Nitrogen making machine compressed air fault diagnosis method based on monitoring data
By using a monitoring data-based approach, real-time detection data of the compressed air delivery system of a nitrogen generator is collected and analyzed. Fault diagnosis is optimized by using correlation weights and collaborative stability range, which solves the problems of response lag and missed detection in traditional methods and achieves more accurate fault identification and adaptive diagnosis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUANGGANG GASES (SHENZHEN) CO LTD
- Filing Date
- 2025-12-25
- Publication Date
- 2026-04-24
AI Technical Summary
Traditional nitrogen generator fault diagnosis methods rely on human experience, resulting in slow response and a high rate of missed diagnoses. They also fail to identify the synergistic relationships between parameters, leading to significant discrepancies between diagnostic results and actual faults, especially under complex operating conditions.
Based on monitoring data, real-time detection data of the compressed air delivery system is collected. By extracting correlation weights and collaborative stability ranges from historical data, the fault diagnosis results are optimized. Combined with load fluctuation characteristic coefficients, multi-dimensional fusion diagnosis is achieved.
It improves the comprehensiveness and accuracy of fault diagnosis, reduces the probability of missing latent faults, adapts to different operating conditions, and enhances the consistency of diagnostic results with actual conditions.
Smart Images
Figure CN121916982A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of fault diagnosis technology, and more specifically, to a method for diagnosing compressed air faults in a nitrogen generator based on monitoring data. Background Technology
[0002] The stable operation of a compressed air delivery system directly determines nitrogen purity, gas production efficiency, and production safety. However, in actual operation, this system often malfunctions due to abnormal pressure pulsations, pipeline thermal expansion, and filter blockage. For example, excessive pressure pulsations can cause carbon molecular sieve pulsation, excessive thermal expansion can easily lead to pipeline interface leaks, and filter blockage can cause insufficient gas supply. These malfunctions can range from minor issues like decreased nitrogen purity and insufficient gas production to major issues like production line shutdowns. Traditional nitrogen generator fault diagnosis methods have significant shortcomings, relying on manual experience and on-site inspections. Maintenance personnel need to judge faults by listening for abnormal noises and measuring pressure, which is not only slow to respond but also has a high rate of missing latent faults. Existing methods can monitor single parameters but ignore the synergistic relationship between parameters. For example, situations where a single pressure is normal but thermal expansion and blockage levels are linked and exceed limits are often not identified. Some physical model-based diagnostic methods are difficult to adapt to dynamic operating scenarios due to the complexity of nitrogen generator system conditions, resulting in significant deviations between diagnostic results and actual faults. Summary of the Invention
[0003] To address the shortcomings of existing technologies, the present invention aims to provide a method for diagnosing compressed air faults in nitrogen generators based on monitoring data.
[0004] To achieve the above objectives, the present invention provides the following technical solution:
[0005] A method for diagnosing compressed air faults in a nitrogen generator based on monitoring data, the method comprising the following steps:
[0006] Real-time monitoring data of compressed air in the compressed air delivery system of the nitrogen generator is collected; wherein, the real-time monitoring data includes pressure pulsation frequency, thermal expansion of pipelines, and degree of filter blockage;
[0007] Based on historical monitoring data, the correlation weights between pressure pulsation frequency deviation, excessive thermal expansion value, and blockage degree coefficient and fault risk are extracted; based on the correlation weights and real-time detection data, the single-dimensional fault warning value corresponding to each parameter is calculated.
[0008] Based on historical monitoring data, the coordinated stability range of each monitoring parameter under rated operating conditions is extracted. Based on whether the pressure pulsation frequency, thermal expansion, and blockage degree are within the coordinated stability range, as well as the correlation weight, coordinated stability range, and single-dimensional fault warning value, the initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized to obtain the target fault diagnosis results.
[0009] Output compressed air fault diagnosis and control instructions based on the target fault diagnosis results.
[0010] Preferably, the initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on whether the pressure pulsation frequency, thermal expansion, and blockage degree are within the cooperative stability range, as well as the correlation weight, cooperative stability range, and single-dimensional fault warning value, to obtain the target fault diagnosis results. This specifically includes the following steps:
[0011] If the pressure pulsation frequency, thermal expansion, and blockage degree are all within the cooperative stability range, the initial fault diagnosis result of the nitrogen generator compressed air delivery system is optimized based on the correlation weight, cooperative stability range, and single-dimensional fault warning value to obtain the first fault diagnosis result.
[0012] If the pressure pulsation frequency, thermal expansion, and blockage degree exceed the coordinated stability range, the characteristic coefficient of the load fluctuation affecting the fault is obtained based on the load fluctuation amplitude, fluctuation duration, and parameter out-of-range amplitude. The initial fault diagnosis result of the nitrogen generator compressed air delivery system is optimized based on the characteristic coefficient, correlation weight, and single-dimensional fault warning value to obtain the second fault diagnosis result.
[0013] The first fault diagnosis result and the second fault diagnosis result are combined to form the target fault diagnosis result.
