Intelligent fault diagnosis method, system and computer program product for solenoid valve test data

Abstract

The application provides an intelligent fault diagnosis method and system for electromagnetic valve test data and a computer program product, relates to the technical field of electromagnetic valve fault diagnosis, and first collects multi-dimensional test data of an electromagnetic valve under a combination of multiple working conditions, covering multiple types of data such as action response; then a data behavior coordination model is constructed based on the behavior coordination relationship between the data, reflecting laws such as synchronous response; abnormal coordination chains deviating from the normal coordination mode are mined from the model, and abnormal performances of each link are determined; the core causes of the faults are deduced reversely through the abnormal coordination chains, and information such as the cause type is determined; finally, executable instruction sequences are generated based on the core causes of the faults and the conduction characteristics of the abnormal coordination chains, and a maintenance system is controlled to perform fault disposal operations. Therefore, the faults can be diagnosed comprehensively and accurately, and the operation reliability of the electromagnetic valve and the system stability are improved.

Description

Technical Field

[0001] This invention relates to the field of solenoid valve fault diagnosis technology, and more specifically, to an intelligent fault diagnosis method, system, and computer program product for solenoid valve test data. Background Technology

[0002] In the wide range of applications of solenoid valves, such as industrial automation control, aerospace, and automotive manufacturing, the reliable operation of solenoid valves is crucial to the stability and safety of the entire system. Solenoid valves operate in complex and variable environments, facing diverse combinations of operating conditions, such as varying pressures, temperatures, flow rates, and electrical control signals. Under these diverse conditions, solenoid valves generate multidimensional test data, including action response data, media interaction data, structural vibration data, environmental impact data, and electrical signal data.

[0003] Currently, the main problems with fault diagnosis methods for solenoid valves are as follows. Firstly, traditional fault diagnosis methods often focus only on a single type of data, such as analyzing only electrical signal data to determine if there is an electrical fault in the solenoid valve, or relying solely on action response data to assess its mechanical performance. However, solenoid valve faults are usually the result of the interaction of multiple factors, and the analysis of a single data type cannot comprehensively and accurately capture fault characteristics, easily leading to misdiagnosis or missed diagnosis. Secondly, existing methods lack in-depth exploration of the behavioral synergistic relationships between multidimensional test data, and cannot effectively identify the synchronous response patterns, mutual triggering mechanisms, and synergistic change patterns of different types of data during operation. This makes it difficult to grasp the overall operating state and fault evolution process of the solenoid valve, thus hindering the accurate and rapid location of the core fault cause and effective fault handling, affecting the maintenance efficiency of the solenoid valve and the reliability of the system. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide an intelligent fault diagnosis method for solenoid valve test data, the method comprising:

[0005] Collect multidimensional test data of the solenoid valve under various operating conditions. The multidimensional test data includes action response data, media interaction data, structural vibration data, environmental impact data, and electrical signal data.

[0006] A data behavior collaboration model is constructed based on the behavioral collaboration relationship between multidimensional test data. The data behavior collaboration model reflects the synchronous response rules, mutual triggering mechanisms and collaborative change patterns of different types of test data during the operation process.

[0007] From the data behavior collaboration model, we can discover abnormal collaboration chains that deviate from the normal collaboration mode. The abnormal collaboration chain includes the collaboration abnormality initiation link, intermediate transmission link, terminal impact link and abnormal collaboration behavior between each link.

[0008] The core cause of the failure is deduced by reverse engineering the abnormal collaborative chain. The core cause of the failure includes the cause type, triggering conditions, collaborative interference mode and initial impact link.

[0009] Based on the transmission characteristics of the core causes of failures and the abnormal collaborative chain, an executable instruction sequence is generated to control the maintenance system to perform corresponding fault handling operations.

[0010] Furthermore, embodiments of the present invention also provide an intelligent fault diagnosis system for solenoid valve test data, comprising:

[0011] A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to execute the above-described intelligent fault diagnosis method for solenoid valve test data by executing the machine-executable instructions.

[0012] In another aspect, embodiments of the present invention also provide a computer program product, the computer program product including machine-executable instructions, the machine-executable instructions being stored in a computer-readable storage medium, the processor of the intelligent fault diagnosis system for solenoid valve test data reading the machine-executable instructions from the computer-readable storage medium, the processor executing the machine-executable instructions, causing the intelligent fault diagnosis system for solenoid valve test data to perform the aforementioned intelligent fault diagnosis method for solenoid valve test data.

[0013] Based on the above, multi-dimensional test data of the solenoid valve under various operating conditions were collected, covering information on action response, media interaction, structural vibration, environmental effects, and electrical signals. This reflects the operating status of the solenoid valve in complex working environments. A data behavior collaboration model was constructed based on the behavioral collaboration relationship between the multi-dimensional test data. This model can reflect the synchronous response law, mutual triggering mechanism, and collaborative change mode of different types of test data during operation, deeply exploring the intrinsic connection between data. Abnormal collaboration chains deviating from the normal collaboration mode were extracted from the data behavior collaboration model, which can present the initiation link, intermediate transmission link, terminal impact link, and abnormal collaboration performance between each link of the collaboration anomaly. The propagation path and impact range of the fault during the operation of the solenoid valve were located. By reverse deducing the core cause of the fault through the abnormal collaboration chain, the cause type, triggering condition, collaborative interference mode, and initial impact link were identified. Based on the transmission characteristics of the core cause of the fault and the abnormal collaboration chain, an executable instruction sequence was generated to control the maintenance system to perform corresponding fault handling operations. This realized the automation and intelligence of fault diagnosis and handling, greatly improved the efficiency and timeliness of fault handling, and effectively ensured the reliable operation of the solenoid valve and the stability of the entire system. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the execution flow of the intelligent fault diagnosis method for solenoid valve test data provided in an embodiment of the present invention.

