A method for diagnosing degradation faults in desulfurization towers during natural gas purification processes.

Abstract

This invention relates to the field of natural gas processing equipment technology, and particularly to a method for diagnosing degradation faults in desulfurization towers during natural gas purification processes. The method comprises nine steps: multi-source synchronous acquisition of desulfurization tower operating information; time-consistency processing of operating information; adaptive fuzzy operating condition identification; equipment functional state construction oriented towards process functional semantics; function-process decoupling based on process constraints; establishment and deviation measurement of functional residual benchmarks; high-time-sensitivity determination oriented towards slow degradation evolution; degradation fault type diagnosis based on functional change structures; and online reconstruction and closed-loop self-calibration of functional benchmarks. This invention provides a desulfurization tower degradation fault diagnosis method capable of establishing stable functional benchmarks under continuously changing multi-condition conditions and effectively decoupling between equipment functional changes and process disturbances in strongly coupled environments. This improves the stability and accuracy of degradation diagnosis under long-term operating conditions, thereby enhancing the safety and reliability of natural gas processing systems.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of natural gas treatment devices, and particularly relates to a degradation fault diagnosis method for a desulfurization tower in a natural gas purification process. BACKGROUND

[0002] A natural gas treatment plant usually adopts an absorption-regeneration process to purify raw natural gas by desulfurization and decarbonization. A typical natural gas purification production line includes, from the upstream raw gas inlet, a raw gas pretreatment unit, a desulfurization tower unit, a rich liquid flash evaporation and heat exchange unit, a regeneration tower unit, a solvent circulation system and a product gas conveying unit, and constitutes a long continuous process production line.

[0003] In such a natural gas purification system, the desulfurization tower is a key device that determines the purification effect of natural gas, and its main function is to use an absorption solvent to absorb and remove hydrogen sulfide, carbon dioxide and other acidic components in raw natural gas, thereby achieving natural gas purification. In actual operation, the purification capacity of the desulfurization tower is not only affected by the structure of the device itself and the operating conditions, but also affected by multiple process factors such as raw gas flow, raw gas acidic component concentration, solvent circulation flow and downstream load changes. Since the natural gas treatment device is a typical continuous long process industrial system, the coupling degree between units is high, and upstream and downstream disturbances are frequent, and the running state of the desulfurization tower often fluctuates dynamically with changes in process conditions.

[0004] In the long-term operation, the desulfurization tower is prone to degradation problems such as packing contamination, deterioration of packing wetting state, uneven liquid distribution or mass transfer efficiency reduction. These degradation phenomena will gradually weaken the mass transfer capacity of the desulfurization tower, making the removal efficiency of acidic components gradually decrease, resulting in weakening of the purification capacity of natural gas. Unlike sudden equipment failure, desulfurization tower degradation usually presents a slow evolution characteristic, and in the early stage, it often does not directly cause the over-limit of single operating parameters such as pressure, temperature and flow rate, but shows functional changes such as gradual decrease of purification efficiency, weakening of purification capacity per unit of processing capacity or decrease of running stability, and has characteristics of slow evolution, strong concealment and difficulty in direct identification by single parameter threshold.

[0005] Currently, there are two main types of methods for monitoring the operation of natural gas processing units. One type is based on alarms for exceeding limits of a single operating parameter. This method monitors key parameters such as pressure, temperature, and flow rate by setting fixed thresholds, triggering an alarm when the parameter exceeds the set threshold. While simple in structure and easy to implement, this method can only identify obvious faults or sudden anomalies. It often lags significantly in detecting slowly evolving degradation faults such as desulfurization tower packing contamination or decreased mass transfer capacity, making it difficult to detect equipment functional degradation in a timely manner. The other type of method is based on data-driven models or statistical analysis to detect anomalies in equipment operating data. This method establishes a normal behavior model of the equipment by analyzing historical operating data, and determines equipment anomalies when the real-time operating state deviates from the normal model. However, in long-process natural gas processing systems, the desulfurization tower's operating state is simultaneously affected by multiple process disturbances, resulting in significant differences in the statistical characteristics of operating parameters under different load conditions. If a unified diagnostic model is established directly without distinguishing operating conditions, normal process fluctuations can easily be misjudged as equipment anomalies, or process disturbances can mask changes in the equipment's own performance, making it difficult to identify desulfurization tower degradation characteristics, thus reducing diagnostic accuracy.

