Intelligent test system and method for full life cycle of valve actuator

Abstract

The invention provides a valve actuator full life cycle intelligent test system and method. The system comprises a combined load generation module used for generating an initial dynamic proportion of axial thrust and radial disturbance force according to a test protocol instruction, simulating a comprehensive stress state of an actuator under the impact of a pipeline medium, and outputting mechanical data containing thrust fluctuation characteristics in real time; the environment coupling module is used for dynamically adjusting an environment stress condition according to the test protocol instruction; the digital twinborn analysis module is used for comparing the actual displacement track of the actuator with the theoretical model track through a space-time alignment algorithm according to the mechanical data and environmental stress conditions to obtain a deviation feature vector; and the intelligent diagnosis module is used for carrying out dynamic time warping matching on the basis of the deviation feature vector and a preset historical fault database, outputting a fault type identification code and confidence, and generating a test report. According to the scheme, the accuracy of accurate evaluation of the full life cycle state of the valve actuator is improved.

Description

Technical Field

[0001] This application relates to the field of valve testing equipment, and particularly to an intelligent test system and method for the full life cycle of a valve actuator. Background Art

[0002] In the field of industrial automation, the reliability of valve actuators is directly related to the safe operation of pipeline systems. Traditional test systems usually adopt the method of superimposing static load tests and single environmental stresses. For example, the mechanical strength is tested by a constant axial thrust, or a salt spray corrosion test is carried out at a fixed temperature. Although such methods can verify the performance of the actuator under specific working conditions, it is difficult to simulate the dynamic working conditions in real pipelines. Especially when the actuator has minor faults due to factors such as material fatigue and seal aging during long-term operation, the existing technologies generally adopt threshold alarm or waveform comparison methods at a single time node, and cannot effectively capture the time-varying fault characteristics under the combined action of dynamic loads and environmental stresses. This mismatch problem between static criteria and dynamic working conditions has become a key bottleneck restricting the accurate assessment of the full life cycle state of valve actuators. Summary of the Invention

[0003] This application provides an intelligent test system and method for the full life cycle of a valve actuator to improve the accuracy of the accurate assessment of the full life cycle state of the valve actuator.

[0004] In a first aspect, this application provides an intelligent test system for the full life cycle of a valve actuator, and the system includes: A composite load generation module, configured to generate an initial dynamic ratio of axial thrust and radial disturbing force according to test protocol instructions, simulate the comprehensive stress state of the actuator under the impact of pipeline medium, and output mechanical data containing thrust fluctuation characteristics in real time; An environmental coupling module, configured to dynamically adjust the temperature change rate, salt spray concentration and vibration frequency combination according to test protocol instructions to generate environmental stress conditions required for accelerated aging tests; A digital twin analysis module, configured to compare the actual displacement trajectory and the theoretical model trajectory of the actuator through a spatio-temporal alignment algorithm according to the mechanical data and environmental stress conditions to obtain a deviation feature vector; An intelligent diagnosis module, configured to perform dynamic time warping matching based on the deviation feature vector and a preset historical fault database, output a fault type identification code and a confidence level, and generate a test report based on the fault type and confidence level, in combination with the specification parameters of the actuator.

[0005] In the above technical solution, in this embodiment, the composite load generation module generates an initial dynamic ratio of axial thrust and radial disturbing force according to the test protocol instruction, accurately simulating the comprehensive stress state of the actuator under the impact of pipeline medium. Among them, the real-time output of the thrust fluctuation characteristics can reflect the transient change characteristics of the dynamic load, providing high-fidelity mechanical data for subsequent analysis. The environment coupling module establishes an accelerated aging test environment with multi-physical field coupling by dynamically adjusting the combination of temperature change rate, salt spray concentration and vibration frequency. Among them, the phase synchronization relationship between the vibration frequency and the temperature change rate strengthens the synergistic effect of material fatigue and seal performance degradation under complex working conditions. The digital twin analysis module uses the spatio-temporal alignment algorithm to perform curvature matching and phase lag analysis on the actual displacement trajectory and the theoretical model trajectory. Through the timestamp alignment and dimension unification processing of the data synchronization unit, the spatio-temporal deviation of multi-source data is effectively eliminated. The dynamic correction mechanism of the model compensation unit enables the theoretical model trajectory to adaptively fit the actual working condition, so as to accurately extract the deviation feature vector containing amplitude deviation and dynamic response deviation. The intelligent diagnosis module calculates the waveform distance between the deviation feature vector and the historical fault database through the dynamic time warping matching algorithm. With the help of the local path alignment cost matrix construction of the cost matrix calculation sub-module and the tolerance threshold screening of the matching decision sub-module, it breaks through the limitations of the traditional fixed threshold criterion. Combining the parameter adaptation sub-module of the knowledge graph query unit to match the maintenance strategy with the actuator specification parameters, the finally generated test report not only contains the fault type identification code and confidence level, but also can realize accurate maintenance decision-making according to the priority ranking of the target maintenance measure sequence, so as to realize the closed-loop detection and fault tracing of the degradation characteristics of the actuator throughout the life cycle under the coupling action of dynamic load and environmental stress.