[0015] Figure 2 This is a schematic diagram of exemplary hardware and software components of an intelligent fault diagnosis system for solenoid valve test data provided in an embodiment of the present invention. Detailed Implementation

[0016] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating an intelligent fault diagnosis method for solenoid valve test data provided in one embodiment of the present invention. The following is a detailed description of this intelligent fault diagnosis method for solenoid valve test data.

[0017] Step S110: Collect multi-dimensional test data of the solenoid valve under various operating conditions. The multi-dimensional test data includes action response data, media interaction data, structural vibration data, environmental effect data, and electrical signal data.

[0018] In this embodiment, an electromagnetic directional valve used to control the flow of hydraulic media on an industrial production line is used as the application object. Multidimensional test data of this electromagnetic valve under various operating conditions are collected. Action response data is acquired through high-precision displacement sensors installed on the movement trajectory of the electromagnetic valve core and limit switches built into the valve body. Specifically, this includes the displacement sequence of the valve core from its initial position to its target position over time, the velocity sequence of the valve core during its movement, and the on / off state sequence of the limit switches. Media interaction data is collected through pressure sensors, flow sensors, and temperature sensors installed at the inlet and outlet of the electromagnetic valve, including the inlet pressure sequence over time, the outlet pressure sequence over time, the media flow rate sequence, and the media temperature sequence. Structural vibration data is collected through triaxial accelerometers attached to different locations on the electromagnetic valve housing, obtaining the vibration acceleration sequence of the housing in the X, Y, and Z axes over time. Environmental impact data is collected through temperature and humidity sensors and air pressure sensors arranged around the electromagnetic valve, including the ambient temperature change sequence, the ambient humidity change sequence, and the ambient air pressure change sequence. Electrical signal data is acquired through current sensors connected in series and voltage sensors connected in parallel on the solenoid valve control circuit. It includes the time-varying sequence of coil current, the time-varying sequence of voltage across the coil, and the pulse width modulation signal sequence of the control signal.

[0019] During data collection, for data involving privacy-sensitive information, such as data that may contain sensitive information like production process parameters, data anonymization techniques are employed. Specifically, sensitive fields in the collected raw data are replaced, and specific process parameter values ​​are replaced with corresponding grade identifiers. Simultaneously, the data is encrypted during transmission and storage. Symmetric encryption algorithms are used to encrypt the transmitted data, and hash algorithms are used to irreversibly transform sensitive data during storage, ensuring privacy protection and preventing data leakage throughout the collection, transmission, and storage processes.

[0020] Step S120: Construct a data behavior collaboration model based on the behavioral collaboration relationship between multidimensional test data. The data behavior collaboration model reflects the synchronous response rules, mutual triggering mechanisms, and collaborative change patterns of different types of test data during the operation process.

[0021] Step S121: Classify the multidimensional test data by type, and record the original data sequence and data acquisition scenario corresponding to each of the action response data, medium interaction data, structural vibration data, environmental action data and electrical signal data. The data acquisition scenario includes the operating load level of the solenoid valve, the physical characteristics of the medium, the ambient temperature and humidity conditions and the electrical power supply status.

[0022] In this embodiment, the collected multidimensional test data is divided into the five types mentioned above. For action response data, the original data sequence includes a set of continuous data points showing the valve core displacement changing over time, a set of continuous data points showing the valve core speed changing over time, and a set of discrete time points showing the on / off state of the limit switch. The operating load level in the data acquisition scenario is divided into three levels: low load, medium load, and high load, based on the number of switching cycles of the solenoid valve per unit time. The physical properties of the medium include its viscosity, density, and compressibility. The environmental temperature and humidity conditions are the ambient temperature and relative humidity values ​​at the time of acquisition. The electrical power supply status includes the stable value and fluctuation range of the power supply voltage. By establishing a correspondence table between data types, original data sequences, and data acquisition scenarios, it is ensured that each type of test data can be accurately associated with its source and the specific scenario information at the time of acquisition.

[0023] Step S122: Extract the time response features of each type of original data sequence. The time response features reflect the response delay of the data after receiving the trigger signal and the change pattern in the response process. They include the delay period from the moment the trigger signal is received to the start of data change, the duration of the data reaching a stable state, and the rate fluctuation performance in the change process.

[0024] Taking the coil current sequence in electrical signal data as an example, when the solenoid valve receives a control signal (i.e., a trigger signal), the coil current begins to rise from its initial value. The trigger signal reception time is determined by detecting the rising edge of the control signal. The time when the data begins to change is determined by setting a current threshold; the moment when the current sequence first exceeds this threshold is the time difference between the two, which is the delay period. The duration during which the data reaches a stable state refers to the length of time after the current rises to a stable operating current value and remains within a certain fluctuation range around that value. This is determined by judging the duration of the current value within the stable threshold range. The rate fluctuation during the change process is reflected by calculating the average rate of change in different time periods during the current rise. The rise process is divided into multiple equally spaced time periods, and the ratio of the change in current to the time interval is calculated for each time period to obtain the rate value for each time period. The set of these rate values ​​constitutes the rate fluctuation performance.

