Safety valve service life evaluation method based on pressure history statistics and dynamic damage accumulation

Abstract

The invention discloses a safety valve service life evaluation method based on pressure history statistics and dynamic damage accumulation. According to the method, only safety valve inlet pressure time sequence data is needed, a load cycle is extracted through an improved rain flow counting method, and the nonlinear damage increment is calculated according to an S-N curve model dynamically degraded along with the accumulated damage degree. According to the method, a self-adaptive failure probability threshold is established based on historical failure data, and probabilistic prediction of the residual life is realized. The model is clear in physical significance, only needs a single data source, significantly reduces the implementation complexity and cost, greatly improves the precision of life prediction by considering the degradation of material performance, and is especially suitable for predicting the accelerated failure risk of the safety valve in the later period of long-term service.

Description

Technical Field

[0001] This invention relates to the field of safety valve life prediction technology, specifically to a method for assessing the service life of safety valves based on pressure history statistics and dynamic damage accumulation. Background Technology

[0002] Safety valves are critical safety protection devices in petrochemical, power, and energy industries, and their reliability directly affects the safety of personnel, equipment, and the environment. Fatigue failure is one of the most common failure modes of safety valves, stemming from the cyclic loads they endure during service, including normal pressure fluctuations, start-up and shutdown operations, and the valve's own opening and reseating motion.

[0003] Currently, the assessment of safety valve lifespan mainly relies on the following methods: Empirical method and periodic calibration: Based on standards or experience, a fixed calibration cycle is set (usually one year). This method is too conservative or risky and cannot reflect the actual operating status of individual valves, which may lead to "over-maintenance" or "under-maintenance".

[0004] Simple counting method: Count the total number of times the safety valve opens, and consider it to have reached the end of its service life when the number reaches a certain preset value. This method completely ignores fatigue damage caused by pressure fluctuations and the differences in the severity of different opening events.

[0005] The method based on linear cumulative damage theory employs Miner's rule and standard SN curves for damage calculation. While this method considers the load spectrum, its core flaw lies in using static and unchanging SN curves. In reality, under cyclic loading, the microstructure of materials changes, and fatigue properties (such as fatigue limit) gradually degrade. Using fixed SN curves systematically underestimates the damage accumulation rate during high-cycle periods, leading to an overestimation of the predicted remaining life and posing significant safety risks in later service life.

[0006] Therefore, there is an urgent need for an accurate life assessment method that can reflect the degradation of material properties and is easy to implement in engineering. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method for assessing the service life of safety valves that requires only pressure data, can simulate the dynamic degradation of material fatigue properties, and predicts remaining service life based on probability statistics. This method aims to achieve low-cost, high-precision predictive maintenance and effectively warn of the risk of accelerated valve failure in later stages.

[0008] To achieve the above objectives, the present invention proposes the following technical solution: A method for assessing the service life of safety valves based on pressure history statistics and dynamic damage accumulation includes the following steps: Load history acquisition steps: Acquire pressure time series data at the safety valve inlet; Load cycle extraction step: The pressure time series data is processed to identify and extract a complete set of pressure load cycles, wherein each cycle is described by its pressure range, average pressure and cycle type identifier, and the cycle type includes main cycle and secondary cycle; Dynamic damage accumulation step: Based on the extracted load cycles, a dynamically degraded material SN curve model is used to calculate the damage caused by each load cycle, and the total accumulated damage is obtained. In the dynamic SN curve model, the number of failure cycles of the material under a specific stress amplitude is a function of the current total accumulated damage. Remaining service life prediction steps: Based on the current total cumulative damage and combined with an adaptive failure probability distribution obtained from historical failure data statistics, predict the remaining service life of the safety valve when it reaches a predetermined failure probability threshold.