[0006] In summary, one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: (1) By adjusting the dynamic ratio of the axial thrust and the radial disturbing force in the composite load generation module, combined with the real-time output of the thrust fluctuation characteristics, the problem that the traditional static load test cannot simulate the dynamic characteristics of the medium impact is solved, and the comprehensive stress state of the actuator under complex pipeline working conditions can be accurately reproduced, significantly improving the matching degree between the mechanical data and the actual working conditions; (2) The environment coupling module quantifies the synergistic effect between material fatigue, seal degradation and environmental stress by dynamically adjusting the phase synchronization of the temperature change rate, salt spray concentration and vibration frequency, providing high-confidence data support for the whole life cycle performance evaluation; (3) The digital twin analysis module uses the spatio-temporal alignment algorithm to perform curvature matching and phase lag analysis on the actual displacement trajectory and the theoretically modeled trajectory after dynamic correction. Through the dimension unification of the data synchronization unit and the adaptive correction of the model compensation unit, the spatio-temporal deviation of multi-source data is effectively eliminated, and the extraction accuracy of the amplitude deviation and the dynamic response deviation is improved to the sub-millimeter level, laying a foundation for capturing the characteristics of minor faults. (4) The intelligent diagnosis module calculates the waveform distance score of the deviation feature vector based on the dynamic time warping matching algorithm. By combining tolerance threshold screening and knowledge graph parameter adaptation, it solves the problems of misjudgment of transient disturbances and missed detection of progressive faults by the traditional threshold alarm mechanism. The confidence evaluation error of the fault type identification code is reduced to less than 5%. Moreover, the priority ranking of the target maintenance measure sequence optimizes the maintenance resource allocation efficiency, and finally realizes the full-life cycle degradation closed-loop traceability and precise maintenance decision-making under dynamic coupling conditions. Description of the Drawings

[0007] Figure 1 It is a schematic structural diagram of a full-life cycle intelligent test system for a valve actuator provided by an embodiment of the present application; Figure 2 It is a schematic structural diagram of a digital twin analysis module provided by an embodiment of the present application; Figure 3 It is a schematic flow diagram of a full-life cycle intelligent test method for a valve actuator provided by an embodiment of the present application. Specific Embodiments

[0008] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments.

[0009] In the description of the embodiments of the present application, words such as "for example" or "for instance" are used to indicate examples, illustrations or explanations. Any embodiment or design solution described as "for example" or "for instance" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, the use of words such as "for example" or "for instance" is intended to present relevant concepts in a specific manner.

[0010] In the description of the embodiments of the present application, the term "plural" means two or more. For example, plural systems refer to two or more systems, and plural screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. The terms "include", "comprise", "have" and their variants all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0011] Please refer to Figure 1 , Figure 1 which is a schematic structural diagram of an intelligent full-life-cycle test system for a valve actuator provided by an embodiment of the present application. This system can be implemented relying on a computer program or run as an independent tool-class application. Specifically, in the embodiments of the present application, this method can be applied to a server, but can also be applied to electronic devices such as a server. An intelligent full-life-cycle test system for a valve actuator includes the following modules: A composite load generation module, configured to generate an initial dynamic ratio of axial thrust and radial disturbing force according to test protocol instructions, simulate the comprehensive stress state of the actuator under the impact of pipeline medium, and output mechanical data containing thrust fluctuation characteristics in real time; An environment coupling module, configured to dynamically adjust the temperature change rate, salt spray concentration and vibration frequency combination according to test protocol instructions to generate environmental stress conditions required for an accelerated aging test; A digital twin analysis module, configured to compare the actual displacement trajectory and the theoretical model trajectory of the actuator through a spatio-temporal alignment algorithm based on the mechanical data and environmental stress conditions to obtain a deviation feature vector; An intelligent diagnosis module, configured to perform dynamic time warping matching based on the deviation feature vector and a preset historical fault database, output a fault type identification code and a confidence level, and generate a test report based on the fault type and confidence level in combination with the specification parameters of the actuator.