[0025] Step S123: Extract the amplitude correlation features of each type of original data sequence. The amplitude correlation features reflect the correspondence and linkage between the amplitude changes of the data and the amplitude changes of other types of data, including the proportional relationship of amplitude changes, the degree of phase synchronization, and the time difference of the amplitude peak.

[0026] Taking the valve core displacement sequence in the action response data and the outlet pressure sequence in the media interaction data as examples, this paper analyzes the amplitude correlation characteristics between the two. The proportional relationship of amplitude change is determined by calculating the ratio of the change in valve core displacement to the change in outlet pressure within the same time period. Statistical analysis of the ratios over multiple time periods yields the average proportional relationship of the amplitude changes. Phase synchronization is assessed by comparing the changing trends of the valve core displacement sequence and the outlet pressure sequence to determine their synchronization at the start and end times of the rising, falling, and stable phases. If the start time of the rising phase and the end time of the falling phase are essentially the same, the phase synchronization is considered high. The time difference of the amplitude peak occurrence refers to the time interval between the moment when the valve core displacement reaches its maximum value and the moment when the outlet pressure reaches its maximum value. This is obtained by finding the time corresponding to the peak point in each of the two sequences and then calculating the time difference.

[0027] Step S124: Analyze the collaborative behavior between different original data sequences within the same type of test data, and record the process by which a change in one data sequence triggers synchronous adjustments in other data sequences of the same category, including the trigger threshold for adjustment, the proportional relationship of the adjustment magnitude, and the duration of the stable state after adjustment.

[0028] Taking the X-axis and Y-axis vibration acceleration sequences from structural vibration data as examples, this study analyzes the cooperative behavior within similar test data. When the amplitude of the X-axis vibration acceleration sequence exceeds a certain set trigger threshold, it is observed whether the Y-axis vibration acceleration sequence changes accordingly. The trigger threshold is determined by analyzing the minimum amplitude at which the X-axis vibration acceleration triggers a change in the Y-axis vibration acceleration in historical normal operation data. The proportional relationship of the adjustment amplitude is obtained by calculating the ratio of the amplitude change of the Y-axis vibration acceleration sequence to the amplitude change of the X-axis vibration acceleration sequence during the adjustment process. The duration of the stable state after adjustment refers to the length of time the Y-axis vibration acceleration sequence reaches a new stable amplitude and remains so after adjustment. This is achieved by monitoring the amplitude fluctuation of the Y-axis vibration acceleration sequence after adjustment; timing begins when the fluctuation range is within the stable threshold and continues until a change occurs again.

[0029] Step S125: Analyze the cross-category collaborative behavior between different types of test data, record the mutual triggering performance between action response data and medium interaction data, structural vibration data and electrical signal data, and environmental action data and action response data, including the transmission path of the trigger signal, the response delay after triggering, and the linkage law of the response amplitude.

[0030] Taking the cross-class collaborative behavior between action response data and media interaction data as an example, when the valve core in the action response data begins to move (i.e., the action response data changes), it triggers changes in pressure and flow rate in the media interaction data. The transmission path of the trigger signal is that the valve core movement changes the flow area of ​​the valve orifice, thereby causing changes in media pressure and flow rate. The response delay after triggering refers to the time interval between the moment the valve core begins to move and the moment the media pressure or flow rate begins to change, which is obtained by comparing the start times of the changes in the two data sequences. The linkage law of the response amplitude is determined by analyzing the relationship between the change in valve core displacement and the changes in media pressure and flow rate. For example, the larger the valve core displacement, the larger the change in the flow area of ​​the valve orifice, and the larger the change in media flow rate. By statistically analyzing the flow rate changes corresponding to different displacement changes, the linkage law between the two is obtained.

[0031] Step S126: Based on time response characteristics and amplitude correlation characteristics, establish collaborative behavior benchmark rules for various test data, record the collaborative boundaries and response standards between data under normal operating conditions, including the range of collaborative triggering conditions, the allowable range of response delay, and the amplitude linkage ratio standard under different operating conditions.

[0032] Based on the extracted time response and amplitude correlation features, collaborative behavior benchmark rules are established for various test data. Taking the collaborative behavior of environmental impact data and action response data as an example, under different ambient temperature conditions (i.e., different operating conditions), changes in ambient temperature will affect the action response of the solenoid valve. The collaborative triggering condition range is set so that a specific change in the action response data will only be triggered when the ambient temperature changes within a certain range. For example, in the range of 20℃ to 30℃, for every 5℃ increase in temperature, the valve core action delay time increases by a certain value. The allowable response delay range is obtained based on historical normal data statistics, representing the normal fluctuation range of the valve core action delay time under different ambient temperatures. When the actual delay time is within this range, the collaborative behavior is considered normal. The amplitude linkage ratio standard refers to the normal proportional relationship between the change in ambient temperature and the change in valve core displacement, for example, the standard value of the change in valve core displacement for every 1℃ change in temperature.

[0033] Step S127: Analyze the sensitivity of various test data to cooperative perturbation, and record the perception threshold of different data types to cooperative anomalies, that is, the specific manifestations of subsequent chain anomalies caused by data under a certain degree of cooperative deviation.

