Graphite heat exchanger fault on-line diagnosis method

Abstract

This invention discloses an online fault diagnosis method for graphite heat exchangers, relating to the field of data processing technology. The method includes: performing data acquisition to establish a time-series monitoring dataset; dividing the time-series monitoring dataset into stages, creating a uniform sampling space; constructing a fitting dataset with constraints of minimum sample size and sampling space, and performing interior and exterior point discrimination; iteratively constructing the fitting dataset and obtaining a working fitting curve; establishing a calibration fitting curve, and establishing a basic deviation evaluation value based on the deviation evaluation between the calibration fitting curve and the working fitting curve; performing exterior point discrete analysis to establish additional deviation evaluation values; setting key coefficients for monitoring parameters in fault anomalies, calculating fault values, and performing fault early warning. This method solves the technical problem of difficulty in timely and accurate diagnosis of graphite heat exchanger faults in existing technologies, achieving online fault diagnosis of graphite heat exchangers and improving the accuracy and efficiency of fault diagnosis.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and more specifically to an online diagnostic method for graphite heat exchanger faults. Background Technology

[0002] With the continuous advancement of industrial technology and the increasing demands of modern production on equipment performance, graphite heat exchangers, as a highly efficient and corrosion-resistant heat exchange device, have been widely used in various fields such as chemical engineering, pharmaceuticals, and metallurgy. However, long-term operation and complex working environments often lead to performance degradation and frequent failures. Traditional fault diagnosis methods often rely on periodic inspections and offline analysis, which suffer from slow response times and low diagnostic accuracy, making it difficult to meet the requirements of modern industrial production for high efficiency, precision, and real-time performance. Summary of the Invention

[0003] This application provides an online diagnostic method for graphite heat exchanger faults, which solves the technical problem in the prior art that it is difficult to diagnose graphite heat exchanger faults in a timely and accurate manner.

[0004] In view of the above problems, this application provides an online diagnostic method for graphite heat exchanger faults.

[0005] This application provides an online fault diagnosis method for graphite heat exchangers, the method comprising:

[0006] Perform data acquisition and establish a time-series monitoring dataset, wherein the time-series monitoring dataset is a monitoring dataset of a graphite heat exchanger, and the monitoring parameters include temperature parameters, pressure parameters, flow parameters, and vibration parameters;

[0007] The time-series monitoring dataset is divided into stages, and a uniform step size is configured within the stage division results. Based on the uniform step size, the stage division results are divided into a uniform sampling space.

[0008] The minimum sample size is configured based on the sampling space model, and an error threshold is set.

[0009] Using the minimum sample size and the sampling space as constraints, a fitted dataset is constructed, and inliers and outliers are distinguished based on the construction results and the error threshold.

[0010] Perform the construction iteration of the fitted dataset and obtain the working fitted curve corresponding to the fitted dataset with the most inliers;

[0011] Establish a calibration fitting curve, evaluate the deviation between the calibration fitting curve and the working fitting curve based on the corresponding parameters, and establish a basic deviation evaluation value;

[0012] Based on the outgoing points of the working fit curve and the working fit curve, an outgoing point discretization analysis is performed to establish an additional deviation evaluation value;

[0013] Key coefficients for monitoring parameters during faults and anomalies are set, and fault values ​​are calculated based on the key coefficients, the basic deviation evaluation values, and the additional deviation evaluation values. Fault warnings are then executed based on the calculation results of the fault values.