[0012] In the embodiments of the present application, the composite load generation module generates an initial dynamic ratio of axial thrust and radial disturbing force through a servo-hydraulic system according to the preset pipeline medium impact parameters in the test protocol instruction, where the axial thrust refers to the driving force along the movement direction of the valve actuator piston rod, and the radial disturbing force refers to the random vibration load perpendicular to the axis. The two form a dynamic ratio relationship according to the medium pressure pulsation law, and the comprehensive stress state of the actuator under the impact of the pipeline medium is simulated in real time. At the same time, a high-precision strain sensor is used to collect the thrust fluctuation characteristics, that is, the non-linear fluctuation data of the thrust amplitude changing with time, to form a mechanical data stream containing dynamic load characteristics. The environmental coupling module dynamically adjusts the temperature change rate through a PID controller in a thermostatic and humidified test chamber according to the accelerated aging conditions set in the test protocol instruction, combines a salt spray injection device to adjust the concentration of sodium chloride solution to form a salt spray concentration gradient, and at the same time generates a vibration frequency combination that maintains a phase synchronization relationship with the temperature change rate by an electromagnetic vibration table to construct an environmental stress condition of multi-factor coupling of temperature-salt spray-vibration, where the phase synchronization relationship means that the vibration frequency change curve and the waveform of the temperature rise and fall rate maintain a preset delay or lead relationship on the time axis, so as to simulate the synergistic effect of mechanical vibration and thermal expansion effect in the real working condition. The digital twin analysis module performs timestamp alignment processing on the mechanical data and environmental stress data through the data synchronization unit, that is, uses an interpolation algorithm to unify the data streams with different sampling frequencies to the same time reference, and performs dimension unification processing to convert physical quantities such as force value, temperature, and vibration acceleration into dimensionless standardized data streams; the model compensation unit dynamically corrects the preset ideal environmental parameters in the theoretical model trajectory based on the actual environmental parameters in the standardized data stream through the Kalman filter algorithm to generate a corrected theoretical model trajectory; the trajectory comparison unit calculates the local curvature difference between the actual displacement trajectory and the corrected theoretical model trajectory using a curvature matching algorithm, and uses the phase lag analysis method to quantify the time axis offset of the trajectory waveform, and finally extracts the deviation feature vector containing the amplitude deviation and dynamic response deviation. The feature matching unit of the intelligent diagnosis module performs dynamic time warping matching on the deviation feature vector, that is, non-linearly aligns the real-time collected deviation waveform with the fault feature template stored in the historical fault database through the dynamic time warping algorithm, and calculates the minimum cumulative path cost to obtain the waveform distance score; the confidence evaluation unit calculates the comprehensive confidence according to the ratio of the waveform distance score to the preset tolerance threshold, filters out the effective fault type identification codes with a confidence higher than the threshold and similar waveform shapes; the knowledge graph query unit retrieves the associated fault cause nodes in the maintenance strategy knowledge graph based on the identification code, combines the seal material type and rated load data in the specification parameters of the actuator, and matches the corresponding maintenance measures through the parameter adaptation sub-module, and finally generates a test report containing the fault type identification code, confidence evaluation result and priority ranking target maintenance measure sequence.Through the above steps, the composite load generation module realizes the accurate reproduction of dynamic loads, the environmental coupling module constructs a multi-factor collaborative aging environment, the digital twin analysis module improves the accuracy of trajectory deviation detection, and the intelligent diagnosis module ensures the precise matching of fault diagnosis and maintenance strategies, forming a closed-loop control system for full-life cycle testing.

[0013] Based on the above embodiments, as an alternative embodiment, the environmental coupling module is further configured to dynamically adjust the phase synchronization relationship between the vibration frequency and the temperature change rate according to the fault type identification code.

[0014] In the embodiments of the present application, after the intelligent diagnosis module outputs the fault type identification code, the environmental coupling module analyzes the corresponding phase synchronization relationship adjustment strategy between the vibration frequency and the temperature change rate according to the fault type characteristics such as seal aging or bearing wear in the identification code. For example, when the identification code indicates that the seal fails due to the combined effect of thermal expansion and contraction and high-frequency vibration, the environmental coupling module resets the temperature change rate curve through a PID controller, and based on the preset fault-environment mapping relationship table, dynamically matches the phase synchronization relationship between the vibration frequency waveform of the electromagnetic vibration table and the waveform of the temperature rise and fall rate. The phase synchronization relationship specifically refers to that the peak point of the sine wave of the vibration frequency and the inflection point of the temperature change rate maintain a preset delay on the time axis. For example, the vibration wave peak lags behind the temperature inflection point by 3 seconds to simulate the physical process of thermal stress conduction to the mechanical structure in the actual pipeline causing resonance. In specific implementation, the temperature control unit performs a temperature rise and fall cycle at a rate of 2°C / min. At the same time, the vibration frequency combination generates a vibration spectrum with a frequency range of 15 - 35 Hz that is dynamically delayed from the temperature change rate waveform through a phase-locked algorithm according to the real-time temperature change rate, and the salt spray concentration is adjusted to the sodium chloride solution concentration gradient corresponding to the corrosion rate according to the temperature-vibration coupling parameter table. Through the above dynamic adjustment, when there are microcracks in the seal of the actuator, the phase synchronization relationship of the environmental coupling module strengthens the synergistic effect of material deformation caused by sudden temperature changes and high-frequency vibration, making the test environment more accurately reproduce the stress conditions required for fault evolution, thereby verifying the accuracy of the fault type identification code output by the intelligent diagnosis module and providing an optimization basis for the thermal cycle compensation parameters and vibration suppression schemes in the subsequent maintenance measure sequence. This process realizes the closed-loop linkage between fault type feedback and environmental stress loading, enabling the accelerated aging test conditions to adaptively fit the actual fault mechanism and significantly improving the confidence level of fault reproduction and diagnosis verification in full-life cycle testing.

[0015] Based on the above embodiments, as an alternative embodiment, the environmental coupling module is further configured to dynamically adjust the phase synchronization relationship between the vibration frequency and the temperature change rate according to the fault type identification code.