[0034] Taking the coordination of electrical signal data and structural vibration data as an example, this paper analyzes the perception threshold of electrical signal data for coordination anomalies. When the amplitude of the coil current in the electrical signal data deviates from the normal amplitude linkage ratio standard to a certain extent, it may trigger anomalies in the structural vibration data. By gradually changing the amplitude of the coil current and observing whether abnormal changes occur in the structural vibration data, the degree of deviation of the coil current amplitude when the vibration acceleration amplitude of the structural vibration data exceeds the normal range is the perception threshold of the electrical signal data for this coordination anomaly. The specific manifestation of subsequent chain anomalies is that structural vibration anomalies may further lead to abnormal pressure fluctuations in the medium interaction data, etc.

[0035] Step S128: Quantify the frequency and stability of different collaborative behaviors through statistical analysis methods, assign a collaborative strength label to each collaborative behavior to reflect the closeness of the collaborative relationship, and associate the strength label with the transmission ability and scope of influence of collaborative anomalies.

[0036] Step S1281: Extract all collaborative behavior instances from the historical, normally functioning multidimensional test data, and distinguish between similar collaborative behavior instances and cross-class collaborative behavior instances. Similar collaborative behavior instances refer to collaboration between different sequences within the same data category, while cross-class collaborative behavior instances refer to collaboration between different data categories.

[0037] From the historical normal operation database of solenoid valves, all multidimensional test data segments recording coordinated behavior were selected. For each data segment, it was classified according to the type of test data involved in the coordination, determining whether it was coordination between different sequences within the same data category (such as coordination between X-axis and Y-axis vibration acceleration sequences in structural vibration data) or coordination between different data categories (such as coordination between electrical signal data and motion response data), and marked as instances of similar coordinated behavior and instances of cross-category coordinated behavior, respectively.

[0038] Step S1282: Count and statistically analyze each type of collaborative behavior instance, record the total number of times the collaborative behavior of this type occurs within a preset time period, and record the number of times the solenoid valve's operating condition changes and the type of operating condition within the time period, forming a collaborative behavior frequency statistics result containing operating condition association information.

[0039] A preset time period of one month is set. For the cross-category collaborative behavior instances mentioned above in "Electrical Signal Data - Action Response Data", the total number of occurrences within one month is counted. At the same time, the number of changes in operating conditions, such as the number of changes in the solenoid valve operating load level, the number of adjustments in the physical properties of the medium, and the number of significant changes in ambient temperature and humidity conditions, are recorded within this month. The specific operating condition type before and after each operating condition change (such as from low load to medium load) is also recorded. The total number of collaborative behaviors and these operating condition association information are recorded together to form frequency statistics results.

[0040] Step S1283: Analyze the duration of each cooperative behavior instance, calculate the duration from trigger to end each time the cooperative behavior occurs, exclude abnormal durations caused by sudden changes in operating conditions, and form an effective duration sequence.

[0041] Taking the example of "structural vibration data - X-axis and Y-axis vibration acceleration coordination" in a similar example of coordinated behavior, the time interval from the start of the trigger signal (e.g., the X-axis vibration acceleration reaches the trigger threshold) to the end of the coordinated behavior (e.g., both the X-axis and Y-axis vibration accelerations return to their normal stable state) for each occurrence of this coordinated behavior is the duration. During the statistical process, if the duration of a particular coordinated behavior is abnormally long, and it is found through reviewing the operating condition records at that time that it was caused by a sudden change in the physical properties of the medium (i.e., a sudden change in operating conditions), then that duration is excluded and not included in the valid duration sequence. The durations under other normal conditions constitute the valid duration sequence.

[0042] Step S1284: Calculate the average duration and duration fluctuation range of this type of cooperative behavior based on the effective duration sequence. The average duration reflects the normal maintenance level of the cooperative behavior, and the duration fluctuation range reflects the temporal stability of the cooperative behavior.

[0043] For the effective duration sequence obtained above, sum all the durations in the sequence and then divide by the number of durations to obtain the average duration. The range of duration fluctuation is reflected by calculating the standard deviation of the sequence. The larger the standard deviation, the greater the fluctuation of the duration and the worse the temporal stability of the cooperative behavior; the smaller the standard deviation, the better the temporal stability.

[0044] Step S1285: Analyze the consistency of data response in each instance of collaborative behavior, calculate the fit of each data sequence participating in the collaboration in terms of change magnitude and change rhythm, including the difference ratio of change magnitude and the sum of time differences in change rhythm, and form response consistency parameters.

[0045] Taking the example of collaborative behavior between "action response data and media interaction data," the two data sequences involved in the collaboration are the valve core displacement sequence and the media flow rate sequence. The difference ratio of the change amplitude refers to the ratio of the difference between the change amplitude of the valve core displacement and the change amplitude of the media flow rate during the collaborative behavior to their average change amplitude. This is obtained by calculating the difference in change amplitude at each time point and then averaging the results. The sum of the time differences in the change rhythm refers to the sum of the time differences between the moment when the valve core displacement sequence reaches a certain change amplitude and the moment when the media flow rate sequence reaches the corresponding change amplitude. Summing these time differences yields the consistency of the change rhythm. Combining the difference ratio of the change amplitude and the sum of the time differences in the change rhythm forms the response consistency parameter.

[0046] Step S1286: Standardize the statistical results of the frequency of collaborative behavior, the average duration, the range of duration fluctuation, and the response consistency parameters. Combine the working condition correlation information to establish a collaborative intensity quantification logic and record the proportion of each parameter in the quantification process. Among them, the frequency and consistency parameters are assigned preset main weights in the quantification logic, which directly determine the level of collaborative intensity.