[0009] The present invention further specifies that the load cyclic extraction step employs an improved rainflow counting method, including: The pressure time series is filtered and denoised; all local peaks and valleys in the denoised signal are identified; an action judgment threshold is set according to the set pressure of the safety valve, which is the set pressure multiplied by a proportional coefficient; the period when the pressure exceeds the action judgment threshold is marked as the action period, and the rest is the static period; within the action period, a complete process of the pressure rising from below the action judgment threshold to the peak and then falling to the reseating pressure is defined as a main cycle, and its pressure range is the difference between the peak and the reseating pressure; within the static period, the standard rainflow counting method is applied to the pressure fluctuation signal to extract all complete secondary cycles, and the pressure range of the secondary cycle is determined by the difference between the peak and the valley of the cycle.

[0010] The present invention further specifies that the dynamic SN curve model is expressed as follows: the initial SN curve equation is that the number of failure cycles of the material under stress amplitude in the initial undamaged state is equal to a material constant multiplied by the negative m power of the stress amplitude, where m is another material constant; The dynamic SN curve model is as follows: Under the current cumulative damage level, the number of failure cycles of the material under the same stress amplitude is equal to the initial number of failure cycles multiplied by one minus the current total cumulative damage level to the power of k. For each extracted load cycle, its pressure parameters are first converted into stress parameters, its local stress range and mean stress are calculated, and then it is equivalent to a symmetrical cyclic stress amplitude by the modified Goodman formula, which states that the equivalent stress amplitude is equal to the stress amplitude divided by one minus the ratio of the mean stress to the tensile strength of the material. The damage increment caused by this cycle is calculated by dividing 1 by the current failure cycle number calculated based on the dynamic SN curve model; The total cumulative damage is updated to the original total cumulative damage plus the damage increment calculated this time.

[0011] The present invention further specifies that the stress transformation process in the dynamic damage accumulation step is as follows: According to the thin-walled cylinder theory, the pressure range is converted into the local stress range of the key parts of the valve body. The calculation formula is that the local stress range is equal to the stress concentration factor multiplied by the product of the pressure range and the average diameter of the flow channel, and then divided by twice the wall thickness. The formula for calculating average stress is: average stress equals the stress concentration factor multiplied by the product of average pressure and average channel diameter, and then divided by twice the wall thickness.

[0012] The present invention further provides that the method for determining the adaptive failure threshold includes: Collect a set of statistically significant fatigue life test data or field failure case data of similar safety valves; From this set of data, the actual cumulative damage degree of each valve at the time of failure is extracted to form a failure damage degree sample set; Assuming the failure damage follows a log-normal distribution, the parameters of this distribution are fitted using the maximum likelihood estimation method; Based on the fitted distribution function, at least two levels of failure thresholds are defined: a warning threshold, which corresponds to a lower cumulative failure probability; and an action threshold, which corresponds to a higher cumulative failure probability. The specific values ​​of the thresholds are obtained by calculating the inverse function of the log-normal distribution.

[0013] The present invention further specifies that the remaining lifetime prediction step specifically includes: Real-time monitoring and calculation of the current total cumulative damage; Based on pressure data from a historical period, the total damage increment during that period is calculated, and then the average damage rate, i.e. the damage increment per unit time, is obtained. Based on the adaptive failure threshold, calculate the remaining damage capacity required for the current damage level to reach the warning threshold and the action threshold; Based on the average damage rate, the remaining lifetime required to reach the warning threshold and the action threshold is predicted, and the value is the remaining damage capacity divided by the average damage rate.

[0014] The present invention further includes a model calibration step: When the target safety valve is subjected to offline calibration or disassembly for repair, its current actual cumulative damage is assessed based on its actual inspection status. Compare the cumulative damage predicted by the model with the actual cumulative damage assessed. By optimizing the algorithm, the key parameters in the model, including the damage sensitivity index k and the material constants in the initial SN curve equation, are adjusted to minimize the deviation between the model prediction and the actual value. The optimized model parameters will be used for subsequent life assessment of the valve.