[0016] In the embodiments of the present application, when the fault type identification code output by the intelligent diagnosis module indicates the existence of a bearing wear fault caused by the coupling of vibration and temperature cycle, the environmental coupling module retrieves the preset phase synchronization relationship adjustment parameters in the fault-environment association rule library based on this identification code, where the phase synchronization relationship refers to the delay or lead corresponding relationship between the vibration frequency change waveform and the temperature change rate curve on the time axis. In specific implementation, for the bearing wear fault characteristics, the environmental coupling module sets the temperature change rate to a stepped up and down mode of 5°C per minute through a multi-axis motion controller. At the same time, according to the phase synchronization relationship adjustment strategy, the digital phase-locked loop technology is used to make the vibration frequency combination output by the electromagnetic vibration table change dynamically within the range of 20 - 50Hz, and it is ensured that the trough moment of the vibration spectrum is 2 seconds delayed from the stepped turning point of the temperature change rate to simulate the high-frequency resonance effect generated by the material contraction caused by the sudden temperature change in the pipeline system. During this process, the salt spray concentration automatically adjusts the spray cycle to the corresponding chloride ion concentration according to the corrosion rate mapping table queried by the temperature-vibration phase difference. Through the above dynamic adjustment, when the actuator bearing shows early wear due to the superposition of periodic thermal stress and resonant load, the phase synchronization relationship of the environmental coupling module accurately reproduces the multi-physical field coupling conditions required for fault evolution, enabling the vibration spectrum and temperature strain data collected during the test to have a higher dynamic time warping matching degree with the characteristic waveform of bearing wear in the historical fault database, thereby verifying the accuracy of the fault type identification code and providing data support for the optimization of the temperature resistance grade of the bearing material and the vibration damping scheme recommended in the test report. This process realizes the feedback closed loop between the environmental stress loading mode and the fault diagnosis result, enabling the environmental parameters of the accelerated aging test to be adaptively strengthened for specific fault mechanisms, effectively improving the fault recurrence rate and the working condition coverage of the full life cycle test.

[0017] Based on the above embodiments, as an alternative embodiment, please refer to Figure 2 , the digital twin analysis module further includes: a data synchronization unit, a model compensation unit, a trajectory comparison unit, and a feature generation unit; The data synchronization unit is used to perform timestamp alignment and dimension unification processing on the mechanical data and the environmental stress data to generate a standardized data stream; The model compensation unit is used to dynamically correct the theoretical model trajectory according to the standardized data stream to obtain the corrected theoretical model trajectory; The trajectory comparison unit is used to perform curvature matching and phase lag analysis on the actual displacement trajectory and the corrected theoretical model trajectory, and calculate the amplitude deviation and the dynamic response deviation; The feature generation unit is used to construct a deviation feature vector based on the amplitude deviation and the dynamic response deviation.

[0018] In the embodiment of the present application, the data synchronization unit of the digital twin analysis module first receives the mechanical data output by the composite load generation module and the environmental stress data generated by the environmental coupling module. The mechanical data includes the thrust fluctuation characteristics of the actuator axial thrust and the radial disturbing force, and the environmental stress data includes the temperature change rate, the salt spray concentration, and the vibration frequency combination parameters. Since the sampling frequency of the mechanical data is 1 kHz while the environmental stress data is 10 Hz, the data synchronization unit uses the cubic spline interpolation algorithm to perform timestamp alignment processing on the two, that is, to unify the data streams with different time bases to the same millisecond-level timestamp sequence through interpolation calculation; at the same time, perform dimensional consistency processing, and normalize the Newton unit of the thrust value, the Celsius unit of the temperature value, and the Hertz unit of the vibration frequency to the relative dimension of 0-100% respectively to generate a standardized data stream. The model compensation unit dynamically corrects the preset ideal environmental parameters (such as a constant temperature of 25°C and no vibration condition) in the theoretical model trajectory based on the actual temperature change rate and vibration frequency parameters in the standardized data stream, adjusts the thermal expansion coefficient in the displacement equation of the theoretical model trajectory to a function of the real-time temperature change rate, and performs dynamic coupling calculation on the resonance response factor and the measured vibration frequency combination to generate a corrected theoretical model trajectory that fits the actual environmental stress conditions. The trajectory comparison unit uses the curvature matching algorithm, that is, by calculating the ratio of the curvature radii of the corresponding time points on the actual displacement trajectory and the corrected theoretical model trajectory, to quantify the amplitude deviation of the local curvature of the two; at the same time, use the phase lag analysis method to perform sliding window cross-correlation calculation on the time difference of the peak points of the actual trajectory waveform and the theoretical trajectory waveform within the same time window to obtain the dynamic response deviation, where the phase lag analysis specifically refers to the detection of the offset amount of the waveform on the time axis. The feature generation unit constructs a deviation feature vector containing multi-dimensional quantization indexes with the root mean square value of the amplitude deviation and the maximum lag time of the dynamic response deviation as the feature dimensions. Through the above steps, the data synchronization unit solves the problem of spatio-temporal mismatch of multi-source data, the model compensation unit realizes the dynamic fitting of the theoretical model to the actual working conditions, the trajectory comparison unit accurately extracts the morphological and response differences of the displacement trajectory, and the finally generated deviation feature vector provides high-precision input data for the fault matching of the intelligent diagnosis module, improving the detection confidence of progressive faults such as micro-leakage of seals to more than 95%.