[0006] In long-process coupled natural gas systems, changes in the functional status of desulfurization towers typically involve two categories of influencing factors: normal process responses caused by changes in feed gas flow rate, gas quality, or downstream load, and functional degradation caused by equipment degradation itself, such as desulfurization tower packing contamination, solvent performance degradation, or decreased mass transfer efficiency. If these two sources of change cannot be effectively distinguished, it is difficult to accurately determine whether the desulfurization tower has experienced a degradation failure.

[0007] Therefore, there is an urgent need for a desulfurization tower degradation fault diagnosis method that can establish a stable functional benchmark under continuous changes in multiple operating conditions and effectively decouple equipment function changes and process disturbances in a strongly coupled process environment, so as to improve the stability and accuracy of degradation diagnosis under long-term operating conditions, and thus enhance the safety and reliability of natural gas processing system operation. Summary of the Invention

[0008] This invention discloses a method for diagnosing degradation faults in a desulfurization tower during a natural gas purification process. The specific method is as follows: Synchronous collection of desulfurization tower operating parameters Upstream process status parameters Downstream load command parameters and environmental auxiliary parameters For multi-source operational information vectors Perform consistency processing and obtain a consistent runtime information sequence. ; Based on the consistent operation information sequence Identify continuously changing effective operating conditions ; Functional status of desulfurization tower based on functional semantic recognition of natural gas processing flow ; Using historical data on the health status of the desulfurization tower, we can analyze the functional status of the desulfurization tower. The separation process can explain the components. Obtain the functional residual components of the desulfurization tower itself. ; During the healthy operation of the desulfurization tower, for each typical operating condition, the residual components of the desulfurization tower's own functions are considered. Establish normal residual benchmarks for each typical working condition; Based on the normal residual benchmark, the comprehensive deviation of the desulfurization tower is calculated. ; Based on comprehensive deviation Constructing cumulative decision quantity Based on cumulative judgment value Determine if the desulfurization tower is degraded or malfunctioning.

[0009] Furthermore, after determining that the desulfurization tower has a degraded abnormal fault, the functional residual components at the entry time are used as the basis for further analysis. Constructing a normalized functional variation structure vector ; Normalized functional change structure vector The degradation fault type is identified by matching it with pre-established degradation fault patterns.

[0010] Furthermore, after identifying an abnormal degradation fault in the desulfurization tower, the degradation stability window is utilized. Estimate new residual statistics from residual data within the range. ; With the new residual statistics Correct Interpretable Functional Components And complete subsequent diagnosis and closed-loop iteration.

[0011] Furthermore, identify continuously changing effective operating conditions. The specific method is as follows: Depend on The selected operating condition indication components are used to construct the operating condition discriminant variables. Introduce fuzzy working condition membership vectors: ; in, Indicates time Belongs to the Typical operating condition range The degree of membership, and satisfying: ; Based on this, a valid set of operating conditions is defined for subsequent matching: ; in, This is the effective threshold for membership.

[0012] Furthermore, the functional status of the desulfurization tower is identified based on the semantic recognition of the natural gas processing flow. The specific method is as follows: Based on the consistent operation information sequence Constructing desulfurization towers under typical operating conditions The functional state vector below: ; in, Construct operators for functional states; Let from Selecting high-frequency signals The scale is defined as coarse-grained sequence: ; Based on scale Symbol pattern probability Calculate entropy features: ; Incorporating multi-scale entropy features into the functional state vector yields the functional state of the desulfurization tower. : .

[0013] Furthermore, the functional residual components of the desulfurization tower itself are obtained. The specific method is as follows: Based on historical data on the health status of desulfurization tower equipment, a virtual health twin process interpretation model was established. Its input uses a consistent runtime information sequence. The output is an interpretable functional component: ; in, Used to characterize the response relationship of functional status under healthy conditions as process conditions change; For model parameters, denoted as; ; in, The time decay factor, These are correction parameters estimated from recent healthy samples; Desulfurization tower functional status Decomposed into process interpretable components and functional residual components ; in, This represents the functional change component that can be explained by changes in process conditions. This represents functional residual components that cannot be explained by process changes. Therefore, we get: ;