[0014] Preferably, the correlation weights between pressure pulsation frequency deviation, excessive thermal expansion value, and blockage degree coefficient and failure risk are extracted from historical monitoring data, specifically including the following steps:
[0015] Collect the correlation frequency between each historical abnormal parameter and the fault for each fault type;
[0016] The correlation weights of pressure pulsation frequency deviation, thermal expansion excess value, and blockage degree coefficient with fault risk are obtained by comparing the correlation frequency with the average value of the corresponding historical abnormal parameters.
[0017] Preferably, the single-dimensional fault warning value corresponding to each parameter is calculated based on the correlation weight and real-time detection data, specifically including the following steps:
[0018] The pressure pulsation frequency, the thermal expansion of the pipeline, and the degree of filter blockage are processed to obtain the pressure pulsation frequency deviation, the excessive value of thermal expansion, and the blockage degree coefficient.
[0019] The pressure pulsation frequency deviation, thermal expansion excess value, and blockage degree coefficient are multiplied by their corresponding associated weights to obtain the single-dimensional fault warning value for each parameter.
[0020] Preferably, the method further includes the following steps:
[0021] The pressure pulsation frequency, thermal expansion, and degree of blockage were compared with the corresponding dynamic cooperative stability ranges.
[0022] If the pressure pulsation frequency is within its corresponding dynamic coordinated stability range, the thermal expansion is within its corresponding dynamic coordinated stability range, and the blockage degree is within its corresponding dynamic coordinated stability range, then it is determined that the pressure pulsation frequency, thermal expansion, and blockage degree are all within the coordinated stability range.
[0023] If any parameter among pressure pulsation frequency, thermal expansion, and blockage degree exceeds its corresponding dynamic coordinated stability range, then it is determined that the pressure pulsation frequency, thermal expansion, and blockage degree exceed the coordinated stability range.
[0024] Preferably, the initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on the correlation weight, cooperative stability range, and single-dimensional fault warning value to obtain the first fault diagnosis result, specifically including the following steps:
[0025] Based on the parameter failure impact priority represented by the correlation weight, the single-dimensional failure warning value is hierarchically corrected according to the parameter failure impact priority and the boundary margin of the cooperative stability range to obtain the corrected failure warning value.
[0026] A multi-dimensional fusion coefficient is constructed by the coupling correlation between parameters, and the corrected fault warning value and the multi-dimensional fusion coefficient are weighted and integrated to obtain a comprehensive fault warning index.
[0027] By comparing the matching relationship between the comprehensive fault early warning index and the preset early warning threshold, the first fault diagnosis result is obtained based on the matching relationship and the confidence level of each fault type in the initial fault diagnosis result.
[0028] Preferably, the characteristic coefficients of the impact of load fluctuations on faults are obtained based on the load fluctuation amplitude, fluctuation duration, and parameter out-of-range amplitude, specifically including the following steps:
[0029] Extract fault cases under the same load fluctuation scenario from historical monitoring data, and determine the correlation between load fluctuation amplitude, fluctuation duration, parameter out-of-range amplitude and fault status based on the fault cases;
[0030] Fault characteristic coefficients are obtained based on load fluctuation amplitude, fluctuation duration, parameter out-of-range amplitude, and correlation patterns.
[0031] Preferably, the initial fault diagnosis result of the nitrogen generator compressed air delivery system is optimized based on characteristic coefficients, correlation weights, and single-dimensional fault warning values to obtain a second fault diagnosis result, specifically including the following steps:
[0032] The first fault diagnosis result is obtained by summing the single-dimensional fault warning values and multiplying them with the feature coefficients, and then optimizing the initial fault diagnosis result by combining the correlation weights.
[0033] Preferably, the compressed air fault diagnosis and control command is output based on the target fault diagnosis result, specifically including the following steps:
[0034] Plot the spatiotemporal distribution curve of the fault risk corresponding to the first fault diagnosis result or the second fault diagnosis result;
[0035] Marking the spatiotemporal distribution curve of fault risk yields marked risk information; wherein, the marked risk information includes the location and time period with the highest fault warning value, and the pipeline path with the rate of increase of the warning value;
[0036] Combine the marked risk information and fault range to output compressed air fault diagnosis and prevention instructions.
[0037] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements a method for diagnosing compressed air faults in a nitrogen generator based on monitoring data.