[0047] The parameters obtained above are standardized by converting parameters with different dimensions into dimensionless relative values. For example, the total number of cooperative behaviors is divided by the maximum possible number within the time period to obtain the standardized frequency value. The average duration, duration fluctuation range, and response consistency parameters are standardized using a similar method. Then, considering operating condition correlation information, such as the higher frequency of cooperative behaviors under high load conditions, the frequency under high load conditions is given a higher weight during quantification. When establishing the quantification logic for cooperative strength, the weights for frequency and consistency parameters are set to 0.4 and 0.3, respectively, and the weights for average duration and duration fluctuation range are 0.15 and 0.15, respectively. Based on these weights, the standardized parameters are weighted and summed to obtain the quantified value of cooperative strength.

[0048] Step S1287: Calculate the collaboration strength value of each collaborative behavior according to the quantification logic. The value reflects the tightness and stability of the collaboration relationship.

[0049] Based on the established logic for quantifying collaborative strength, the standardized statistical results of collaborative behavior frequency, average duration, duration fluctuation range, and response consistency parameter are substituted into a weighted summation formula to calculate the collaborative strength value for each collaborative behavior. For example, the standardized parameters for the collaborative behavior of "electrical signal data - action response data" are 0.8 (frequency), 0.7 (average duration), 0.6 (duration fluctuation range), and 0.9 (response consistency parameter). The collaborative strength value calculated according to the weights is 0.8×0.4+0.7×0.15+0.6×0.15+0.9×0.3. The larger this value, the closer and more stable the collaborative relationship.

[0050] Step S1288: Set the classification criteria for the synergy strength identifier. Divide the identifier into different levels according to the distribution range of the synergy strength values. Each level corresponds to a fixed value range. Each level is associated with the corresponding synergy anomaly transmission rate and impact range prediction.

[0051] Based on the calculated coordination strength values ​​of all coordinated behaviors, their distribution ranges are statistically analyzed, and classification criteria are established. For example, the coordination strength values ​​are divided into five levels: Level 1 corresponds to a value range of 0.8 to 1.0, Level 2 corresponds to 0.6 to 0.8, Level 3 corresponds to 0.4 to 0.6, Level 4 corresponds to 0.2 to 0.4, and Level 5 corresponds to 0 to 0.2. Among them, coordination anomalies associated with Level 1 have a fast transmission rate and a large impact range; coordination anomalies associated with Level 5 have a slow transmission rate and a small impact range.

[0052] Step S1289: Adopt a dynamic adjustment method for the coordination strength, taking into account the impact of the solenoid valve's running time on the coordination relationship, i.e., an increase in running time may lead to a decrease in coordination strength, set an attenuation coefficient, and correct the coordination strength values ​​for different operating stages.

[0053] Considering that with increased operating time, internal components of the solenoid valve may experience wear and aging, leading to a decrease in the tightness and stability of the synergy, i.e., a decrease in synergy strength. A decrease coefficient is set that is related to operating time; the longer the operating time, the larger the decrease coefficient. For example, the decrease coefficient is 1.0 for operating time less than 1000 hours; 0.9 for operating time between 1000 and 2000 hours; and 0.8 for operating time exceeding 2000 hours. Multiplying the original synergy strength value by the corresponding decrease coefficient yields the corrected synergy strength value.

[0054] Step S12810: Compare the corrected collaboration strength value of each collaborative behavior with the classification criteria, assign a corresponding collaboration strength identifier to it, and bind the identifier to the triggering condition and response delay parameter of the collaborative behavior.

[0055] The corrected collaboration strength value is compared with the aforementioned collaboration strength identification criteria to determine the level to which the collaboration behavior belongs, i.e., to assign it a corresponding collaboration strength identification. Simultaneously, this collaboration strength identification is bound to the triggering conditions of the collaboration behavior (such as the threshold range of the trigger signal) and response delay parameters (such as the allowable range of normal response delay), and stored in the relevant parameter library of the data behavior collaboration model.

[0056] Step S12811: Bind the coordination strength identifier to the corresponding coordination behavior to form a coordination feature list that includes coordination behavior type, occurrence frequency, operating condition association information, stability parameters, coordination strength identifier and attenuation coefficient.

[0057] All relevant information on the aforementioned collaborative behaviors is compiled into a collaborative feature list. Each record in the list corresponds to a type of collaborative behavior, including the frequency of occurrence of the collaborative behavior (from frequency statistics), operating condition association information (such as under which operating conditions it is likely to occur), stability parameters (average duration and duration fluctuation range), collaborative strength identifier, and attenuation coefficient, etc., to facilitate subsequent model construction and analysis.

[0058] Step S129: Based on the collaborative behavior benchmark rules, collaborative disturbance sensitivity threshold, and collaborative strength identifier, construct the basic framework of the data behavior collaborative model, and set the nodes and edges of the data behavior collaborative model. The nodes include data type, collection scenario, and feature parameters, and the edges include collaborative triggering conditions, response delay parameters, and amplitude linkage ratio.

[0059] The basic framework of the model is constructed using data types, acquisition scenarios, and feature parameters as nodes. For example, "Electrical signal data" is a data type node, which contains sub-nodes for different acquisition scenarios, such as the "Low load-normal temperature-hydraulic oil" scenario node. Each scenario node is connected to a corresponding feature parameter node, such as the time response feature node and amplitude correlation feature node for coil current. Edges are used to connect different nodes, representing cooperative relationships. The attributes of the edges include cooperative triggering conditions (such as the coil current reaching a certain threshold), response delay parameters (such as the allowable delay range from the change in coil current to the change in valve core displacement), and amplitude linkage ratios (such as the ratio standard between the change in coil current and the change in valve core displacement).