[0014] Furthermore, the overall deviation of the desulfurization tower is calculated. The specific method is as follows: During the healthy operation phase of the desulfurization tower equipment, typical operating conditions are addressed. Based on residual samples Establish a normal function residual baseline and calculate the residual mean vector and covariance matrix: ; ; in, ; During real-time operation, for each effective working condition Calculate the residual deviation: ; in, Used to quantify operating conditions The degree to which the lower residuals deviate from the healthy baseline; To reflect the overall deviation under fuzzy conditions, a membership-weighted overall deviation is constructed. : ;

[0015] Furthermore, the specific methods for determining whether the desulfurization tower has degraded or malfunctioned are as follows: Based on the overall deviation Construct the cumulative decision value: ; in, This indicates the deviation of the overall functional residual after fuzzy working condition weighting; The overall deviation reference level of the desulfurization tower during the healthy operation phase is defined as follows: ; When the following conditions are met: ; And the duration is not less than When the desulfurization tower enters the abnormal degradation evolution stage, the trigger time is recorded as... ;in: For rapid diagnosis of thresholds.

[0016] Furthermore, the types of degradation faults are identified using the following methods: At the trigger time Next, first determine the main operating condition number of the equipment at that moment. The main operating condition number is defined as follows: ; in, The membership degree of the working condition obtained in step S3; Select the main operating condition range From the functional residual vector, construct the functional change structure vector: ; in, It is a 2-norm; Used to depict the relative changes between different functional indicators; A set of degradation failure modes is pre-established based on the structural mechanism of the desulfurization tower and historical failure data: ; in, For the first Reference structure vectors corresponding to typical degradation mechanisms The number of degradation modes; The current functional change structure vector is matched with each degradation failure mode, and the similarity is calculated: ; Determine the functional fault diagnosis results of the desulfurization tower: .

[0017] Furthermore, with the new residual statistics Correct Interpretable Functional Components And complete subsequent diagnosis and closed-loop iteration, the specific method is as follows: Estimating operating conditions within the degradation stability window Residual statistics below: ; ; in, For degradation window and working conditions Matched sample set, Indicates the number of samples; Smoothly update the residual baseline: ; ; in, To update the smoothing factor; The updated residual mean As working condition The following degradation compensation items: ; And its correction process can be used to explain the functional components: ; The corrected residual is obtained accordingly: ; in, Will replace the original Enter S6 to calculate new and It will be used in the next round of diagnosis; To adapt the process interpretation model to slow drift during long-term operation, parameter time decay updates are adopted: ; in, Estimated from recent samples, This is the attenuation factor.

[0018] Due to the adoption of the above technical solutions, this application has the following beneficial effects: 1. Enables early identification of desulfurization tower degradation faults: This invention comprehensively characterizes the functional changes of the desulfurization tower by constructing functional status indicators such as equipment purification efficiency and purification capacity per unit processing volume, so that the functional degradation trend can be identified in the early stage of degradation before the operating parameters exceed the limit.

[0019] 2. It can adapt to multiple operating conditions in the natural gas processing process: By introducing a fuzzy operating condition identification mechanism, corresponding functional benchmarks are established under different load conditions to achieve operating condition matching diagnosis, thereby reducing misdiagnosis caused by load changes or gas quality fluctuations.

[0020] 3. Effectively eliminates the impact of process disturbances on diagnostic results: By establishing a function-process decoupling model, the functional changes of the desulfurization tower are decomposed into process interpretable components and equipment functional residual components, thereby improving the accuracy of degradation fault diagnosis.

[0021] 4. Capable of diagnosing different types of desulfurization tower degradation faults: By constructing functional change structures and matching them with preset fault modes, it is possible to identify different types of degradation faults such as packing contamination, decreased mass transfer efficiency, or solvent performance degradation.

[0022] 5. Improves diagnostic stability under long-term operating conditions: After detecting the degradation of the desulfurization tower, the functional residuals are updated and the explained functional components under different operating conditions are corrected by using the updated residuals, forming a closed-loop iterative diagnostic mechanism, which improves the stability of the diagnostic model under long-term operating conditions.

[0023] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0024] The accompanying drawings of this invention are described below.

[0025] Figure 1 This is a schematic diagram of the overall process of the present invention. Detailed Implementation

[0026] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0027] The natural gas processing plant employs an absorption-regeneration process to desulfurize and decarbonize the feedstock natural gas. The natural gas purification production line, starting from the upstream feedstock gas inlet, sequentially includes a feedstock gas pretreatment unit, a desulfurization tower unit, a rich liquid flash evaporation and heat exchange unit, a regeneration tower unit, a solvent circulation system, and a product gas conveying unit, forming a continuously operating long-process production line.

