentration waveform type is obtained based on the target feature vector and classification model; The adjustment of the feature weights corresponding to the first and second feature vectors based on the pulse jet interval can be achieved using the following formula: ； in, This represents the feature weights corresponding to the first feature vector. This represents the feature weights corresponding to the second feature vector. Indicates the pulse jet interval. Represents the baseline value of the time-domain characteristic sensitive interval. Indicates the width of the time-domain feature sensitive interval. This indicates the pulse jet intensity correction factor. Represents the frequency domain attenuation coefficient. Represents the frequency domain gain constant. This indicates the elimination of the zero constant.

2. The dust collector filter bag damage detection method as described in claim 1, characterized in that, The step of extracting the first feature vector from the dust concentration waveform in the time domain includes: Extract the peak value, valley value, number of peak values, number of valley values, rise time, fall time, pulse width, and average dust concentration between peak values ​​from the dust concentration waveform; The peak value, the valley value, the number of peak values, the number of valley values, the rise time, the fall time, the pulse width, and the average dust concentration between peak values ​​are used as the first feature vector. Wherein, the rise time represents the time from the adjacent valley point to the peak point for any given peak point; the fall time represents the time from the peak point to the adjacent valley point for any given peak point; and the pulse width represents the duration during which the dust concentration is greater than the target concentration threshold.

3. The dust collector filter bag damage detection method as described in claim 1, characterized in that, The training process of the classification model includes: Multiple sets of analysis and training data were obtained by grouping historical dust concentration waveforms based on pulse jet intervals. The main loss function and similarity constraint loss function are calculated based on multiple sets of analysis training sets. The main loss function and similarity constraint loss function are then weighted and fused to obtain the target loss function. When the target loss function satisfies the preset conditions, the trained classification model is obtained.

4. The dust collector filter bag damage detection method as described in claim 3, characterized in that, The formula for the similarity constraint loss function is: Where i,j represent the sample numbers of two samples selected from the training set for multiple analysis groups. This represents the loss value due to similarity constraints. Represents the feature extraction function. This represents the dust concentration waveform of the i-th sample. This represents the dust concentration waveform of the j-th sample. Describes the 2-norm. , These represent the pulse jet intervals for the i-th and j-th samples, respectively. Indicates the attenuation coefficient under operating conditions. This represents the penalty function for the difference in pulse jet intervals.

5. The dust collector filter bag damage detection method as described in claim 1, characterized in that, The preprocessing of the dust concentration data to obtain the first dust concentration data includes: A dynamic baseline model for dust concentration is established, which is used to determine a reference baseline for dust concentration. The dust concentration data is input into the dust concentration dynamic baseline model. The dynamic standard deviation algorithm is used to identify outliers. Outliers exceeding ±a times the standard deviation in N consecutive sampling periods are replaced with the average value of the dust concentration data in the N+1th sampling period to obtain the first dust concentration data.

6. The dust collector filter bag damage detection method as described in claim 1, characterized in that, The determination of the cause of filter bag damage based on the fault location of the filter bag, the pressure difference and temperature of the chamber where the filter bag is located, and the material level detection data of the hopper corresponding to the filter bag includes: When the fault location is the bag opening, the pressure difference in the bag opening area is compared with a preset pressure difference threshold to obtain a first comparison result, and the temperature at the bag opening is compared with the temperature at other locations on the filter bag to obtain a second comparison result. The cause of the filter bag damage is determined based on the first comparison result and the second comparison result. When the fault location is at the bottom of the bag, calculate the first proportion of the material level detection data exceeding the height threshold within the target time period, calculate the second proportion of the synchronous change time of pressure difference and temperature within the target time period, and determine the cause of filter bag damage based on the first proportion and the second proportion.

7. A dust collector filter bag damage detection system, characterized in that, include: The data acquisition module is used to acquire dust concentration data and filter bag operating parameter data within a target time period. The data analysis module is used to preprocess the dust concentration data to obtain first dust concentration data, obtain a dust concentration waveform based on the first dust concentration data, extract features from the dust concentration waveform, and determine the dust concentration waveform type by combining the working condition parameter data of the filter bag. The damage detection module is used to determine whether the filter bag is damaged based on the dust concentration waveform type. If the filter bag is damaged, the fault location of the filter bag is located based on the dust concentration waveform type. Based on the fault location of the filter bag, the pressure difference and temperature of the chamber where the filter bag is located, and the material level detection data of the hopper corresponding to the filter bag, the cause of the filter bag damage is determined. The operating parameters of the filter bag include the pulse jet interval; The data analysis module is specifically used for: A first feature vector is obtained by performing time-domain feature extraction on the dust concentration waveform, and a second feature vector is obtained by performing frequency-domain feature extraction on the dust concentration waveform. The feature weights corresponding to the first feature vector and the second feature vector are adjusted based on the pulse jet interval, and the target feature vector is obtained by weighted fusion based on the first feature vector, the second feature vector and their corresponding feature weights. The dust concentration waveform type is obtained based on the target feature vector and classification model; The adjustment of the feature weights corresponding to the first and second feature vectors based on the pulse jet interval can be achieved using the following formula: ； in, This represents the feature weights corresponding to the first feature vector. This represents the feature weights corresponding to the second feature vector. Indicates the pulse jet interval. Represents the baseline value of the time-domain characteristic sensitive interval. Indicates the width of the time-domain feature sensitive interval. This indicates the pulse jet intensity correction factor. Represents the frequency domain attenuation coefficient. Represents the frequency domain gain constant. This indicates the elimination of the zero constant.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images