[0038] Compared with the prior art, the present invention has the following beneficial effects:
[0039] This invention improves the comprehensiveness and accuracy of fault diagnosis by selecting pressure pulsation frequency, pipeline thermal expansion, and filter clogging degree as monitoring objects, covering key dimensions of pressure stability, pipeline safety, and airflow unobstructedness in the compressed air system of a nitrogen generator, avoiding the one-sidedness of monitoring a single parameter. Simultaneously, by extracting correlation weights from historical data, the impact of different parameters on faults is quantified and differentiated. For example, pressure pulsation frequency has a higher weighting on molecular sieve damage, and the diagnosis prioritizes focusing on anomalies in this parameter, reducing misjudgments caused by treating all parameters equally. The introduction of a coordinated stability range takes into account the linkage relationship between parameters. Even if a single parameter is within the normal range, if multiple parameters exceed the coordinated range, it can still be identified as a potential risk, significantly reducing the probability of missing latent faults. Differentiated optimization is performed based on whether the parameter is within the coordinated stability range for stable and out-of-limit scenarios: when parameters are stable, hierarchical correction and multi-dimensional fusion improve diagnostic accuracy; when parameters exceed limits, the diagnostic results are adjusted by combining the characteristic coefficients of load fluctuations, allowing the diagnostic logic to adapt to different operating conditions of the nitrogen generator, making the diagnostic results more consistent with the actual operating state. Attached Figure Description
[0040] Figure 1 This is a schematic diagram illustrating a nitrogen generator compressed air fault diagnosis method based on monitoring data proposed in this invention;
[0041] Figure 2 This is a schematic diagram illustrating a nitrogen generator compressed air fault diagnosis method based on monitoring data proposed in this invention;
[0042] Figure 3 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention.
[0043] 610. Processor; 620. Communication interface; 630. Memory; 640. Communication bus. Detailed Implementation
[0044] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.
[0045] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.
[0046] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.
[0047] Reference Figures 1-3 As shown.
[0048] The embodiments further illustrate the method for diagnosing compressed air faults in nitrogen generators based on monitoring data proposed in this invention.
[0049] A method for diagnosing compressed air faults in a nitrogen generator based on monitoring data, the method comprising the following steps:
[0050] Real-time monitoring data of compressed air in the compressed air delivery system of the nitrogen generator is collected; the real-time monitoring data includes pressure pulsation frequency, thermal expansion of pipelines, and degree of filter blockage.
[0051] Based on historical monitoring data, the correlation weights between pressure pulsation frequency deviation, excessive thermal expansion value, and blockage degree coefficient and fault risk are extracted; based on the correlation weights and real-time detection data, the single-dimensional fault warning value corresponding to each parameter is calculated.
[0052] Based on historical monitoring data, the coordinated stability range of each monitoring parameter under rated operating conditions is extracted. Based on whether the pressure pulsation frequency, thermal expansion, and blockage degree are within the coordinated stability range, as well as the correlation weight, coordinated stability range, and single-dimensional fault warning value, the initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized to obtain the target fault diagnosis results.
[0053] Output compressed air fault diagnosis and control instructions based on the target fault diagnosis results.
[0054] The initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on whether the pressure pulsation frequency, thermal expansion, and blockage degree are within the cooperative stability range, as well as the correlation weight, cooperative stability range, and single-dimensional fault warning value, to obtain the target fault diagnosis results. This optimization process includes the following steps:
[0055] If the pressure pulsation frequency, thermal expansion, and blockage degree are all within the cooperative stability range, the initial fault diagnosis result of the nitrogen generator compressed air delivery system is optimized based on the correlation weight, cooperative stability range, and single-dimensional fault warning value to obtain the first fault diagnosis result.
[0056] If the pressure pulsation frequency, thermal expansion, and blockage degree exceed the coordinated stability range, the characteristic coefficient of the load fluctuation affecting the fault is obtained based on the load fluctuation amplitude, fluctuation duration, and parameter out-of-range amplitude. The initial fault diagnosis result of the nitrogen generator compressed air delivery system is optimized based on the characteristic coefficient, correlation weight, and single-dimensional fault warning value to obtain the second fault diagnosis result.
[0057] The first fault diagnosis result and the second fault diagnosis result are combined to form the target fault diagnosis result.
[0058] Based on historical monitoring data, the correlation weights between pressure pulsation frequency deviation, excessive thermal expansion, and blockage degree coefficient and failure risk are extracted, specifically including the following steps:
[0059] Collect the correlation frequency between each historical abnormal parameter and the fault for each fault type;
[0060] The correlation weights of pressure pulsation frequency deviation, thermal expansion excess value, and blockage degree coefficient with fault risk are obtained by comparing the correlation frequency with the average value of the corresponding historical abnormal parameters.
[0061] The correlation frequency between historical abnormal parameters and faults for each fault type is collected. Fault types can include pipeline leaks and insufficient gas supply. The correlation frequency refers to the actual number of times a fault occurs after a certain parameter becomes abnormal. For example, in the pipeline leak fault type, the number of times pipeline leaks occurred after an abnormal pressure pulsation frequency deviation in historical data is the correlation frequency between pressure pulsation frequency deviation and pipeline leak faults. Similarly, the correlation frequencies between excessive thermal expansion and blockage coefficient and these fault types are obtained.
[0062] The correlation frequency of each parameter is compared with the average value of the corresponding historical abnormal parameter. This ratio represents the correlation weight between that parameter and the failure risk. For example, to calculate the average value of the pressure pulsation frequency deviation in historical abnormal data for pipeline leakage failures, the correlation frequency between pressure pulsation frequency deviation and pipeline leakage is divided by the average value to obtain the correlation weight between pressure pulsation frequency deviation and pipeline leakage failure risk. Similarly, the correlation frequency of thermal expansion exceeding the standard value is divided by the average value of its historical abnormal parameters to obtain the correlation weight corresponding to the thermal expansion exceeding the standard value. The correlation weight of the blockage degree coefficient is also obtained through the same calculation method.