[0015] The present invention also provides an apparatus for implementing the method, comprising: The data interface module is used to receive in real time the inlet pressure time series data of safety valves from the process control system or to read in batches from the historical database. The signal processing and loop counting module, which has a built-in improved rainflow counting algorithm, is used to perform the load loop extraction step and output a structured load spectrum. The dynamic damage calculation engine encapsulates the dynamic SN curve model and the nonlinear damage accumulation algorithm, and is used to receive the load spectrum and execute the dynamic damage accumulation steps to output the real-time accumulated damage degree. The lifespan prediction and reporting module stores the adaptive failure threshold and generates a report containing the current health status, cumulative damage, predicted remaining lifespan, and maintenance recommendations based on the remaining lifespan prediction step. The human-computer interaction interface is used to visually display evaluation results and historical trends.

[0016] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the method when executing the program.

[0017] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps of the method.

[0018] The beneficial effects of this invention are as follows: By utilizing pressure data, the implementation threshold and cost are significantly reduced, making it easy to promote in existing factory systems. Through the "dynamic SN curve," the continuous degradation of material properties is considered in the life assessment of safety valves, breaking the limitations of traditional linear cumulative damage theory. The nonlinear damage model can more accurately describe the accelerated damage process of valves in the later stages of service, and its early warning capability is significantly superior to traditional methods. Based on probabilistic statistics, failure thresholds and remaining life predictions provide a more scientific and flexible quantitative basis for maintenance decisions, achieving true predictive maintenance. Through model calibration, the evaluation model can adapt to the individual differences of specific valves and actual operating conditions, becoming more accurate with use. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating Embodiment 1 of the present invention.

[0020] Figure 2 This is a schematic diagram of the dynamic SN curve evolution according to an embodiment of the present invention. Detailed Implementation

[0021] The following will describe in detail the implementation of this application with reference to the accompanying drawings and embodiments, so that the implementation process of how this application uses technical means to solve technical problems and achieve technical effects can be fully understood and implemented accordingly.

[0022] Example 1

[0023] like Figures 1-2 As shown, this invention provides a method for assessing the service life of safety valves based on pressure history statistics and dynamic damage accumulation, which includes the following steps: Load history acquisition steps: Obtain pressure time series data at the safety valve inlet. , where t is time; Load cycle extraction step: Process the pressure time series data to identify and extract the complete set of pressure load cycles. Each cycle Due to its pressure range Mean pressure and loop type identifier Describing it using (main loop or secondary loop); Dynamic damage accumulation step: Based on the extracted load cycles, a dynamically degenerate material SN curve model is used to calculate the damage caused by each load cycle. The total cumulative damage degree D is obtained by accumulating the values, where, in the dynamic SN curve model, the material is subjected to a specific stress amplitude. Failure cycle count It is a function of the current total cumulative damage D, and the relationship is expressed as:

[0024] in, For the material in an initially undamaged state under stress amplitude The number of failure cycles, where k is a damage sensitivity index greater than 1; Remaining lifetime prediction steps: Based on the current total cumulative damage level And combined with an adaptive failure probability distribution obtained from historical failure data statistics. Predicting that the safety valve will reach a predetermined failure probability threshold. The remaining service life (RUL).

[0025] The load cyclic extraction step employs an improved rainflow counting method, specifically including: Pressure time series Filtering and denoising are performed to obtain a smooth signal. ; Identification signal All local peak point sequences Valley value sequence ; According to the set pressure of the safety valve Set an action detection threshold Where α is a proportionality coefficient, ranging from 0.8 to 0.95, and the marking pressure exceeds... The period during which the action phase occurs is the active phase, and the rest is the resting phase; During the action phase, a complete pressure is increased from below... Rise to peak Then it descends to the reseating pressure The process is defined as a main loop. Its pressure range ; During the quiescent period, the standard rainflow counting method was applied to the pressure fluctuation signal to extract all complete sub-cycles. Its pressure range It is determined by the difference between the peak and trough values ​​of the cycle.