[0060] Step S1210: Integrate cross-class collaborative behavior and same-class collaborative behavior into the basic framework, record the details of the collaborative triggering conditions corresponding to each edge, the dynamic adjustment range of the response delay parameter, and the fluctuation threshold of the amplitude linkage ratio, and improve the collaborative logic of the data behavior collaborative model.

[0061] Within the basic framework, cross-class cooperative behaviors (such as cooperation between electrical signal data nodes and action response data nodes) and similar cooperative behaviors (such as cooperation between nodes with different characteristic parameters under an action response data node) are integrated through edge connections. For the details of the cooperative triggering conditions corresponding to each edge, such as the triggering condition details for the "electrical signal data - action response data" edge, the control signal pulse width is greater than 10 milliseconds and the coil voltage is between 90% and 110% of the rated voltage. The dynamic adjustment range of the response delay parameter is set according to different operating conditions; for example, under high-temperature conditions, the allowable range of response delay is expanded by a certain proportion compared to normal-temperature conditions. The fluctuation threshold of the amplitude linkage ratio is set to ±10% of the normal ratio standard; when the actual linkage ratio is within this threshold range, the cooperative behavior is considered normal.

[0062] Step S1211: Input the historical normal operation sequence of multidimensional test data into the model for collaborative verification, simulate the collaborative behavior under different working conditions, and adjust the collaborative parameters of the data behavior collaborative model according to the verification results, including the weight allocation of the collaborative strength identifier and the calibration of the collaborative disturbance sensitivity threshold, to form a data behavior collaborative model that reproduces the normal collaborative mode.

[0063] A large amount of historical normal operation sequences from multidimensional test data were selected and input into the constructed basic framework model. The cooperative behavior of each node and edge in the model was simulated under different operating conditions (such as high load-high temperature-high viscosity medium), and the consistency between the cooperative behavior output by the model and the actual cooperative behavior in the historical normal data was observed. If discrepancies exist, the cooperative parameters of the model were adjusted based on the verification results. For example, when the response delay of a certain cooperative behavior in the model deviates significantly from the actual data, the allowable range of the response delay for that cooperative behavior was calibrated; if the weight allocation of the cooperative strength identifier caused the model to inaccurately judge the importance of certain cooperative behaviors, the weight values ​​were adjusted. After multiple adjustments and verifications, a data behavior cooperative model that can accurately reproduce normal cooperative patterns was finally formed.

[0064] Step S130: Mine abnormal collaboration chains that deviate from the normal collaboration mode from the data behavior collaboration model. The abnormal collaboration chain includes the collaboration abnormal initiation link, intermediate transmission link, terminal impact link and abnormal collaboration performance between each link.

[0065] Step S131: Extract the normal collaboration mode parameters from the data behavior collaboration model, record the trigger condition threshold, response delay standard range, amplitude linkage standard ratio and collaboration disturbance sensitivity threshold of various collaborative behaviors, and form a normal collaboration parameter set.

[0066] From the constructed data behavior collaboration model, normal collaboration mode parameters for all collaborative behaviors are extracted. For example, for the collaborative behavior of "medium interaction data - structural vibration data", the trigger condition threshold is set to a change in medium pressure exceeding 0.5 MPa, the standard response delay range is 0.1 to 0.3 seconds, the standard amplitude linkage ratio is the ratio of pressure change to vibration acceleration change of 2:1, and the collaborative disturbance sensitivity threshold is considered to indicate a potential anomaly when the amplitude linkage ratio deviates from the standard ratio by more than 20%. These parameters for all collaborative behaviors are then summarized to form a set of normal collaboration parameters.

[0067] Step S132: Input the real-time collected multi-dimensional test data into the data behavior collaboration model to simulate the real-time collaborative behavior process and generate a real-time collaborative behavior parameter sequence that includes the trigger condition satisfaction status, the actual value of the response delay, and the actual proportion of amplitude linkage.

[0068] The currently acquired multidimensional test data is input into the data behavior collaboration model according to data type and acquisition scenario. The model simulates the real-time collaborative behavior process based on the internal node and edge connections and collaboration logic. For example, after the coil current sequence from real-time electrical signal data is input into the model, the model determines whether it meets the triggering conditions for collaboration with the action response data, calculates the actual response delay time (i.e., the actual response delay value), and the actual ratio of the coil current amplitude change to the valve core displacement amplitude change (i.e., the actual amplitude linkage ratio). This information is then arranged in chronological order to generate a real-time collaborative behavior parameter sequence.

[0069] Step S133: Compare the real-time collaborative behavior parameter sequence with the normal collaborative mode parameters to identify abnormal collaborative behaviors in real-time collaborative behavior, such as failure to meet trigger conditions, response delay exceeding the standard range, amplitude linkage ratio deviating from the standard, or exceeding the collaborative disturbance sensitivity threshold.

[0070] Each parameter in the generated real-time cooperative behavior parameter sequence is compared one by one with the corresponding parameter in the normal cooperative parameter set. For example, if the triggering condition of a certain cooperative behavior in the real-time parameter sequence is not met (such as the control signal pulse width being less than the threshold set by the model), or the actual value of the response delay exceeds the standard range of the response delay, or the actual ratio of amplitude linkage deviates from the standard ratio of amplitude linkage and exceeds the cooperative disturbance sensitivity threshold, then the cooperative behavior is identified as abnormal cooperative behavior.