[0014] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0015] A time-series monitoring dataset was established through data acquisition. This dataset is a monitoring dataset for a graphite heat exchanger, and the monitored parameters include temperature, pressure, flow rate, and vibration. The time-series monitoring dataset was divided into stages, and a uniform step size was configured within each stage. Based on this uniform step size, the stage division results were segmented into a uniform sampling space. A minimum sample size for the model was configured based on the sampling space, and an error threshold was set. Using the minimum sample size and sampling space as constraints, a fitting dataset was constructed, and interior and exterior point discrimination was performed based on the construction results and the error threshold. Iterative construction of the fitting dataset was performed, and the working fitting curve corresponding to the fitting dataset with the most interior points was obtained. A calibration fitting curve was established, and deviation evaluation between the calibration fitting curve and the working fitting curve was performed based on the corresponding parameters, establishing a basic deviation evaluation value. Exterior point discretization analysis was performed based on the exterior points of the working fitting curve, establishing an additional deviation evaluation value. Key coefficients for the monitored parameters in fault anomalies were set. Fault values ​​were calculated based on the key coefficients, basic deviation evaluation values, and additional deviation evaluation values. Fault warnings were implemented based on the calculated fault values. This invention solves the technical problem of difficulty in timely and accurate diagnosis of graphite heat exchanger faults in existing technologies, and realizes online diagnosis of graphite heat exchanger faults, thereby improving the accuracy and efficiency of fault diagnosis. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic flowchart of an online fault diagnosis method for graphite heat exchangers provided in an embodiment of this application.

[0018] Figure 2 This is a schematic diagram illustrating the process of constructing a fitted dataset in the online fault diagnosis method for graphite heat exchangers provided in this application embodiment. Detailed Implementation

[0019] This application provides an online fault diagnosis method for graphite heat exchangers, which solves the technical problem in the prior art that it is difficult to diagnose graphite heat exchanger faults in a timely and accurate manner.

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0021] It should be noted that the terms “comprising” and “having”, and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0022] Example 1

[0023] like Figure 1 As shown in the embodiment of this application, an online fault diagnosis method for graphite heat exchangers is provided, wherein the method includes:

[0024] Perform data acquisition and establish a time-series monitoring dataset, wherein the time-series monitoring dataset is a monitoring dataset of a graphite heat exchanger, and the monitoring parameters include temperature parameters, pressure parameters, flow parameters, and vibration parameters;

[0025] By collecting temperature, pressure, flow, and vibration parameters of the graphite heat exchanger during operation using sensors, a time-series monitoring dataset is established. This dataset is a monitoring dataset of the graphite heat exchanger arranged in chronological order, containing the operating status information of the graphite heat exchanger at each point in time.

[0026] The time-series monitoring dataset is divided into stages, and a uniform step size is configured within the stage division results. Based on the uniform step size, the stage division results are divided into a uniform sampling space.

[0027] To obtain the characteristics of graphite heat exchangers at different operating stages or under different working conditions, the time-series monitoring dataset will be divided into different stages. A uniform step size will be configured within each stage division result to divide the stage division result into a uniform sampling space.

[0028] The minimum sample size is configured based on the sampling space model, and an error threshold is set.

[0029] Using the minimum sample size and the sampling space as constraints, a fitted dataset is constructed, and inliers and outliers are distinguished based on the construction results and the error threshold.

[0030] Taking the collected temperature parameters as an example, pressure, flow, and vibration parameters will be processed in the same way. Optionally, the temperature parameters can be divided into different stages, with a uniform step size (equal number of temperature data points) configured within each stage. These temperature data points constitute a sampling space. The number of temperature data points in the sampling space is used as the minimum sample size for the fitting model, and an error threshold is set to evaluate the difference between the data points and the fitted curve, i.e., the error. Based on the collected temperature data points and the constraints of the minimum sample size and sampling space, a fitting dataset is constructed. For each data point in the fitting dataset, an inlier / outlier distinction is made using the error threshold. An inlier is a data point whose distance from the fitting dataset is less than or equal to the error threshold, while an outlier is a data point whose distance from the fitting dataset is greater than the error threshold.

[0031] Furthermore, such as Figure 2 As shown, the fitted dataset construction is performed with the minimum sample size and the sampling space as constraints, including:

[0032] For each sampling space, perform adaptive data clustering within the space, establish adaptive data clustering results, and establish the first edge value of the data based on the adaptive data clustering results;

[0033] Establish spatial associations in the sampling space, wherein the spatial associations are adjacent spatial associations;

[0034] The largest clustering family in the adjacent space is determined based on the adaptive data clustering results, and the second edge value of the data in the sampling space is established based on the largest clustering family.