[0026] In this embodiment, the dynamic SN curve model is specifically represented as follows: The initial SN curve equation is: Where C and m are material constants, Stress amplitude; The dynamic SN curve model is as follows:

[0027] Where D is the current total cumulative damage, 0≤D<1, and k is the damage sensitivity index, k>1; For each extracted load cycle First, its pressure parameters are converted into stress parameters, and its local stress range is calculated. and average stress Then, by modifying the Goodma formula, it is equivalent to a symmetrical cyclic stress amplitude. ;in ; in, , The tensile strength of the material; The incremental damage caused by this cycle The calculation is as follows: ; Total cumulative damage updated as follows:

[0028] In this embodiment, the stress transformation process in the dynamic damage accumulation step is specifically as follows: Based on the thin-walled cylinder theory, the pressure range Converted to local stress range of key parts of the valve body The formula is: ; in, Where is the average diameter of the valve body flow channel, and t is the wall thickness of the calculated part of the valve body. The stress concentration factor is used to account for geometric discontinuities (such as fillets and holes); Mean stress The calculation formula is: ; in, This represents the average pressure during the load cycle.

[0029] Furthermore, the method for determining the adaptive failure threshold in this embodiment includes: Collect fatigue life test data or field failure case data of N similar safety valves to form a sample set. ,in The cumulative damage degree when the j-th valve fails; Assumption It follows a log-normal distribution, that is Fitting distribution parameters using maximum likelihood estimation and σ; Define warning threshold and action threshold They are respectively: ; ; in, It is the inverse function of the standard normal distribution.

[0030] The remaining lifetime prediction step described in this embodiment specifically includes: Real-time calculation of the current total cumulative damage ; Based on a period of history Calculate the total damage increment within the pressure data during that time period. Thus, the average damage rate is obtained. : ; Predicted to reach the warning threshold and action threshold Required remaining lifespan and : in ;

[0031] This embodiment also includes a model calibration step: When the target safety valve is subjected to offline calibration or disassembly for repair, its actual condition is recorded. If an initial microcrack is found, its current cumulative damage level is determined. If the condition is good, then ; The cumulative damage predicted by the model and Compare and calculate the deviation ; Adjust model parameters by optimizing algorithms. To minimize the sum of squared deviations at all available calibration points: ; Where M represents the number of calibration points, the optimized parameters will be used for the subsequent life assessment of the valve to achieve individualized and accurate prediction.

[0032] Example 2

[0033] This embodiment provides an apparatus for implementing the method, comprising: The data interface module is used to receive in real time the inlet pressure time series data of safety valves from the process control system or to read in batches from the historical database. The signal processing and loop counting module, which has a built-in improved rainflow counting algorithm, is used to perform the load loop extraction step and output a structured load spectrum. The dynamic damage calculation engine encapsulates the dynamic SN curve model and the nonlinear damage accumulation algorithm, and is used to receive the load spectrum and execute the dynamic damage accumulation step to output the real-time accumulated damage degree D. The lifetime prediction and reporting module stores the adaptive failure threshold. and Based on the remaining lifespan prediction step, a report is generated that includes the current health status, cumulative damage level, predicted remaining lifespan, and maintenance recommendations. The human-computer interaction interface is used to visually display evaluation results and historical trends.

[0034] Example 3

[0035] This embodiment also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method.

[0036] Example 4

[0037] This embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method.

[0038] The present invention will now be described in further detail. The method in this embodiment 1 mainly includes four steps: data acquisition and preprocessing, load spectrum analysis, dynamic damage accumulation, and remaining life prediction.

[0039] The data acquisition and preprocessing methods are as follows: The data source comes from the pressure transmitter at the inlet of the safety valve. The data is acquired through the device's DCS or SACDA system. The following requirements apply when acquiring the data: the sampling frequency of the pressure time series is not less than 1Hz to ensure effective identification of pressure fluctuations and rapid opening events; the preprocessing method is to perform moving average filtering or low-pass filtering on the raw pressure signal to suppress high-frequency noise interference and obtain a smooth signal for subsequent analysis. The size of the filtering window is selected according to the fluctuation characteristics of the actual pressure signal, usually 0.1s-1s.