[0063] This allows the influence of each parameter on different fault types to be quantified in the form of associated weights. The higher the weight value, the greater the probability or degree of impact of the corresponding fault when the parameter is abnormal.
[0064] The single-dimensional fault warning value corresponding to each parameter is calculated based on the correlation weight and real-time detection data. The specific steps include:
[0065] The pressure pulsation frequency, the thermal expansion of the pipeline, and the degree of filter blockage are processed to obtain the pressure pulsation frequency deviation, the excessive value of thermal expansion, and the blockage degree coefficient.
[0066] The pressure pulsation frequency deviation, thermal expansion excess value, and blockage degree coefficient are multiplied by their corresponding associated weights to obtain the single-dimensional fault warning value for each parameter.
[0067] The first step is to preprocess the real-time monitoring data, which involves processing the collected pressure pulsation frequency, pipeline thermal expansion, and filter clogging degree into corresponding fault-related parameters. Taking pressure pulsation frequency as an example, the pressure pulsation frequency deviation is calculated by subtracting the standard value of the parameter under rated operating conditions from the real-time collected pressure pulsation frequency value. For pipeline thermal expansion, the excess value is calculated by subtracting the maximum allowable thermal expansion under rated operating conditions from the real-time thermal expansion value. If the real-time value does not exceed the standard, the excess value is 0; otherwise, the actual excess value is used. Typically, the clogging degree coefficient is obtained by dividing the real-time clogging degree by the allowable clogging threshold under rated operating conditions. If the rated standard value of pressure pulsation frequency is 5Hz and the real-time acquisition value is 7Hz, then the pressure pulsation frequency deviation is 7 minus 5 equals 2Hz; if the maximum allowable value of pipeline thermal expansion is 0.8mm and the real-time value is 1.0mm, then the thermal expansion exceeding the standard value is 1.0 minus 0.8 equals 0.2mm; if the proportion corresponding to the filter's allowable clogging threshold is 30% and the real-time clogging degree is 20%, then the clogging degree coefficient is 20% divided by 30%, approximately equal to 0.67.
[0068] The method for calculating the single-dimensional fault warning value for each parameter is to multiply the pre-processed pressure pulsation frequency deviation, thermal expansion excess value, and blockage degree coefficient by their respective correlation weights. If the correlation weight for the pressure pulsation frequency deviation is 0.4, then the single-dimensional fault warning value for that parameter is the pressure pulsation frequency deviation multiplied by 0.4; if the correlation weight for the thermal expansion excess value is 0.3, then the single-dimensional fault warning value is the thermal expansion excess value multiplied by 0.3; and if the correlation weight for the blockage degree coefficient is 0.3, then the corresponding single-dimensional fault warning value is the blockage degree coefficient multiplied by 0.3.
[0069] It also includes the following steps:
[0070] The pressure pulsation frequency, thermal expansion, and degree of blockage were compared with the corresponding dynamic cooperative stability ranges.
[0071] If the pressure pulsation frequency is within its corresponding dynamic coordinated stability range, the thermal expansion is within its corresponding dynamic coordinated stability range, and the blockage degree is within its corresponding dynamic coordinated stability range, then it is determined that the pressure pulsation frequency, thermal expansion, and blockage degree are all within the coordinated stability range.
[0072] If any parameter among pressure pulsation frequency, thermal expansion, and blockage degree exceeds its corresponding dynamic coordinated stability range, then it is determined that the pressure pulsation frequency, thermal expansion, and blockage degree exceed the coordinated stability range.
[0073] First, the real-time collected pressure pulsation frequency, thermal expansion, and blockage degree are compared one by one with their corresponding dynamic coordinated stability ranges. These dynamic coordinated stability ranges are not fixed numerical ranges, but rather reasonable fluctuation ranges of parameters adapted from historical stable operating data, taking into account the actual operating conditions of the nitrogen generator under current load and ambient temperature. For example, the dynamic coordinated stability range for pressure pulsation frequency is 4Hz to 6Hz under low load conditions; under high load conditions, this range is adjusted to 5Hz to 7Hz; the range for thermal expansion changes with ambient temperature, and the range for blockage degree is dynamically updated with operating time.
[0074] The system's stability is determined by matching the three parameters with their corresponding intervals. The first scenario is full parameter matching: if the real-time value of the pressure pulsation frequency falls within its corresponding dynamic stable interval, and the real-time value of the thermal expansion also falls within its own dynamic stable interval, and the real-time value of the blockage degree is also within its corresponding interval, then these three parameters are considered to be within a coordinated stable range. For example, if the current pressure pulsation frequency is 5Hz (range 4Hz-6Hz), the thermal expansion is 0.5mm (range 0.3mm-0.7mm), and the blockage degree is 15% (range 10%-20%), and all three parameters are within their respective intervals, then the system is in a coordinated stable state.