[0071] Step S134: Use a cross-level collaborative anomaly identification method to distinguish between surface collaborative anomalies and deep collaborative anomalies. Surface collaborative anomalies refer to anomalies that are directly manifested as parameter deviations, while deep collaborative anomalies refer to anomalies where the parameters are not deviated but the collaborative logic is broken.

[0072] For identified abnormal collaborative behaviors, further distinctions are made between surface-level and deep-level collaborative anomalies. Surface-level collaborative anomalies are relatively intuitive; for example, the actual ratio of amplitude linkage deviates significantly from the standard ratio, which can be directly detected through parameter comparison. Deep-level collaborative anomalies, on the other hand, involve parameters that may be within the normal range, but the collaborative logic is disrupted. For instance, the order of change in action response data and media interaction data is reversed. Normally, the valve core should actuate first, followed by a change in media flow rate. However, in real-time data, the media flow rate changes first, followed by the valve core actuation. In this case, the parameters may still be within the normal range, but the collaborative logic is abnormal. This situation is what constitutes a deep-level collaborative anomaly.

[0073] Step S135: Mark the test data category, data sequence and collection scenario corresponding to the first occurrence of abnormal collaborative behavior, determine the link where the abnormal collaborative behavior is located as the starting link of the collaborative anomaly, and record the time of occurrence of the anomaly, the source of the trigger signal and the initial abnormal behavior of the starting link.

[0074] Among the identified abnormal cooperative behaviors, the first abnormal cooperative behavior is determined based on the time of occurrence. The test data category (e.g., electrical signal data), specific data sequence (e.g., coil current sequence), and acquisition scenario (e.g., medium load-normal temperature condition) corresponding to this abnormal cooperative behavior are labeled. The link consisting of the model node and edge where the abnormal cooperative behavior occurs is determined as the starting link of the cooperative anomaly. The time of occurrence of the anomaly in the starting link (accurate to milliseconds), the source of the trigger signal (e.g., control signal issued by an external control unit), and the initial abnormal manifestation (e.g., slow rise of coil current) are recorded.

[0075] Step S136: Track the impact of abnormal collaborative behavior at the initiation stage of collaborative anomaly on subsequent related collaborative behaviors. Based on the collaborative strength identifier and transmission logic in the data behavior collaborative model, monitor the process by which the abnormal collaborative behavior causes other collaborative behaviors to deviate from the normal pattern.

[0076] Step S1361: Extract the abnormal collaborative behavior characteristics of the initiation link of the collaborative anomaly, record the triggering method, response deviation, amplitude linkage anomaly performance and working conditions when the anomaly occurs. The triggering method is divided into active triggering and passive triggering, and the response deviation is divided into delay deviation and amplitude deviation.

[0077] For the initial stage of the coordinated anomaly, its abnormal coordinated behavior characteristics are extracted in detail. If the triggering method is an anomaly caused by changes in the data of the initial stage itself, it is considered active triggering; if the anomaly is caused by the influence of other stages, it is considered passive triggering. In response deviation, delay deviation refers to the difference between the actual response delay value and the standard range, and amplitude deviation refers to the difference between the actual amplitude linkage ratio and the standard ratio. Anomalies in amplitude linkage manifest as an abnormal linkage ratio between current amplitude and displacement amplitude. The operating conditions at the time of the anomaly include the operating load, ambient temperature and humidity, and medium parameters.

[0078] Step S1362: Based on the data behavior collaboration model, query all related collaboration links that have a direct collaboration relationship with the starting link of the collaboration anomaly, sort them according to the collaboration strength identifier, give priority to the related links corresponding to the collaboration strength identifier, and identify potential collaboration objects that may be affected.

[0079] In the data behavior collaboration model, based on the connection relationships between nodes and edges, all associated collaboration links directly connected to the initiating link of the collaboration anomaly are queried. For example, if the initiating link is a coil current sequence in electrical signal data, its directly associated collaboration links may include a valve core displacement sequence in action response data. These associated links are sorted according to their collaboration strength indicators, with those having higher collaboration strength indicators (such as level one or level two) given priority. These links are then identified as potential collaboration objects that may be affected.

[0080] Step S1363: Obtain the real-time test data sequence corresponding to the potential collaborative object, and analyze the changes in the trigger signal of the potential collaborative object after the abnormal collaborative behavior occurs in the initiation stage of the collaborative anomaly, including the amplitude change, frequency change and phase change of the trigger signal.

[0081] For a identified potential collaborating object, obtain its corresponding real-time test data sequence. Analyze whether the trigger signal of the potential collaborating object changes after the abnormal collaborating behavior occurs at the initiation stage of the collaborating anomaly. Taking the valve core displacement sequence of the action response data as an example, its trigger signal comes from the coil current in the electrical signal data. Analyze whether the amplitude of the valve core displacement trigger signal increases or decreases, whether the frequency changes (such as the frequency change of the pulse signal), and whether the phase shifts from the normal situation after the coil current becomes abnormal.

[0082] Step S1364: Monitor the response process of potential cooperating objects after the trigger signal changes, compare the difference between their response delay and amplitude changes and the normal cooperating mode, calculate the delay deviation ratio and amplitude deviation ratio, and quantify the abnormal situation.