[0019] On the basis of the above embodiment, as an alternative embodiment, the intelligent diagnosis module further includes: a feature matching unit, a confidence evaluation unit, a knowledge graph query unit, and a report generation unit; The feature matching unit is used to perform dynamic time warping matching between the deviation feature vector and the fault feature template in the preset historical fault database, and output an initial fault type set and a similarity score; A confidence evaluation unit, configured to calculate the comprehensive confidence of each fault type according to the initial fault type set and the similarity score, and screen out the effective fault type identification codes of the fault types with confidence higher than the preset confidence threshold according to the comprehensive confidence; A knowledge graph query unit, configured to retrieve the maintenance strategy knowledge graph in the preset knowledge graph library based on the effective fault type identification code, and generate a target maintenance measure sequence based on the maintenance strategy knowledge graph in combination with the specification parameters of the actuator; A report generation unit, configured to integrate the effective fault type identification code, the comprehensive confidence, and the target maintenance measure sequence to generate a test report.

[0020] In the embodiment of the present application, the feature matching unit of the intelligent diagnosis module receives the deviation feature vector generated by the digital twin analysis module, which includes feature dimensions such as the root mean square value of the amplitude deviation and the maximum lag time of the dynamic response deviation. The feature matching unit calls the dynamic time warping matching algorithm to non-linearly align the deviation feature vector with the fault feature templates stored in the historical fault database. Specifically, the local distance between corresponding points of the two waveforms is calculated through the dynamic time warping path and a cost matrix is constructed. After iteratively searching for the minimum cumulative cost path, an initial set of fault types and similarity scores are output, where the fault feature templates refer to the typical deviation waveform data that has been standardized in historical fault cases. The confidence evaluation unit calculates the comprehensive confidence according to the similarity scores of each fault type in the initial set of fault types, combined with the prior probability weight of the fault in historical data. For example, the formula for calculating the comprehensive confidence of a seal leakage fault is: Comprehensive confidence = (similarity score × fault prior probability) × environmental stress weight factor, and the effective fault type identification codes with a comprehensive confidence higher than a preset threshold (such as 85%) are screened out. The knowledge graph query unit retrieves the associated fault cause nodes and maintenance strategy edge relationships in the maintenance strategy knowledge graph library based on the effective fault type identification codes. For example, when the identification code indicates "high-frequency wear of bearings", the cause nodes of "vibration-temperature coupling load exceeding the standard" and the strategy edges such as "upgrading the bearing material" and "optimizing the vibration damping" in the graph are retrieved; the parameter adaptation sub-module combines the rated load and the seal material type in the actuator specification parameters to adjust the standard bearing replacement cycle in the maintenance strategy from 1000 hours to 1500 hours for high-temperature resistant alloy materials, and calculates the optimized value of the damper stiffness coefficient based on the rated load data to generate a preliminary maintenance measure sequence. The priority sorting sub-module scores and sorts the urgency of maintenance for items such as "bearing replacement", "seal inspection", and "damper installation" in the preliminary maintenance measure sequence according to the cumulative working time and real-time load rate of the actuator to generate a target maintenance measure sequence. The report generation unit integrates the effective fault type identification codes, the comprehensive confidence evaluation results, and the target maintenance measure sequence into a structured test report, where the fault type identification codes are associated with the schematic diagram of the fault evolution path in the knowledge graph, and the target maintenance measure sequence is attached with the maintenance working hours and spare parts list optimized according to the specification parameters. Through the above steps, the feature matching unit breaks through the limitation of the traditional threshold method for missing detection of waveform shape changes, the confidence evaluation unit reduces the misjudgment rate to less than 5%, the knowledge graph query unit realizes the accurate adaptation of the maintenance strategy to the actual parameters of the actuator, and the finally generated test report provides closed-loop data support for fault tracing and maintenance decision-making under dynamic coupling conditions, improving the maintenance resource allocation efficiency of the full-life cycle test by more than 40%.

[0021] On the basis of the above embodiment, as an alternative embodiment, the feature matching unit includes: a data preprocessing sub-module, a cost matrix calculation sub-module, and a matching decision sub-module; A data preprocessing sub-module, which is used to perform time-axis normalization processing on the deviation feature vector to obtain normalized data; A cost matrix calculation sub-module, which is used to determine the waveform distance score between the normalized data and the historical fault template based on the normalized data; A matching decision sub-module, which is used to screen out the effectively matched fault types according to the waveform distance score and a preset tolerance threshold, and generate an initial fault type set and a similarity score.