[0035] The penalty factor of the data is calculated using the first edge value and the second edge value, and the fitting dataset is constructed using the penalty factor calculation result.

[0036] Within each sampling space, the DBSCAN algorithm dynamically determines the number and shape of clusters based on the density and distance of data points, obtaining adaptive data clustering results. The minimum or maximum value of clustered data points from the adaptive clustering results is determined as the first marginal value, used to identify the cluster boundaries. Spatial association is determined based on adjacent spatial associations, and spatial associations between multiple sampling spaces are determined according to periodicity. Based on the adaptive clustering results, the cluster with the densest data point aggregation in adjacent sampling spaces is identified as the largest cluster. By comparing the largest clusters in adjacent spaces, the second marginal value of the data in the sampling space is established. A penalty factor is calculated using the first and second marginal values. The penalty factor is a parameter used to adjust the weight of data points during the construction of the fitted dataset. When a data point approaches the marginal value, the penalty factor increases accordingly. The penalty factor is used to integrate data points; data points with excessively large penalty factors are considered marginal and removed. The remaining data points are used to construct the fitted dataset.

[0037] Perform the construction iteration of the fitted dataset and obtain the working fitted curve corresponding to the fitted dataset with the most inliers;

[0038] The construction of the fitting dataset is iterated multiple times based on the time-series monitoring dataset. The number of inliers in the fitting datasets of multiple iterations is compared. The fitting dataset with the most inliers is selected, and the working fitting curve corresponding to the fitting dataset with the most inliers is generated. The working fitting curve reflects the current operating status of the graphite heat exchanger.

[0039] Establish a calibration fitting curve, evaluate the deviation between the calibration fitting curve and the working fitting curve based on the corresponding parameters, and establish a basic deviation evaluation value;

[0040] Data collected under normal and stable operating conditions of the graphite heat exchanger is used as the calibration dataset. A calibration fitting curve is established based on the calibration data, reflecting the operating data of the graphite heat exchanger under standard conditions. The calibration fitting curve is compared with the operating fitting curve, and the difference or distance between the two curves is calculated to quantify the degree of deviation. Optionally, the deviation between the calibration fitting curve and the operating fitting curve is calculated to establish a baseline deviation evaluation value, which reflects the deviation between the current state of the equipment and the calibrated normal state.

[0041] Based on the outgoing points of the working fit curve and the working fit curve, an outgoing point discretization analysis is performed to establish an additional deviation evaluation value;

[0042] An out-of-line discretization analysis is performed based on the out-of-line points of the working curve and the working curve itself. By calculating the distance between the out-of-line points and the working curve, an additional deviation evaluation value is obtained. The greater the distance, the higher the deviation evaluation value. The additional deviation evaluation value reflects the degree of deviation between the out-of-line points and the working curve.

[0043] Key coefficients for monitoring parameters during faults and anomalies are set, and fault values ​​are calculated based on the key coefficients, the basic deviation evaluation values, and the additional deviation evaluation values. Fault warnings are then executed based on the calculation results of the fault values.

[0044] Key coefficients are set for monitoring parameters during fault anomalies. These key coefficients reflect the importance of different monitoring parameters in fault detection. Fault values ​​are calculated based on these key coefficients, basic deviation evaluation values, and additional deviation evaluation values. Fault warnings are then issued based on the calculated fault values.

[0045] Furthermore, key coefficients for monitoring parameters during fault anomalies are defined, including:

[0046] Historical fault data of the graphite heat exchanger is collected and obtained. Based on the historical fault data, the impact analysis of each monitoring parameter is performed, and the first impact analysis result is generated.

[0047] Fault data of similar graphite heat exchangers are collected using big data, and the impact analysis of various monitoring parameters is performed based on the fault data collection results to generate a second impact analysis result.

[0048] Trust-weighted calculations are performed on the first and second impact analysis results, and key coefficients of the monitoring parameters corresponding to each fault anomaly are constructed based on the trust-weighted calculation results.