[0040] The load spectrum analysis method is as follows: An improved rainflow counting method is used to adapt it for pressure signals and correlate it with safety valve actions. The specific operation is as follows: First, peak / valley detection is performed: on the smoothed pressure signal, all local maxima (peaks) and local minima (valleys) are identified to form a peak sequence. Valley value sequence .

[0041] Action event recognition: Set an action judgment threshold. This threshold is typically the set pressure. 85%~95%, that is α∈[0.85,0.95]. When the pressure exceeds... At this point, the safety valve is considered to have started to open or is in the open state. Based on this, the pressure time history is divided into the "action period" and the "stationary period".

[0042] Loop refactoring includes the main loop and the secondary loop, as detailed below: Major Cycles: Correspond to the complete opening and closing action of the safety valve. During the operating period, the valve operates from a pressure below... Starting from the trough point, it rises to the peak point. Then it drops to the pressure of the return seat. (usually slightly lower) The process constitutes a main loop. Its pressure range .

[0043] Minor Cycles: During the quiescent period, the pressure fluctuates around a certain mean. These fluctuation signals are analyzed using the standard four-peak-valley rainflow counting method to extract all complete stress cycles, which are then denoted as minor cycles. Its pressure range This is the difference between the peak and valley values ​​of the cycle.

[0044] The output will be a set containing all load cycles. Each loop Using triplet ( ) description, in which It is a loop type.

[0045] Subsequently, a dynamic damage accumulation model was established, which includes stress transformation and dynamic SN curve models.

[0046] In the stress transformation process, the pressure cycle is converted into a local stress cycle in key parts of the valve body, such as the inner fillet of the valve body-valve cover connection or the valve seat support area. The stress range calculation is based on thin-walled cylinder theory and considers stress concentration effects, yielding the following formula: , among which Where is the average diameter of the valve body flow channel, and t is the wall thickness of the calculated part. The stress concentration factor is the one obtained from theory or experiment.

[0047] The mean stress is as follows: The mean stress is corrected by using the modified Goodman formula, which transforms the asymmetric cycle into a symmetric cycle. ,in , This refers to the tensile strength of the valve body material.

[0048] The initial fatigue properties of the valve material in this invention (taking ASTM A182 F22 as an example) are described by the standard SN curve: Where C and m are material constants, which can be obtained by consulting material handbooks or fatigue experiments. Furthermore, in this invention, the fatigue strength of the material decreases with increasing cumulative damage; therefore, a dynamic SN curve model is introduced: Here, D is the current total cumulative damage (0 ≤ D < 1), and k is the damage sensitivity index (k > 1). This index controls the rate at which the SN curve shifts downward as damage intensifies. A larger k value indicates faster material property degradation, and in the later stages of damage, the lifespan corresponding to the same stress amplitude is shorter. The shorter.

[0049] Damage increment calculation: For each load cycle, the resulting damage increment is... ; The total damage level is updated as follows: ; because The calculation depends on the current Therefore, the damage accumulation process is nonlinear. In the early stages of service (when D is small), This aligns with the traditional Miner's Law. In the later stages of service (when D approaches 1). It becomes very small, making The damage increment is drastically reduced, resulting in a much larger damage increment for the same load cycle. This accurately simulates the physical process by which materials fail more rapidly after damage accumulates.

[0050] In remaining service life prediction, an adaptive failure threshold is used: the failure threshold is not fixed at 1, but is based on a statistical distribution. Damage data of a batch of valves of the same model at failure are collected. Assuming it follows a log-normal distribution, parameters μ and σ are fitted. Then, two key thresholds are defined: the warning threshold and the... and action threshold ,in, , Then, the remaining lifespan is calculated based on the current degree of damage. Based on a period of history Calculate the total damage increment within the pressure data during that time period. Thus, the average damage rate is obtained. Then, the remaining lifetime required to reach the threshold is predicted; In this invention, to improve the prediction accuracy for specific valves, a calibration step is also included. When the valve undergoes offline calibration or maintenance, the calibration is performed based on its actual condition (e.g., if micro-cracks are found). If it is in perfect condition, then it is considered as good as new. ), and model predictions A comparison is then performed. The parameters θ=[k,C,m] are adjusted using an optimization algorithm (such as the Levenberg-Marquardt algorithm) to minimize the deviation between the predicted and actual values. The calibrated model will be used for subsequent, more accurate evaluations of the valve.