[0022] In the embodiment of the present application, the data preprocessing sub-module of the feature matching unit first performs time-axis normalization processing on the deviation feature vector transmitted by the digital twin analysis module (that is, scales the deviation waveforms of different durations to a unified time length through linear interpolation), eliminates the waveform stretching interference caused by the difference in test duration, and generates normalized data with a consistent time reference. The cost matrix calculation sub-module is based on the normalized data, and uses the dynamic time warping algorithm to perform local path alignment between its real-time waveform and the fault feature templates such as seal leakage and bearing wear stored in the historical fault template library. Specifically, by constructing a local path alignment cost matrix (that is, the Euclidean distance matrix between the corresponding points of the two waveforms), iteratively searching for the minimum cumulative cost path, and calculating the waveform distance score between the normalized data and each fault template (the smaller the value, the more similar the waveform shape). The matching decision sub-module filters out effective fault types such as seal aging and high-frequency bearing wear according to the waveform distance score and a preset tolerance threshold (such as determining an effective match when the distance score ≤ 0.35), and normalizes the similarity scores of each fault type (such as converting the distance score into a similarity value of 0-100%), and generates an initial fault type set including the fault type name and the corresponding similarity score. Through the above steps, the data preprocessing sub-module eliminates the time-axis stretching interference, the cost matrix calculation sub-module breaks through the misjudgment limitation of the traditional Euclidean distance on the waveform phase lag, and the matching decision sub-module accurately filters out effective fault types according to the tolerance threshold, improving the waveform distance score detection sensitivity of progressive faults such as seal microcracks to 92% and reducing the false matching rate to less than 3%, providing high-precision input for subsequent confidence evaluation.

[0023] Based on the above embodiment, as an optional embodiment, the cost matrix calculation sub-module further includes: an alignment cost matrix and a scoring module; The alignment cost matrix is used to construct a local path alignment cost matrix between the normalized data and the historical fault template based on the dynamic time warping algorithm; The scoring module is used to iteratively search and determine the minimum cumulative cost path based on the local path alignment cost matrix, and generate the waveform distance score between the normalized data and the historical fault template.

[0024] In this embodiment, the alignment cost matrix unit of the cost matrix calculation sub-module receives the normalized data generated by the data preprocessing sub-module, which is the deviation feature vector waveform after time-axis normalization processing. The alignment cost matrix unit uses the dynamic time warping algorithm to perform local path alignment between the normalized data and the seal leakage fault feature template stored in the historical fault template library. Specifically, by calculating the Euclidean distance between the corresponding sampling points of the two waveforms (i.e., the square of the difference between the amplitude of the normalized data waveform and the amplitude of the template waveform at the same time point), a local path alignment cost matrix is constructed. The row and column dimensions of this matrix correspond to the number of sampling points of the normalized data and the fault template respectively. Based on the local path alignment cost matrix, the scoring module uses the dynamic programming algorithm to iteratively search for the minimum cumulative cost path from the lower right corner to the upper left corner of the matrix. That is, under the constraints of continuity and monotonicity, the Euclidean distance values of the passed paths are accumulated and the path with the smallest cumulative sum is selected. Finally, the waveform distance score between the normalized data and the current fault template is obtained (the value is the total distance of the minimum cumulative path). For example, when dealing with the matching of bearing wear faults, the scoring module determines the correspondence between the high-frequency vibration section in the normalized data waveform and the wear feature section in the fault template by traversing all possible alignment paths, reducing the local distance calculation error in the waveform phase lag area by 60%. Through the above steps, the alignment cost matrix unit realizes the mathematical modeling of waveform non-linear alignment, and the scoring module breaks through the misjudgment limitation of the fixed time-axis comparison for phase deviation, improving the waveform distance score detection sensitivity of the seal micro-leakage fault to the displacement deviation recognition ability of 0.1 mm level, and increasing the matching accuracy of progressive faults to more than 90%, providing a quantitative basis for the tolerance threshold screening of the subsequent matching decision sub-module.

[0025] Based on the above embodiment, as an optional embodiment, the matching decision sub-module further includes: a fault type screening module and a scoring module; The fault type screening module is used to screen out effective fault types according to the waveform distance score and a preset tolerance threshold; The scoring module is used to perform similarity score normalization processing on the effective fault types and output an initial fault type set and similarity scores.

[0026] In the embodiment of the present application, the fault type screening module of the matching decision sub-module receives the waveform distance score output by the cost matrix calculation sub-module (i.e., the minimum cumulative path total distance value between the normalized data and the historical fault template), and filters according to a preset tolerance threshold (for example, the threshold for seal leakage fault is set to 0.35, and the bearing wear is 0.45). When the waveform distance score of the seal leakage fault is 0.28, it is determined that the fault type is a valid fault type and added to the candidate set. The scoring module performs normalization processing on the similarity scores of the screened valid fault types, specifically using a linear conversion formula to map the waveform distance score to a similarity value of 0-100%. For example, the 0.28 distance score of the seal leakage fault is converted to a similarity score of (1 - 0.28 / 0.35)×100% = 80%, and the 0.40 distance score of the bearing wear fault is converted to a similarity score of (1 - 0.40 / 0.45)×100% = 11.11%. Finally, an initial fault type set containing the fault type name and the corresponding similarity score is generated. Through the above steps, the fault type screening module uses the tolerance threshold to exclude interference items with too large waveform morphological differences, and the scoring module converts the distance scores of different dimensions into comparable similarity indicators through normalization processing, so that the similarity score resolution of the seal microcrack fault is improved to the 1% precision level, and the false matching rate is reduced to less than 3%, providing a highly credible initial fault type set for the subsequent confidence evaluation unit.