[0049] Historical fault data of graphite heat exchangers was collected, including the causes, manifestations, and impact of the faults. Impact analysis was performed on various monitoring parameters based on this historical fault data. Optionally, the frequency, variation, or correlation coefficient of each monitoring parameter in different fault types was calculated to assess their impact on the faults. The analysis results generated a first impact analysis, reflecting the correlation between each monitoring parameter and the faults in the historical data. Simultaneously, fault data from similar graphite heat exchangers was collected through a big data platform. This data came from other equipment, equipment of the same type, or equipment on the same production line. Collecting fault data from these similar devices further expanded the data sample size and improved the generality and accuracy of the analysis results. Based on the fault data collection results, impact analysis was performed on various monitoring parameters, generating a second impact analysis. Historical fault data and similar equipment fault data may have certain differences and uncertainties; therefore, the reliability of both needs to be considered comprehensively. Different weights were assigned based on factors such as data source, quality, and quantity. A trust-weighted calculation was performed on the first and second impact analysis results to obtain the final impact analysis result. Based on the trust-weighted calculation results, key coefficients for monitoring parameters corresponding to each fault anomaly are constructed. These key coefficients reflect the relative importance of different monitoring parameters in fault early warning and diagnosis.

[0050] Furthermore, the methods include:

[0051] The fault value is calculated using the following formula:

[0052] ;

[0053] in, The fault value characterizes the i-th type of fault. , , , These correspond to the key coefficients for temperature, pressure, flow rate, and vibration under the i-th type of fault. , , , These correspond to the basic deviation evaluation values ​​for temperature, pressure, flow rate, and vibration, respectively. , , , These correspond to additional deviation evaluation values ​​for temperature, pressure, flow rate, and vibration, respectively.

[0054] The fault value calculation formula comprehensively considers the basic deviation evaluation values ​​and additional deviation evaluation values ​​of different monitoring parameters (temperature, pressure, flow rate, and vibration), and performs a weighted summation using key coefficients to derive the fault value for a specific fault. In the fault value calculation formula, The fault value characterizes the i-th type of fault. , , , These correspond to the key coefficients for temperature, pressure, flow rate, and vibration under the i-th type of fault. , , , These correspond to the basic deviation evaluation values ​​for temperature, pressure, flow rate, and vibration, respectively. , , , These correspond to additional deviation evaluation values ​​for temperature, pressure, flow rate, and vibration, respectively.

[0055] Furthermore, the methods also include:

[0056] The parameters are predicted and fitted using the working fitting curve, and the prediction fitting results are established.

[0057] Based on the predicted fitting results, the deviation evaluation of the calibration fitting curve is performed, and a compensation deviation evaluation result is established.

[0058] The warning level of the fault warning is updated based on the compensation deviation evaluation results.

[0059] Based on the working fitting curve, parameters are predicted and fitted to establish a predicted fitting curve, i.e., the predicted fitting result. The predicted fitting result is then compared with the calibration fitting curve to evaluate the deviation, calculating the degree and frequency of the deviation. Based on the severity of the deviation, a compensation deviation evaluation result is established. The compensation deviation evaluation result is used to update the fault warning level. If the compensation deviation evaluation result shows that the deviation between the predicted value and the actual value is within an acceptable range, the warning level can remain unchanged. If the deviation is large, it indicates that the equipment may have an anomaly or is about to malfunction; in this case, the warning level should be increased to allow for timely intervention.

[0060] Furthermore, the methods also include:

[0061] Maintenance and detection are performed based on the aforementioned fault warnings, and matching data between maintenance and detection results and fault warnings is established.

[0062] The error threshold is generated using the matching data, along with a set compensation coefficient.

[0063] The error threshold is iteratively optimized based on the set compensation coefficient.