[0075] The second scenario involves any parameter exceeding its limit. If the real-time value of any one of the parameters—pressure pulsation frequency, thermal expansion, or blockage degree—exceeds its corresponding dynamic coordinated stability range, the system is deemed to be outside the coordinated stability range, regardless of whether other parameters are within their respective ranges. For example, if the pressure pulsation frequency is 7.5Hz (exceeding the range of 4Hz-6Hz), even if the thermal expansion of 0.5mm and the blockage degree of 15% are both within their ranges, it will still be considered outside the coordinated stability range. Similarly, if the thermal expansion is 0.8mm (exceeding the range of 0.3mm-0.7mm), it will also trigger an over-limit determination.
[0076] This judgment method considers both the independent rationality of each parameter and emphasizes the coordinated and stable relationship between parameters during system operation, providing a clear preliminary judgment basis for subsequent scenario-based optimization of fault diagnosis results.
[0077] The initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on correlation weights, collaborative stability ranges, and single-dimensional fault warning values to obtain the first fault diagnosis result. This optimization includes the following steps:
[0078] Based on the parameter failure impact priority represented by the correlation weight, the single-dimensional failure warning value is hierarchically corrected according to the parameter failure impact priority and the boundary margin of the cooperative stability range to obtain the corrected failure warning value.
[0079] A multi-dimensional fusion coefficient is constructed by the coupling correlation between parameters, and the corrected fault warning value and the multi-dimensional fusion coefficient are weighted and integrated to obtain a comprehensive fault warning index.
[0080] By comparing the matching relationship between the comprehensive fault early warning index and the preset early warning threshold, the first fault diagnosis result is obtained based on the matching relationship and the confidence level of each fault type in the initial fault diagnosis result.
[0081] Based on the parameter fault impact priority represented by correlation weights, and combined with the boundary margin of the cooperative stability range, the single-dimensional fault warning value is hierarchically corrected to obtain the corrected fault warning value. The parameter fault impact priority is determined by the correlation weight; the higher the correlation weight, the higher the priority of the parameter's impact on the fault. The boundary margin refers to the distance between the parameter's real-time value and the boundary of the cooperative stability interval. For example, the closer the parameter's real-time value is to the upper limit of the interval, the smaller the boundary margin. For parameters with higher priority, if their boundary margin is smaller, the correction magnitude of their single-dimensional fault warning value will be greater. For example, the correlation weight of pressure pulsation frequency is 0.4, its boundary margin from the upper limit of the cooperative stability interval is 0.2, and the corresponding correction coefficient is 1.2; the correlation weight of thermal expansion is 0.3, its boundary margin is 0.5, and the correction coefficient is 1.0; the correlation weight of blockage degree is 0.3, its boundary margin is 0.6, and the correction coefficient is 0.9. At this point, the corrected fault warning value for pressure pulsation frequency is its single-dimensional fault warning value multiplied by 1.2, the corrected value for thermal expansion is the single-dimensional value multiplied by 1.0, and the corrected value for blockage degree is the single-dimensional value multiplied by 0.9.
[0082] A multi-dimensional fusion coefficient is constructed by leveraging the coupling correlation between parameters. The corrected fault warning value and the multi-dimensional fusion coefficient are then weighted and integrated to obtain a comprehensive fault warning index. The coupling correlation between parameters refers to the degree of correlation between abnormal parameters. For example, a high pressure pulsation frequency is often accompanied by an increase in thermal expansion; the two have a high coupling correlation. The construction of the fusion coefficient needs to consider the frequency of parameter linkage anomalies in historical data; the higher the coupling correlation, the greater the weight of the corresponding fusion coefficient. For example, the coupling correlation between pressure pulsation frequency and thermal expansion is 0.8, the coupling correlation between thermal expansion and blockage degree is 0.5, and the coupling correlation between pressure pulsation frequency and blockage degree is 0.6. Therefore, the multi-dimensional fusion coefficient is calculated as (0.8×0.4+0.5×0.3+0.6×0.3)=0.67. The corrected fault warning values of each parameter are summed, and then multiplied by the multi-dimensional fusion coefficient to obtain the comprehensive fault warning index. For example, if the sum of the corrected fault warning values is 5, the comprehensive index is 5×0.67=3.35.
[0083] The matching relationship between the comprehensive fault warning index and the preset warning threshold is compared, and the first fault diagnosis result is obtained by combining the confidence levels of each fault type in the initial fault diagnosis results. The preset warning thresholds are usually divided into multiple levels, such as low-risk threshold 2, medium-risk threshold 4, and high-risk threshold 6. If the comprehensive fault warning index is 3.35, it matches the medium-risk threshold range. Meanwhile, the confidence level of loose pipe connections in the initial fault diagnosis results is 0.7, and the confidence level of initial filter blockage is 0.4. At this point, combining the risk level of the comprehensive index, the judgment weight of high-confidence fault types is increased, ultimately obtaining the first fault diagnosis result. For example, if the fault is judged as medium-risk (loose pipe connections), the confidence level is adjusted to 0.8.