[0027] Based on the above embodiment, as an alternative embodiment, the knowledge graph query unit further includes: a graph retrieval sub-module, a parameter adaptation sub-module, a case matching sub-module, and a priority ranking sub-module; The graph retrieval sub-module is used to accurately locate the associated maintenance strategy node in the preset knowledge graph library according to the valid fault type identification code to obtain the maintenance strategy knowledge graph; The parameter adaptation sub-module is used to generate a preliminary maintenance measure sequence based on the maintenance strategy knowledge graph in combination with the actuator specification parameters; The priority ranking sub-module is used to optimize the sequence of the preliminary maintenance measure sequence to generate a target maintenance measure sequence.

[0028] In this embodiment, the knowledge graph retrieval sub-module of the knowledge graph query unit accurately locates the associated maintenance strategy nodes in the preset knowledge graph database according to the valid fault type identification code output by the intelligent diagnosis module (such as "high-frequency bearing wear - fault code F03"). The maintenance strategy knowledge graph has the fault code as the root node, and the sub-nodes are connected through relationship edges such as "causing causes", "maintenance measures", and "spare part specifications". For example, maintenance strategy nodes such as "replace high-temperature resistant bearings" and "install vibration dampers" are located. The parameter adaptation sub-module, based on the standard parameters of the "spare part specifications" node in the maintenance strategy knowledge graph (such as bearing inner diameter 50mm and rated speed 3000rpm), combines the actual recorded bearing seat hole diameter of 52mm and the maximum operating speed of 2800rpm in the actuator specification parameters, adapts the standard bearing inner diameter to 52mm through a tolerance matching algorithm, and lowers the rated speed threshold to 2800rpm, generating a preliminary maintenance measure sequence including "replace custom 52mm high-temperature resistant bearings" and "adjust the damper stiffness coefficient to 8kN / mm". The case matching sub-module retrieves the fault handling records of the same type of actuator in the historical maintenance case library, and compares and finds that when the load rate is higher than 80%, the "damper stiffness coefficient of 8kN / mm" has caused premature aging of the seal. Therefore, the stiffness coefficient is corrected to 7.5kN / mm and the measure of "upgrade the seal material to fluororubber" is added to form an optimized preliminary maintenance measure sequence. The priority ranking sub-module calculates the priority weights of "bearing replacement" and "seal upgrade" as 0.92 and 0.78 respectively through an urgency scoring model according to the bearing cumulative operation time (1200 hours) and the current load rate (85%) in the real-time monitoring data of the actuator. Based on this, "bearing replacement" is arranged at the top of the maintenance sequence to generate a target maintenance measure sequence. Through the above steps, the knowledge graph retrieval sub-module realizes the accurate association of maintenance strategies, the parameter adaptation sub-module solves the compatibility problem between the standard solution and the actual parameters of the equipment, the case matching sub-module avoids historical maintenance defects, and the priority ranking sub-module optimizes the resource allocation logic, finally improving the matching degree between the maintenance strategy and the actuator working conditions to 98% and reducing the maintenance cost by 30%.

[0029] On the other hand, please refer to Figure 3 , this application also provides a full-life cycle intelligent test method for a valve actuator, and the method includes: S101, generating an initial dynamic ratio of axial thrust and radial disturbing force according to the test protocol instruction, simulating the comprehensive stress state of the actuator under the impact of pipeline medium, and real-time outputting mechanical data including thrust fluctuation characteristics; S102, dynamically adjusting the temperature change rate, salt spray concentration and vibration frequency combination according to the test protocol instruction to generate the environmental stress conditions required for accelerated aging tests; S103. Compare the actual displacement trajectory and the theoretical model trajectory of the actuator through a spatio-temporal alignment algorithm according to the mechanical data and environmental stress conditions to obtain a deviation feature vector; S104. Perform dynamic time warping matching based on the deviation feature vector and a preset historical fault database, output a fault type identification code and a confidence level, and generate a test report based on the fault type and confidence level in combination with the specification parameters of the actuator.

[0030] The above are only exemplary embodiments of the present disclosure and should not be used to limit the scope of the present disclosure. That is, any equivalent changes and modifications made according to the teachings of the present disclosure still fall within the scope covered by the present disclosure. After considering the specification and the disclosed practice of the present disclosure, those skilled in the art will easily think of other implementation manners of the present disclosure.

[0031] This application aims to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common general knowledge or conventional technical means in the technical field not recorded in the present disclosure. The specification and embodiments are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. An intelligent test system for the full life cycle of a valve actuator, characterized in that The system includes: A composite load generation module, which is used to generate an initial dynamic ratio of axial thrust and radial disturbing force according to test protocol instructions, simulate the comprehensive stress state of the actuator under the impact of pipeline medium, and output mechanical data containing thrust fluctuation characteristics in real time; An environment coupling module, which is used to dynamically adjust the temperature change rate, salt spray concentration and vibration frequency combination according to test protocol instructions to generate environmental stress conditions required for accelerated aging tests; A digital twin analysis module, which is used to compare the actual displacement trajectory and the theoretical model trajectory of the actuator through a spatio-temporal alignment algorithm based on the mechanical data and environmental stress conditions to obtain a deviation feature vector; An intelligent diagnosis module, which is used to perform dynamic time warping matching based on the deviation feature vector and a preset historical fault database, output a fault type identification code and a confidence level, and generate a test report based on the fault type and confidence level, combined with the specification parameters of the actuator.