[0064] When an early warning is issued, the graphite heat exchanger should be maintained and inspected to identify and resolve potential problems. During maintenance, the actual condition of the equipment should be recorded in detail, including but not limited to replaced parts, repaired issues, and observed anomalies. The results of the maintenance inspection should be matched with the fault warnings to record which warnings are accurate and which may be inaccurate. Based on the matching data of the maintenance inspection results and fault warnings, the error situation of the warning system should be analyzed. If significant errors are found in certain types of warnings, the corresponding error thresholds may need to be adjusted. A compensation coefficient is set, and the error threshold is dynamically adjusted based on the analysis results of the matching data. The compensation coefficient is applied to the iterative optimization process of the error threshold. By continuously adjusting the error threshold, the most suitable error threshold for the current equipment and operating environment can be gradually found, improving the accuracy and reliability of the fault warnings.

[0065] Furthermore, the methods also include:

[0066] Idle time records are maintained for the graphite heat exchanger to establish a maintainable time cycle;

[0067] The fault warnings of the graphite heat exchanger are cumulatively recorded, and a cumulative warning value is generated;

[0068] A maintenance cycle is established by performing time-breakthrough analysis based on the cumulative early warning value and the maintainable time period;

[0069] The maintenance and management of the graphite heat exchanger shall be performed according to the aforementioned maintenance cycle.

[0070] Idle time of the graphite heat exchanger is recorded. By collecting this idle time data, a maintainable time cycle can be established. Simultaneously, fault warnings for the graphite heat exchanger are cumulatively recorded. Each time a warning occurs, relevant information is recorded, such as warning type, occurrence time, and warning level. By accumulating this warning data, a cumulative warning value is generated, reflecting the frequency and severity of warnings within a certain period. Time-break analysis is performed using the cumulative warning value and the maintainable time cycle to determine the most appropriate time for maintenance. If the cumulative warning value reaches or exceeds a set threshold within the maintainable time cycle, maintenance is required. Maintenance management of the graphite heat exchanger is then executed according to the established maintenance cycle.

[0071] Furthermore, by performing time-breakthrough analysis using the cumulative early warning value and the maintainable time period, a maintenance cycle is established, including:

[0072] If the maintainable time period cannot be exceeded, the maintainable time period will be used as the maintenance period for maintenance management.

[0073] If the maintainable time period of the graphite heat exchanger is not exceeded by the cumulative warning value, it means that the cumulative warning value has not exceeded the set threshold within the maintainable time period. In this case, the current maintainable time period is directly used as the maintenance cycle, and corresponding maintenance management is performed.

[0074] In summary, the embodiments of this application have at least the following technical effects:

[0075] A time-series monitoring dataset was established through data acquisition. This dataset is a monitoring dataset for a graphite heat exchanger, and the monitored parameters include temperature, pressure, flow rate, and vibration. The time-series monitoring dataset was divided into stages, and a uniform step size was configured within each stage. Based on this uniform step size, the stage division results were segmented into a uniform sampling space. A minimum sample size for the model was configured based on the sampling space, and an error threshold was set. Using the minimum sample size and sampling space as constraints, a fitting dataset was constructed, and interior and exterior point discrimination was performed based on the construction results and the error threshold. Iterative construction of the fitting dataset was performed, and the working fitting curve corresponding to the fitting dataset with the most interior points was obtained. A calibration fitting curve was established, and deviation evaluation between the calibration fitting curve and the working fitting curve was performed based on the corresponding parameters, establishing a basic deviation evaluation value. Exterior point discretization analysis was performed based on the exterior points of the working fitting curve, establishing an additional deviation evaluation value. Key coefficients for the monitored parameters in fault anomalies were set. Fault values ​​were calculated based on the key coefficients, basic deviation evaluation values, and additional deviation evaluation values. Fault warnings were implemented based on the calculated fault values. This invention solves the technical problem of difficulty in timely and accurate diagnosis of graphite heat exchanger faults in existing technologies, and realizes online diagnosis of graphite heat exchanger faults, thereby improving the accuracy and efficiency of fault diagnosis.