[0028] In this production line, the desulfurization tower, as a key piece of equipment responsible for absorbing and removing acidic components, is affected not only by the internal operating conditions but also by process conditions such as the feed gas flow rate, the concentration of acidic components in the feed gas, and the solvent circulation status. During long-term operation, the packing material inside the desulfurization tower may become contaminated or its mass transfer performance may decline, leading to a gradual decrease in natural gas purification efficiency. In the early stages, this type of degradation typically does not cause single operating parameters such as pressure, temperature, or flow rate to exceed limits; instead, it manifests as a slow decline in purification capacity, making it highly insidious.

[0029] A method for diagnosing degradation faults in desulfurization towers during natural gas purification processes is presented. Taking the desulfurization tower in a natural gas processing unit as the object, the method analyzes the operating status of the desulfurization tower based on unit operating data. (See also...) Figure 1 The following steps in this embodiment enable the diagnosis of desulfurization tower degradation faults: S1: Multi-source synchronous collection of desulfurization tower operation information.

[0030] In this embodiment, the natural gas purification production line is a long-process system that operates continuously. The desulfurization tower, as a key unit responsible for removing acidic components, is not only related to its own operating conditions, but is also affected by the raw gas load and downstream operating requirements.

[0031] Firstly, focusing on the desulfurization tower, based on the production line process structure, we identify the set of measurement points that have a direct or indirect impact on its purification function. The set of measurement points includes: Measurement points on the desulfurization tower body: internal temperature Pressure before and after the tower ; Upstream measurement point: Raw gas flow rate Concentration of acidic components in the raw gas ; Downstream measurement point: Concentration of acidic components in purified natural gas ; Auxiliary measurement point: solvent circulation flow rate .

[0032] At any moment The operational information from the set of measurement points is collected synchronously and organized into an operational information vector: ; S2: Time consistency processing of runtime information.

[0033] Considering the differences in data sampling frequency and communication delay at different measurement points in the natural gas purification production line, directly using the raw collected data for analysis can easily introduce time misalignment errors, thereby affecting the accuracy of diagnosis.

[0034] The runtime information collected in step S1 is processed for time consistency, including time alignment, removal of outlier sampling points, and compensation for missing data, to obtain a time-consistent runtime information sequence: ; in, This indicates the time consistency handling operator.

[0035] S3: Adaptive fuzzy condition recognition.

[0036] In the natural gas purification and production process, the operating load of the desulfurization tower exhibits a continuous evolution characteristic with changes in the flow rate and quality of the raw gas, making it difficult to divide the operating conditions using fixed intervals.

[0037] Therefore, from Selecting the feed gas flow rate Concentration of acidic components in the feed gas As a component indicating the operating condition, a variable for determining the operating condition is constructed, and a fuzzy operating condition membership vector is introduced: ; in, Indicates the time of the desulfurization tower The following belongs to the Typical operating condition range The degree of membership, and satisfying: ； Based on this, weighted operating condition information is constructed: ; And define the set of valid operating conditions: ; in, This is the effective threshold for membership.

[0038] S4: Construction of device functional status based on purification function semantics.

[0039] Based on weighted operating data, and starting from the acid component removal function of the desulfurization tower, a model of the equipment under typical operating conditions is constructed. The functional state vector below.

[0040] In this embodiment, the core function of the desulfurization tower is to remove acidic components from the raw gas; therefore, the following functional state quantities are constructed: Purification efficiency characteristics: ; Purification capacity characteristics per unit processing volume: ; in, Used to characterize the purification efficiency of the desulfurization tower at time t; Used to characterize the purification capacity under unit feed gas flow conditions. When a desulfurization tower experiences degradation such as packing fouling and decreased mass transfer efficiency, it typically manifests as... , It decreases slowly over time.

[0041] Therefore, a desulfurization tower is constructed within the operating range. The functional state vector below: ; Furthermore, a working condition-weighted functional state vector is introduced: ; in, The membership degree of the working condition obtained in step S3 is used to achieve smooth matching of functional states under multiple working conditions.

[0042] S5: Functions based on process constraints – process decoupling.