[0084] The characteristic coefficients of the impact of load fluctuations on faults are obtained based on the load fluctuation amplitude, fluctuation duration, and parameter out-of-range amplitude. The specific steps include:
[0085] Extract fault cases under the same load fluctuation scenario from historical monitoring data, and determine the correlation between load fluctuation amplitude, fluctuation duration, parameter out-of-range amplitude and fault status based on the fault cases;
[0086] Fault characteristic coefficients are obtained based on load fluctuation amplitude, fluctuation duration, parameter out-of-range amplitude, and correlation patterns.
[0087] First, fault cases under the same load fluctuation scenario are extracted from historical monitoring data. Based on these cases, the correlation between load fluctuation amplitude, duration of fluctuation, out-of-range parameter magnitude, and fault condition is determined. The same load fluctuation scenario refers to a consistent type of load fluctuation, such as a sudden 30% load increase. The fault cases are actual fault records occurring under this scenario. By statistically analyzing multiple sets of cases, correlation patterns can be summarized. For example, in a scenario of a sudden 30% load increase, if the load fluctuation amplitude exceeds 20% and the duration exceeds 5 minutes, and the parameter out-of-range magnitude exceeds 15%, then 80% of the cases will result in a sudden increase in pipeline pressure and leakage. If the fluctuation amplitude is between 10% and 20% and the duration is less than 3 minutes, and the parameter out-of-range magnitude is less than 10%, then the fault incidence rate is only 10%. Combinations of load fluctuation amplitude, duration, and out-of-range magnitude correspond to different probabilities or severity of fault occurrence.
[0088] The fault characteristic coefficient is obtained by combining the out-of-range magnitude of load fluctuations and their duration with the correlation pattern. This coefficient quantifies the degree of fault risk under the current load fluctuation conditions and needs to be calculated by combining the weights of each condition in the correlation pattern. For example, based on the above correlation pattern, the weight of load fluctuation magnitude is set to 0.4, the weight of fluctuation duration is 0.3, and the weight of parameter out-of-range magnitude is 0.3; if the current load fluctuation magnitude is 25%, corresponding to the high-risk interval in the pattern, it is assigned a value of 1.2; the fluctuation duration is 6 minutes, assigned a value of 1.1; and the parameter out-of-range magnitude is 18%, assigned a value of 1.3. Then the fault characteristic coefficient is 25%×0.4×1.2+6 minutes×0.3×1.1+18%×0.3×1.3. In actual calculation, each parameter needs to be normalized to the same dimension, such as converting time into a relative percentage of fluctuation duration. The higher the final fault characteristic coefficient value, the greater the impact of fault risk under the current load fluctuation scenario.
[0089] The initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on characteristic coefficients, correlation weights, and single-dimensional fault warning values to obtain a second fault diagnosis result. This optimization process includes the following steps:
[0090] The first fault diagnosis result is obtained by summing the single-dimensional fault warning values and multiplying them with the feature coefficients, and then optimizing the initial fault diagnosis result by combining the correlation weights.
[0091] First, it is necessary to calculate the sum of the single-dimensional fault warning values. Here, the single-dimensional fault warning value is the independent warning value of each parameter obtained by multiplying the parameter preprocessing and the associated weights. For example, the single-dimensional fault warning value of pressure pulsation frequency is 2, the single-dimensional fault warning value of thermal expansion is 1.5, and the single-dimensional fault warning value of blockage degree is 1. The sum of the three is 2 plus 1.5 plus 1 equals 4.5.
[0092] The summation result is then multiplied by the fault characteristic coefficient. The fault characteristic coefficient is a quantified value calculated based on the load fluctuation amplitude, fluctuation duration, and parameter out-of-range amplitude. It represents the degree of impact of the current load fluctuation scenario on the fault. For example, if the fault characteristic coefficient in the current scenario is 1.2, then the result after multiplication is 4.5 multiplied by 1.2 equals 5.4.
[0093] The initial fault diagnosis results are optimized by combining correlation weights to obtain the second fault diagnosis results. Correlation weights characterize the priority of each parameter's influence on the fault. At this point, the confidence level of each fault type in the initial fault diagnosis results needs to be adjusted according to the proportion of the correlation weights, based on the sum of the single-dimensional warning values multiplied by the feature coefficients. For example, in the initial fault diagnosis results, the confidence level of abnormal pipeline pressure is 0.6, and the confidence level of worsening filter blockage is 0.5. However, the correlation weight of pressure pulsation frequency (0.4) is higher than the correlation weight of blockage degree (0.3). By increasing the confidence level of abnormal pipeline pressure and combining it with the calculation results in section 5.4, the final confidence level of abnormal pipeline pressure is corrected to 0.8, and the confidence level of worsening filter blockage is corrected to 0.6. The resulting fault judgment, including the corrected confidence levels, is the second fault diagnosis result.