[0076] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0077] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0078] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. An online fault diagnosis method for graphite heat exchangers, characterized in that, The method includes: Perform data acquisition and establish a time-series monitoring dataset, wherein the time-series monitoring dataset is a monitoring dataset of a graphite heat exchanger, and the monitoring parameters include temperature parameters, pressure parameters, flow parameters, and vibration parameters; The time-series monitoring dataset is divided into stages, and a uniform step size is configured within the stage division results. Based on the uniform step size, the stage division results are divided into a uniform sampling space. The minimum sample size is configured based on the sampling space model, and an error threshold is set. Using the minimum sample size and the sampling space as constraints, a fitted dataset is constructed, and inliers and outliers are distinguished based on the construction results and the error threshold. Perform the construction iteration of the fitted dataset and obtain the working fitted curve corresponding to the fitted dataset with the most inliers; Establish a calibration fitting curve, evaluate the deviation between the calibration fitting curve and the working fitting curve based on the corresponding parameters, and establish a basic deviation evaluation value; Based on the outgoing points of the working fit curve and the working fit curve, an outgoing point discretization analysis is performed to establish an additional deviation evaluation value; Set key coefficients for monitoring parameters during faults and anomalies, calculate fault values ​​based on the key coefficients, the basic deviation evaluation values, and the additional deviation evaluation values, and execute fault warnings based on the calculation results of the fault values. The process of constructing a fitted dataset, constrained by the minimum sample size and the sampling space, includes: For each sampling space, perform adaptive data clustering within the space, establish adaptive data clustering results, and establish the first edge value of the data based on the adaptive data clustering results; Establish spatial associations in the sampling space, wherein the spatial associations are adjacent spatial associations; The largest clustering family in the adjacent space is determined based on the adaptive data clustering results, and the second edge value of the data in the sampling space is established based on the largest clustering family. The penalty factor of the data is calculated using the first edge value and the second edge value, and the fitting dataset is constructed using the penalty factor calculation result.

2. The method as described in claim 1, characterized in that, The key coefficients for monitoring parameters during fault anomalies also include: Historical fault data of the graphite heat exchanger is collected and obtained. Based on the historical fault data, the impact analysis of each monitoring parameter is performed, and the first impact analysis result is generated. Fault data of similar graphite heat exchangers are collected using big data, and the impact analysis of various monitoring parameters is performed based on the fault data collection results to generate a second impact analysis result. Trust-weighted calculations are performed on the first and second impact analysis results, and key coefficients of the monitoring parameters corresponding to each fault anomaly are constructed based on the trust-weighted calculation results.

3. The method as described in claim 2, characterized in that, The method further includes: The fault value is calculated using the following formula: ； in, The fault value characterizes the i-th type of fault. , , , These correspond to the key coefficients for temperature, pressure, flow rate, and vibration under the i-th type of fault. , , , These correspond to the basic deviation evaluation values ​​for temperature, pressure, flow rate, and vibration, respectively. , , , These correspond to additional deviation evaluation values ​​for temperature, pressure, flow rate, and vibration, respectively.

4. The method as described in claim 1, characterized in that, The method further includes: The parameters are predicted and fitted using the working fitting curve, and the prediction fitting results are established. Based on the predicted fitting results, the deviation evaluation of the calibration fitting curve is performed, and a compensation deviation evaluation result is established. The warning level of the fault warning is updated based on the compensation deviation evaluation results.

5. The method as described in claim 1, characterized in that, The method further includes: Maintenance and detection are performed based on the aforementioned fault warnings, and matching data between maintenance and detection results and fault warnings is established. The error threshold is generated using the matching data, along with a set compensation coefficient. The error threshold is iteratively optimized based on the set compensation coefficient.

6. The method as described in claim 1, characterized in that, The method further includes: Idle time records are maintained for the graphite heat exchanger to establish a maintainable time cycle; The fault warnings of the graphite heat exchanger are cumulatively recorded, and a cumulative warning value is generated; A maintenance cycle is established by performing time-breakthrough analysis based on the cumulative early warning value and the maintainable time period; The maintenance and management of the graphite heat exchanger shall be performed according to the aforementioned maintenance cycle.

7. The method as described in claim 6, characterized in that, The step of establishing a maintenance cycle by performing time-breakthrough analysis based on the cumulative early warning value and the maintainable time period includes: If the maintainable time period cannot be exceeded, the maintainable time period will be used as the maintenance period for maintenance management.

Images