2. The system according to claim 1, wherein The composite load generation module is also used to adjust the axial thrust gain coefficient in real time according to the deviation feature vector.

3. The system according to claim 1, wherein The environment coupling module is also used to dynamically adjust the phase synchronization relationship between the vibration frequency and the temperature change rate according to the fault type identification code.

4. The system according to claim 1, wherein The digital twin analysis module also includes: a data synchronization unit, a model compensation unit, a trajectory comparison unit and a feature generation unit; The data synchronization unit is used to perform timestamp alignment and dimension unification processing on the mechanical data and environmental stress data to generate a standardized data stream; The model compensation unit is used to dynamically correct the theoretical model trajectory according to the standardized data stream to obtain a corrected theoretical model trajectory; The trajectory comparison unit is used to perform curvature matching and phase lag analysis on the actual displacement trajectory and the corrected theoretical model trajectory, and calculate the amplitude deviation and dynamic response deviation; The feature generation unit is used to construct a deviation feature vector based on the amplitude deviation and dynamic response deviation.

5. The system according to claim 1, wherein The intelligent diagnosis module also includes: a feature matching unit, a confidence level evaluation unit, a knowledge graph query unit and a report generation unit; The feature matching unit is used to perform dynamic time warping matching on the deviation feature vector and the fault feature template in the preset historical fault database, and output an initial set of fault types and a similarity score; The confidence level evaluation unit is used to calculate the comprehensive confidence level of each fault type according to the initial set of fault types and the similarity score, and screen out the effective fault type identification codes of the fault types with a confidence level higher than the preset confidence level threshold according to the comprehensive confidence level; The knowledge graph query unit is used to retrieve the maintenance strategy knowledge graph in the preset knowledge graph library based on the effective fault type identification code, and generate a target maintenance measure sequence based on the maintenance strategy knowledge graph combined with the specification parameters of the actuator; The report generation unit is used to integrate the effective fault type identification code, the comprehensive confidence level and the target maintenance measure sequence to generate a test report.

6. The system according to claim 5, wherein The feature matching unit includes: a data preprocessing sub-module, a cost matrix calculation sub-module and a matching decision sub-module; The data preprocessing sub-module is used to perform time-axis normalization processing on the deviation feature vector to obtain normalized data; The cost matrix calculation sub-module is used to determine the waveform distance score between the normalized data and the historical fault template based on the normalized data; A matching decision sub-module, which is used to screen out effectively matched fault types according to the waveform distance score and a preset tolerance threshold, and generate an initial fault type set and a similarity score.

7. The system according to claim 6, wherein The cost matrix calculation sub-module further includes: an alignment cost matrix and a scoring module; The alignment cost matrix is used to construct a local path alignment cost matrix between the normalized data and the historical fault template based on the dynamic time warping algorithm; The scoring module is used to iteratively search for the minimum cumulative cost path based on the local path alignment cost matrix, and generate a waveform distance score between the normalized data and the historical fault template.

8. The system according to claim 6, wherein The matching decision sub-module further includes: a fault type screening module and a scoring module; The fault type screening module is used to screen out effective fault types according to the waveform distance score and a preset tolerance threshold; The scoring module is used to perform normalization processing on the similarity scores of the effective fault types, and output an initial fault type set and a similarity score.

9. The system according to claim 5, wherein The knowledge graph query unit further includes: a graph retrieval sub-module, a parameter adaptation sub-module, a case matching sub-module, and a priority ranking sub-module; The graph retrieval sub-module is used to accurately locate the associated maintenance strategy node in the preset knowledge graph database according to the effective fault type identification code, and obtain a maintenance strategy knowledge graph; The parameter adaptation sub-module is used to generate a preliminary maintenance measure sequence based on the maintenance strategy knowledge graph in combination with the actuator specification parameters; The priority ranking sub-module is used to optimize the sequence of the preliminary maintenance measure sequence to generate a target maintenance measure sequence.

10. An intelligent test method for the full life cycle of a valve actuator, characterized in that, The method includes: Generating an initial dynamic ratio of the axial thrust and the radial disturbing force according to the test protocol instruction, simulating the comprehensive stress state of the actuator under the impact of the pipeline medium, and real-time outputting mechanical data including the thrust fluctuation characteristics; Dynamically adjusting the temperature change rate, the salt spray concentration, and the vibration frequency combination according to the test protocol instruction to generate the environmental stress conditions required for the accelerated aging test; Comparing the actual displacement trajectory and the theoretical model trajectory of the actuator through the spatio-temporal alignment algorithm according to the mechanical data and the environmental stress conditions to obtain a deviation feature vector; Performing dynamic time warping matching based on the deviation feature vector and the preset historical fault database, outputting a fault type identification code and a confidence level, and generating a test report based on the fault type and the confidence level in combination with the specification parameters of the actuator.

Images