[0094] Based on the target fault diagnosis results, output compressed air fault diagnosis and control instructions, specifically including the following steps:
[0095] Plot the spatiotemporal distribution curve of the fault risk corresponding to the first fault diagnosis result or the second fault diagnosis result;
[0096] Marking the spatiotemporal distribution curve of fault risk yields marked risk information; the marked risk information includes the location and time period of the highest fault warning value, as well as the pipeline path of the warning value growth rate;
[0097] Combine the marked risk information and fault range to output compressed air fault diagnosis and prevention instructions.
[0098] First, a spatiotemporal distribution curve of the fault risk corresponding to the first or second fault diagnosis result is plotted. The spatiotemporal distribution curve is a visual chart combining time and space dimensions. The time dimension corresponds to different periods of nitrogen generator operation, and the space dimension corresponds to different pipeline locations in the compressed air delivery system. The vertical axis of the curve represents the level of the fault warning value. For example, the curve for the second fault diagnosis result shows the trend of the fault warning value of pipeline section A increasing from 2 to 4 during the period 08:00-09:00; and the warning value of pipeline section B increasing from 3 to 5 during the period 10:00-11:00, intuitively presenting the distribution of fault risk at different times and locations.
[0099] The risk information is obtained by marking the spatiotemporal distribution curve of failure risk. The markings include the location with the highest failure warning value, such as pipeline section B with a warning value of 5; the time period with the highest warning value, such as 10:00-11:00; and the pipeline path with the fastest rate of warning value increase, such as pipeline section C where the warning value rises from 1 to 4 within one hour, a rate of increase higher than other pipelines. This marking information extracts the key risk points from the curve, clarifying the core areas and pace of change of failure risk.
[0100] Combining risk information and fault scope, the system outputs compressed air fault diagnosis and control instructions. The fault scope refers to the area of pipeline or equipment affected by the fault, such as the adsorption tower unit associated with pipeline section B. The control instructions will formulate targeted measures based on the marked information. For example, for a high warning value in pipeline section B between 10:00 and 11:00, the instruction requires a pressure inspection of pipeline section B before 10:00; for the rapid increase rate in pipeline section C, the instruction requires increasing the real-time monitoring frequency of pipeline section C; and simultaneously, based on the fault scope, supplementary control requirements include suspending high-load operation of the adsorption tower unit.
[0101] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for diagnosing compressed air faults in a nitrogen generator based on monitoring data.
[0102] like Figure 3 As shown, the electronic device may include a processor 610, a communication interface 620, a memory 630, and a communication bus 640, wherein the processor 610, the communication interface 620, and the memory 630 communicate with each other through the communication bus 640. The processor 610 can call logical instructions in the memory 630 to execute a nitrogen generator compressed air fault diagnosis method based on monitoring data.
[0103] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.
[0104] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program that can be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer is able to execute a nitrogen generator compressed air fault diagnosis method based on monitoring data.
[0105] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform a method for diagnosing compressed air faults in a nitrogen generator based on monitoring data.
[0106] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0107] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.
[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for diagnosing compressed air faults in a nitrogen generator based on monitoring data, characterized in that, The method includes the following steps: Real-time monitoring data of compressed air in the compressed air delivery system of the nitrogen generator is collected; wherein, the real-time monitoring data includes pressure pulsation frequency, thermal expansion of pipelines, and degree of filter blockage; Based on historical monitoring data, the correlation weights between pressure pulsation frequency deviation, excessive thermal expansion value, and blockage degree coefficient and fault risk are extracted; based on the correlation weights and real-time detection data, the single-dimensional fault warning value corresponding to each parameter is calculated. Based on historical monitoring data, the coordinated stability range of each monitoring parameter under rated operating conditions is extracted. Based on whether the pressure pulsation frequency, thermal expansion, and blockage degree are within the coordinated stability range, as well as the correlation weight, coordinated stability range, and single-dimensional fault warning value, the initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized to obtain the target fault diagnosis results. Output compressed air fault diagnosis and control instructions based on the target fault diagnosis results.
2. The method for diagnosing compressed air faults in a nitrogen generator based on monitoring data according to claim 1, characterized in that, The initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on whether the pressure pulsation frequency, thermal expansion, and blockage degree are within the cooperative stability range, as well as the correlation weight, cooperative stability range, and single-dimensional fault warning value, to obtain the target fault diagnosis results. This optimization process includes the following steps: If the pressure pulsation frequency, thermal expansion, and blockage degree are all within the cooperative stability range, the initial fault diagnosis result of the nitrogen generator compressed air delivery system is optimized based on the correlation weight, cooperative stability range, and single-dimensional fault warning value to obtain the first fault diagnosis result. If the pressure pulsation frequency, thermal expansion, and blockage degree exceed the coordinated stability range, the characteristic coefficient of the load fluctuation affecting the fault is obtained based on the load fluctuation amplitude, fluctuation duration, and parameter out-of-range amplitude. The initial fault diagnosis result of the nitrogen generator compressed air delivery system is optimized based on the characteristic coefficient, correlation weight, and single-dimensional fault warning value to obtain the second fault diagnosis result. The first fault diagnosis result and the second fault diagnosis result are combined to form the target fault diagnosis result.