[0051] This invention significantly reduces the implementation threshold and cost by utilizing pressure data, making it easy to promote in existing factory systems. Through the "dynamic SN curve," the continuous degradation of material properties is considered in the life assessment of safety valves, breaking the limitations of traditional linear cumulative damage theory. The nonlinear damage model can more accurately describe the accelerated damage process of valves in the later stages of service, and its early warning capability is significantly superior to traditional methods. Based on probabilistic statistics, failure thresholds and remaining life predictions provide a more scientific and flexible quantitative basis for maintenance decisions, achieving true predictive maintenance. Through model calibration, the evaluation model can adapt to the individual differences of specific valves and actual operating conditions, becoming more accurate with use.

[0052] As used in the specification and claims, certain terms refer to specific components. Those skilled in the art will understand that hardware manufacturers may use different names to refer to the same component. This specification and claims do not distinguish components based on differences in name, but rather on differences in function. The term "comprising" throughout the specification and claims is an open-ended term and should be interpreted as "comprising but not limited to." "Approximately" means that within an acceptable margin of error, those skilled in the art can solve the technical problem and substantially achieve the technical effect within a certain margin of error.

[0053] The foregoing description illustrates and describes several preferred embodiments of the present invention. However, as previously stated, it should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for evaluating service life of a safety valve based on pressure history statistics and dynamic damage accumulation, characterized in that, The method comprises the following steps: A load history collecting step: obtaining the pressure time series data at the inlet of the safety valve; A load cycle extracting step: processing the pressure time series data, identifying and extracting a complete set of pressure load cycles, wherein each cycle is described by its pressure range, average pressure and cycle type identification, and the cycle type includes main cycles and secondary cycles; A dynamic damage accumulation step: based on the extracted load cycles, a dynamic degradation material S-N curve model is used to calculate the damage degree caused by each load cycle and accumulate the total accumulated damage degree, wherein in the dynamic S-N curve model, the failure cycle number of the material under a certain stress amplitude is a function of the current total accumulated damage degree; A residual life prediction step: based on the current total accumulated damage degree and in combination with an adaptive failure probability distribution obtained based on historical failure data statistics, the residual service life of the safety valve reaching a predetermined failure probability threshold is predicted.

2. The method for evaluating service life of safety valve based on pressure history statistics and dynamic damage accumulation according to claim 1, characterized in that, The load cycle extracting step uses an improved rainflow counting method, which comprises: Filtering and denoising the pressure time series, identifying all local peak points and valley points in the denoised signal, setting an action determination threshold according to the set pressure of the safety valve, which is the set pressure multiplied by a proportional coefficient, marking the period when the pressure exceeds the action determination threshold as the action period, and the rest as the static period, in the action period, defining a complete pressure from rising from below the action determination threshold to the peak value and then falling to the closing pressure as a main cycle, and the pressure range of the main cycle is the difference between the peak value and the closing pressure, in the static period, applying the standard rainflow counting method to the pressure fluctuation signal to extract all complete secondary cycles, and the pressure range of the secondary cycle is determined by the difference between the peak value and the valley value of the cycle.

3. The method for evaluating service life of safety valve based on pressure history statistics and dynamic damage accumulation according to claim 1, characterized in that, The dynamic S-N curve model is represented as: the initial S-N curve equation is that the failure cycle number of the material under a certain stress amplitude in the initial undamaged state is equal to a material constant multiplied by the negative mth power of the stress amplitude, wherein m is another material constant; The dynamic S-N curve model is: under the current accumulated damage degree, the failure cycle number of the material under the same stress amplitude is equal to the initial failure cycle number multiplied by a minus kth power of the current total accumulated damage degree; For each extracted load cycle, first convert its pressure parameters to stress parameters, calculate its local stress range and average stress, and then convert it to an equivalent symmetric cycle stress amplitude by the modified Goodman formula, which is equivalent stress amplitude equal to stress amplitude divided by a minus the ratio of average stress and material tensile strength; The damage increment caused by the cycle is calculated as 1 divided by the current failure cycle number calculated according to the dynamic S-N curve model; The total accumulated damage degree is updated to the original total accumulated damage degree plus the damage increment calculated this time.