[0043] In natural gas purification production lines, changes in the functional status of desulfurization towers can originate from normal fluctuations in feed gas flow and composition, or from performance degradation of the desulfurization towers themselves. To avoid process disturbances masking diagnostic results, this step introduces function-process decoupling.

[0044] During the healthy operation phase of the desulfurization tower, a normal response relationship between functional status and process conditions is established based on historical data. The model input is... The output consists of process-interpretable functional components: ; in, These are model parameters; This represents the functional change component that can be explained by changes in process conditions.

[0045] During real-time operation, the functional state is decomposed into: ; in, This represents the functional change component that can be explained by changes in process conditions. This indicates the component representing the change in the device's own function.

[0046] Therefore, the functional residual components are: ; in, These are functional residual components that cannot be explained by process changes, used to characterize functional anomalies caused by the degradation of the desulfurization tower itself.

[0047] Considering the slow drift of the health baseline due to long-term operation of the desulfurization tower, a time decay update mechanism is introduced to adaptively update the model parameters: ; in, This is the time decay factor; These are correction parameters estimated from recent healthy samples.

[0048] S6: Establishment of equipment's own functional benchmarks and calculation of deviation.

[0049] During the healthy operation phase of the desulfurization tower equipment, typical operating conditions are addressed. Based on residual samples Establish a normal function residual baseline and calculate the residual mean vector and covariance matrix: ; ; in, .

[0050] During real-time operation, for each effective working condition Calculate the deviation: ; Furthermore, the comprehensive deviation is constructed using the fuzzy condition membership degree: ; S7: Degradation and abnormal evolution determination for high time sensitivity requirements.

[0051] Considering that the functional degradation of desulfurization towers typically exhibits a continuous evolutionary characteristic, a cumulative judgment quantity is constructed based on the functional deviation obtained in step S6: ; When the cumulative determination value exceeds the threshold The desulfurization tower was determined to have entered a stage of abnormal degradation and evolution.

[0052] S8: Degradation Fault Diagnosis Based on Functional Changes in Structure. This step, based on abnormal triggering, further analyzes the structural characteristics of functional changes in the desulfurization tower itself to diagnose the type of functional fault.

[0053] S81: Determine the primary operating condition. Since a fuzzy operating condition identification method was used in step S3, the desulfurization tower may belong to multiple operating conditions at any given time. To ensure the diagnostic conclusion has clear engineering implications, at the moment the functional anomaly is triggered, the operating condition with the highest membership value is selected from all operating conditions as the primary operating condition upon which the current diagnosis is based.

[0054] ; in, Indicates the time of the desulfurization tower Belongs to the The degree of each working condition.

[0055] Select the operating condition number with the largest membership value as the main operating condition number used in the current diagnosis.

[0056] S82: Construction of the functional change structure of the desulfurization tower. After determining the main operating condition, select the functional residual vector of the desulfurization tower obtained in step S5 under this condition.

[0057] The functional residual vector is used to reflect the actual changes in the functional status of the desulfurization tower after eliminating the effects of raw gas fluctuations and process disturbances.

[0058] In this embodiment, the functional residual vector of the desulfurization tower includes the following two components: The change in purification efficiency relative to the health baseline; the change in purification capacity per unit volume relative to the health baseline.

[0059] The two functional residual components mentioned above are combined to form the functional change vector of the desulfurization tower: ; To eliminate the differences in numerical magnitude and units among different functional indicators, the functional change vector is normalized to obtain the functional change structure vector: ; The functional change structure vector is used to characterize the relative change relationship between different functional indicators of the desulfurization tower.

[0060] S83: Degradation Fault Mode Matching Diagnosis Based on Functional Change Structure. A degradation mode library is pre-established based on the structural characteristics and historical operating experience of the desulfurization tower.

[0061] Each functional failure mode corresponds to a standardized functional change structure, used to describe the changing characteristics of various functional indicators under a specific functional degradation mechanism. A degradation mode library is pre-established: ; in, For the first Reference structure vectors for typical degradation modes, This represents the number of patterns. The similarity between the current structure vector and each reference pattern is calculated: ; in, Indicates the current functional change structure and the first The degree of similarity between different functional failure modes.

[0062] The fault mode with the highest similarity is used as the current functional fault diagnosis result of the desulfurization tower: ; S84: Degradation Diagnosis Result Output. Based on the functional change and structural matching results, output the functional fault diagnosis conclusion of the desulfurization tower.