3. The method for diagnosing compressed air faults in a nitrogen generator based on monitoring data according to claim 1, characterized in that, Based on historical monitoring data, the correlation weights between pressure pulsation frequency deviation, excessive thermal expansion, and blockage degree coefficient and failure risk are extracted, specifically including the following steps: Collect the correlation frequency between each historical abnormal parameter and the fault for each fault type; The correlation weights of pressure pulsation frequency deviation, thermal expansion excess value, and blockage degree coefficient with fault risk are obtained by comparing the correlation frequency with the average value of the corresponding historical abnormal parameters.
4. The method for diagnosing compressed air faults in a nitrogen generator based on monitoring data according to claim 1, characterized in that, The single-dimensional fault warning value corresponding to each parameter is calculated based on the correlation weight and real-time detection data. The specific steps include: The pressure pulsation frequency, the thermal expansion of the pipeline, and the degree of filter blockage are processed to obtain the pressure pulsation frequency deviation, the excessive value of thermal expansion, and the blockage degree coefficient. The pressure pulsation frequency deviation, thermal expansion excess value, and blockage degree coefficient are multiplied by their corresponding associated weights to obtain the single-dimensional fault warning value for each parameter.
5. The method for diagnosing compressed air faults in a nitrogen generator based on monitoring data according to claim 1, characterized in that, It also includes the following steps: The pressure pulsation frequency, thermal expansion, and degree of blockage were compared with the corresponding dynamic cooperative stability ranges. If the pressure pulsation frequency is within its corresponding dynamic coordinated stability range, the thermal expansion is within its corresponding dynamic coordinated stability range, and the blockage degree is within its corresponding dynamic coordinated stability range, then it is determined that the pressure pulsation frequency, thermal expansion, and blockage degree are all within the coordinated stability range. If any parameter among pressure pulsation frequency, thermal expansion, and blockage degree exceeds its corresponding dynamic coordinated stability range, then it is determined that the pressure pulsation frequency, thermal expansion, and blockage degree exceed the coordinated stability range.
6. The method for diagnosing compressed air faults in a nitrogen generator based on monitoring data according to claim 2, characterized in that, The initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on correlation weights, collaborative stability ranges, and single-dimensional fault warning values to obtain the first fault diagnosis result. This optimization includes the following steps: Based on the parameter fault impact priority represented by the correlation weight, the single-dimensional fault warning value is hierarchically corrected according to the parameter fault impact priority and the boundary margin of the cooperative stability range to obtain the corrected fault warning value. A multi-dimensional fusion coefficient is constructed by the coupling correlation between parameters, and the corrected fault warning value and the multi-dimensional fusion coefficient are weighted and integrated to obtain a comprehensive fault warning index. By comparing the matching relationship between the comprehensive fault early warning index and the preset early warning threshold, the first fault diagnosis result is obtained based on the matching relationship and the confidence level of each fault type in the initial fault diagnosis result.
7. The method for diagnosing compressed air faults in a nitrogen generator based on monitoring data according to claim 6, characterized in that, The characteristic coefficients of the impact of load fluctuations on faults are obtained based on the load fluctuation amplitude, fluctuation duration, and parameter out-of-range amplitude. The specific steps include: Extract fault cases under the same load fluctuation scenario from historical monitoring data, and determine the correlation between load fluctuation amplitude, fluctuation duration, parameter out-of-range amplitude and fault status based on the fault cases; Fault characteristic coefficients are obtained based on load fluctuation amplitude, fluctuation duration, parameter out-of-range amplitude, and correlation patterns.
8. The method for diagnosing compressed air faults in a nitrogen generator based on monitoring data according to claim 7, characterized in that, The initial fault diagnosis results of the nitrogen generator compressed air delivery system are optimized based on characteristic coefficients, correlation weights, and single-dimensional fault warning values to obtain a second fault diagnosis result. This optimization process includes the following steps: The first fault diagnosis result is obtained by summing the single-dimensional fault warning values and multiplying them with the feature coefficients, and then combining the correlation weights to optimize the initial fault diagnosis result.
9. A method for diagnosing compressed air faults in a nitrogen generator based on monitoring data, as described in claim 8, is characterized in that... Based on the target fault diagnosis results, output compressed air fault diagnosis and control instructions, specifically including the following steps: Plot the spatiotemporal distribution curve of the fault risk corresponding to the first fault diagnosis result or the second fault diagnosis result; Marking risk information is obtained by marking the spatiotemporal distribution curve of fault risk; wherein, the marked risk information includes the location and time period of the highest fault warning value and the pipeline path of the warning value growth rate; Combine the marked risk information and fault range to output compressed air fault diagnosis and prevention instructions.
10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the nitrogen generator compressed air fault diagnosis method based on monitoring data as described in any one of claims 1 to 9.