4. The method for evaluating service life of safety valve based on pressure history statistics and dynamic damage accumulation according to claim 3, characterized in that, The stress conversion process in the dynamic damage accumulation step is: According to the thin-walled cylinder theory, the pressure range is converted to the local stress range of the key parts of the valve body, and the calculation formula is local stress range equal to stress concentration coefficient multiplied by the product of pressure range and average diameter of flow passage, and then divided by twice the wall thickness; The formula for calculating the average stress is: average stress equals stress concentration factor multiplied by the product of average pressure and average diameter of the flow passage, divided by twice the wall thickness.

5. The method for evaluating service life of safety valve based on pressure history statistics and dynamic damage accumulation according to claim 1, characterized in that, The method for determining the adaptive failure threshold comprises: collecting fatigue life experimental data or field failure case data of a group of safety valves of the same type with statistical significance; extracting the actual cumulative damage degree of each valve at the time of failure from the group of data to form a failure damage degree sample set; assuming that the failure damage degree follows a lognormal distribution, fitting the parameters of the distribution using the maximum likelihood estimation method; defining at least two levels of failure threshold values according to the fitted distribution function: a pre-warning threshold value corresponding to a lower cumulative failure probability; and an action threshold value corresponding to a higher cumulative failure probability; the specific values of the threshold values are obtained by the inverse function calculation of the lognormal distribution.

6. The method for evaluating service life of safety valve based on pressure history statistics and dynamic damage accumulation according to claim 5, characterized in that, The residual life prediction step specifically comprises: monitoring and calculating the current total cumulative damage degree in real time; calculating the total damage increment in a certain period of time based on the pressure data in the period of time, and then obtaining the average damage rate, i.e. the damage increment per unit time; calculating the residual damage capacity required for the current damage degree to reach the pre-warning threshold value and the action threshold value according to the adaptive failure threshold; predicting the residual life required for reaching the pre-warning threshold value and the action threshold value based on the average damage rate, the value being the residual damage capacity divided by the average damage rate.

7. The method for evaluating service life of safety valve based on pressure history statistics and dynamic damage accumulation according to claim 1, characterized in that, It also includes a model calibration step: when the target safety valve is offline checked or disassembled for maintenance, evaluating its current real cumulative damage degree according to its actual inspection state; comparing the cumulative damage degree predicted by the model with the evaluated real cumulative damage degree; adjusting the key parameters in the model, including the damage sensitivity index k and the material constant in the initial S-N curve equation, through an optimization algorithm to minimize the deviation between the predicted value and the real value; the optimized model parameters will be used for subsequent life assessment of the valve.

8. An apparatus for implementing the method of any one of claims 1-7, characterized by It comprises: a data interface module for receiving the inlet pressure time series data of the safety valve in real time from the process control system or reading the data in batches from the historical database; a signal processing and cycle counting module with an improved rainflow counting algorithm built-in, for executing the load cycle extraction step and outputting a structured load spectrum; a dynamic damage calculation engine encapsulating the dynamic S-N curve model and the nonlinear damage accumulation algorithm, for receiving the load spectrum and executing the dynamic damage accumulation step to output the real-time cumulative damage degree; a life prediction and reporting module storing the adaptive failure threshold and generating a report containing the current health status, cumulative damage degree, predicted residual life and maintenance recommendations based on the residual life prediction step; a human-computer interaction interface for visualizing the evaluation results and historical trends.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the steps of the method of any one of claims 1 to 7 when executing the program.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the steps of the method of any one of claims 1 to 7 when executed by the processor.

Images