[0063] In this embodiment, the diagnostic results include, but are not limited to, the following types: If the purification efficiency and the residual purification capacity per unit throughput both show a continuous negative deviation under the main operating conditions, it is diagnosed as a degradation type of packing contamination / mass transfer capacity decline. If the purification efficiency residual deviates significantly negatively while the purification capacity residual per unit throughput does not change significantly, the diagnosis is solvent performance degradation or solvent circulation efficiency decline. The final output is the functional fault diagnosis result of the desulfurization tower.

[0064] S9: Evolutionary online functional baseline reconstruction and closed-loop self-calibration. After degradation is confirmed, the residual baseline is updated using the functional residuals from the degradation stage, and the updated residuals are used to correct the process-explainable functional components under different operating conditions. This enables long-term online closed-loop iteration and adaptive baseline evolution, avoiding long-term false alarms or sensitivity degradation caused by baseline drift.

[0065] When the degradation confirmation condition of S7 is met, i.e., the duration Then, the operating conditions are estimated within the degradation stability window. Residual statistics below: ; ; in, For degradation window and working conditions Matched sample set, Indicates the number of samples.

[0066] Smoothly update the residual baseline: ; ; in, To update the smoothing factor. Updated , It will be reused for S6 deviation calculation to realize the residual benchmark that evolves with the degradation stage.

[0067] Furthermore, the updated residual mean As working condition The following degradation compensation items: ; And its correction process can be used to explain the functional components: ; The corrected residual is obtained accordingly: ; in, Will replace the original Enter S6 to calculate new and Then, it continues to enter S7 and S8 to complete the next round of judgment and diagnosis, thus forming a closed loop iteration.

[0068] To adapt the process interpretation model to slow drift under long-term operation, parameter time decay updates can continue to be used: ; in, The time decay factor, The correction parameters are estimated from recent samples to ensure the model's adaptability in long-term operation.

[0069] 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 it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for diagnosing degradation faults in a desulfurization tower during a natural gas purification process, characterized in that, The specific method is as follows: Synchronous collection of desulfurization tower operating parameters Upstream process status parameters Downstream load command parameters and environmental auxiliary parameters For multi-source operational information vectors Perform consistency processing and obtain a consistent runtime information sequence. ; Based on the consistent operation information sequence Identify continuously changing effective operating conditions ; Functional status of desulfurization tower based on functional semantic recognition of natural gas processing flow ; Using historical data on the health status of the desulfurization tower, we can analyze the functional status of the desulfurization tower. The separation process can explain the components. Obtain the functional residual components of the desulfurization tower itself. ; During the healthy operation of the desulfurization tower, for each typical operating condition, the residual components of the desulfurization tower's own functions are considered. Establish normal residual benchmarks for each typical working condition; Based on the normal residual benchmark, the comprehensive deviation of the desulfurization tower is calculated. ; Based on comprehensive deviation Constructing cumulative decision quantity Based on cumulative judgment value Determine if the desulfurization tower is degraded or malfunctioning.

2. The method for diagnosing degradation faults in a desulfurization tower during natural gas purification as described in claim 1, characterized in that, After determining that the desulfurization tower has a degraded abnormal fault, the functional residual components at the entry time are used as the basis. Constructing a normalized functional variation structure vector ; Normalized functional change structure vector The degradation fault type is identified by matching it with pre-established degradation fault patterns.

3. The method for diagnosing degradation faults in a desulfurization tower during natural gas purification as described in claim 1, characterized in that, After identifying an abnormal degradation fault in the desulfurization tower, utilize the degradation stability window. Estimate new residual statistics from residual data within the range. ; With the new residual statistics Correct Interpretable Functional Components And complete subsequent diagnosis and closed-loop iteration.

4. The method for diagnosing degradation faults in a desulfurization tower during natural gas purification as described in claim 1, characterized in that, Identifying continuously changing effective operating conditions The specific method is as follows: Depend on The selected operating condition indication components are used to construct the operating condition discriminant variables. Introduce fuzzy working condition membership vectors: ； in, Indicates time Belongs to the Typical operating condition range The degree of membership, and satisfying: ； Based on this, a valid set of operating conditions is defined for subsequent matching: ； in, This is the effective threshold for membership.

5. The method for diagnosing degradation faults in a desulfurization tower during natural gas purification as described in claim 1, characterized in that, Functional status of desulfurization tower based on functional semantic recognition of natural gas processing flow The specific method is as follows: Based on the consistent operation information sequence Constructing desulfurization towers under typical operating conditions The functional state vector below: ； in, Construct operators for functional states; Let from Selecting high-frequency signals The scale is defined as coarse-grained sequence: ； Based on scale Symbol pattern probability Calculate entropy features: ； Incorporating multi-scale entropy features into the functional state vector yields the functional state of the desulfurization tower. : ；。 6. The method for diagnosing degradation faults in a desulfurization tower during natural gas purification as described in claim 5, characterized in that, Obtain the functional residual components of the desulfurization tower itself The specific method is as follows: Based on historical data on the health status of desulfurization tower equipment, a virtual health twin process interpretation model was established. Its input uses a consistent runtime information sequence. The output is an interpretable functional component: ； in, Used to characterize the response relationship of functional status under healthy conditions as process conditions change; For model parameters, denoted as; ； in, The time decay factor, These are correction parameters estimated from recent healthy samples; Desulfurization tower functional status Decomposed into process interpretable components and functional residual components ； in, This represents the functional change component that can be explained by changes in process conditions. This represents functional residual components that cannot be explained by process changes. Therefore, we get: ；。 7. The method for diagnosing degradation faults in a desulfurization tower during natural gas purification as described in claim 6, characterized in that, Calculate the overall deviation of the desulfurization tower The specific method is as follows: During the healthy operation phase of the desulfurization tower equipment, typical operating conditions are addressed. Based on residual samples Establish a normal function residual baseline and calculate the residual mean vector and covariance matrix: ； ； in, ; During real-time operation, for each effective working condition Calculate the residual deviation: ； in, Used to quantify operating conditions The degree to which the lower residuals deviate from the healthy baseline; To reflect the overall deviation under fuzzy conditions, a membership-weighted overall deviation is constructed. : 。 8. The method for diagnosing degradation faults in a desulfurization tower during a natural gas purification process as described in claim 2 or 3, characterized in that, The specific methods for determining whether a desulfurization tower is experiencing degradation or abnormal malfunction are as follows: Based on the overall deviation Construct the cumulative decision value: ； in, This indicates the deviation of the overall functional residual after fuzzy working condition weighting; The overall deviation reference level of the desulfurization tower during the healthy operation phase is defined as follows: ； When the following conditions are met: ； And the duration is not less than When the desulfurization tower enters the abnormal degradation evolution stage, the trigger time is recorded as... ;in: For rapid diagnosis of thresholds.

9. The method for diagnosing degradation faults in a desulfurization tower during natural gas purification as described in claim 9, characterized in that, The specific methods for identifying degradation fault types are as follows: At the trigger time Next, first determine the main operating condition number of the equipment at that moment. The main operating condition number is defined as follows: ； in, The membership degree of the working condition obtained in step S3; Select the main operating condition range From the functional residual vector, construct the functional change structure vector: ； in, It is a 2-norm; Used to depict the relative changes between different functional indicators; A set of degradation failure modes is pre-established based on the structural mechanism of the desulfurization tower and historical failure data: ； in, For the first Reference structure vectors corresponding to typical degradation mechanisms The number of degradation modes; The current functional change structure vector is matched with each degradation failure mode, and the similarity is calculated: ； Determine the functional fault diagnosis results of the desulfurization tower: ；。 10. The method for diagnosing degradation faults in a desulfurization tower during a natural gas purification process as described in claim 9, characterized in that, With the new residual statistics Correct Interpretable Functional Components And complete subsequent diagnosis and closed-loop iteration, the specific method is as follows: Estimating operating conditions within the degradation stability window Residual statistics below: ； ； in, For degradation window and working conditions Matched sample set, Indicates the number of samples; Smoothly update the residual baseline: ； ； in, To update the smoothing factor; The updated residual mean As working condition The following degradation compensation items: ； And its correction process can be used to explain the functional components: ； The corrected residual is obtained accordingly: ； in, Will replace the original Enter S6 to calculate new and It will be used in the next round of diagnosis; To adapt the process interpretation model to slow drift during long-term operation, parameter time decay updates are adopted: ； in, Estimated from recent samples, This is the attenuation factor.

Images