[0083] After the trigger signal of a potential collaborating object changes, its response process is continuously monitored. The actual response delay time is compared with the standard range of response delay in normal collaborating mode, and the delay deviation ratio is calculated as (actual delay time - standard delay time) / standard delay time. Similarly, the difference between the actual value and the standard value of the amplitude change is calculated and then divided by the standard value to obtain the amplitude deviation ratio. These two ratios are used to quantify the abnormal situation of the potential collaborating object.

[0084] Step S1365: Record the collaborative links in the potential collaborative objects that have abnormal response delays or abnormal amplitude changes, mark them as initially affected collaborative links, and record the time difference of the abnormality and the change in operating conditions. The time difference refers to the time interval between the abnormality and the initial link.

[0085] During monitoring, when the delay deviation ratio or amplitude deviation ratio of a potential collaborating link exceeds the set abnormal threshold, the collaborating link is marked as a preliminarily affected collaborating link. The time difference between the time of the abnormality of this link and the time of the abnormality of the link that initiated the collaborating abnormality is recorded, as well as whether the operating conditions of the solenoid valve have changed during this period (such as load level adjustment, medium temperature change, etc.).

[0086] Step S1366: Analyze the abnormal collaborative behavior characteristics of the initially affected collaborative links, compare them with the abnormal characteristics of the collaborative abnormal initiation link, and determine whether there is a causal relationship between them and the abnormal collaborative behavior of the collaborative abnormal initiation link. The relationship judgment is based on the transmission logic and collaborative strength in the collaborative model.

[0087] Extract the abnormal collaborative behavior characteristics of the initially affected collaborative links, such as the type of abnormal response delay deviation and abnormal amplitude linkage, and compare them with the abnormal characteristics of the initiating link of the collaborative anomaly. Combine the transmission logic in the data behavior collaboration model (such as whether the current anomaly leads to the transmission path of the displacement anomaly) and the collaboration strength indicator (high collaboration strength level indicates a high probability of causal relationship) to determine whether the anomaly of the initially affected collaborative links is caused by the abnormal collaborative behavior of the initiating link, that is, whether there is a causal relationship.

[0088] Step S1367: Based on the preset collaborative logic rules in the data behavior collaborative model, determine whether the causal relationship conforms to the transmission path and triggering conditions defined by the model, and mark abnormal collaborative behaviors that do not conform to the preset rules as isolated anomalies, and not include them in the transmission sequence of the current abnormal collaborative chain.

[0089] The data behavior coordination model predefines the transmission paths and triggering conditions between various coordinated behaviors. For example, it stipulates that electrical signal data anomalies can only affect motion response data through specific paths, and cannot directly affect structural vibration data. The model determines whether the causal relationship between the initially affected coordinated link and the initial link conforms to these predetermined rules. If not, for example, if the initial link is an electrical signal data anomaly and the initially affected link is structural vibration data, and there is no direct transmission path or triggering condition between the two in the model, then the anomaly in the structural vibration data is marked as an isolated anomaly and not included in the current anomaly coordination chain's transmission sequence.

[0090] Step S1368: Track the impact of abnormal collaborative behavior of the initially affected collaborative link on its own subsequent collaborative links, repeat the steps of potential collaborative object query, trigger signal analysis, response process monitoring and anomaly marking to form a multi-level affected chain.

[0091] For a preliminarily affected synergistic link that is determined to have a causal relationship, it is regarded as a new "starting link". Repeat the above steps S1362 to S1367, that is, query its associated subsequent synergistic links, analyze the changes in trigger signals, monitor the response process, mark abnormal links, determine causal relationships, etc., thereby forming a multi-level affected chain, such as starting link → preliminarily affected link → secondary affected link → ...

[0092] Step S1369: Record the abnormal behavior of each affected collaborative link, the association method with the abnormal collaborative behavior of the previous level, and the time interval of the impact transmission. At the same time, record the operating parameters corresponding to the abnormal collaborative behavior of this level. Abnormal behavior includes delay deviation, amplitude deviation, and collaborative logic destruction. The association method is divided into direct triggering and indirect triggering.

[0093] During the formation of a multi-level affected chain, detailed records are kept of the specific abnormal manifestations of each affected collaborative link, such as the specific numerical range of delay deviation, the direction of amplitude deviation (too large or too small), and whether collaborative logic is disrupted. The correlation between this link and the abnormal collaborative behavior of the previous level is recorded: direct triggering means the previous level's abnormality directly causes this link's abnormality, while indirect triggering means the previous level's abnormality indirectly causes this link's abnormality by affecting other intermediate links. The time interval for the impact transmission is recorded, i.e., the time difference between the occurrence of the previous level's abnormality and the occurrence of this level's abnormality. Simultaneously, the operating parameters at the time of the occurrence of this level's abnormality are recorded, such as the ambient temperature and medium viscosity.

[0094] Step S13610: Arrange all affected collaborative links in the order of impact transmission, mark the degree of abnormality, correlation status and working condition correlation information of each link, form the transmission sequence of abnormal collaborative behavior, and reflect the spread process of abnormality from the starting link and the changing pattern under different working conditions.

[0095] All affected collaborative links recorded above are arranged in chronological order of impact transmission to form a transmission sequence of abnormal collaborative behavior. In the sequence, each link is labeled with its degree of abnormality (e.g., mild, moderate, severe, categorized according to the proportion of deviation), its relationship with preceding and following links (direct or indirect), and operating condition information (e.g., under what operating conditions the degree of abnormality is aggravated or mitigated). This transmission sequence reflects how the abnormality spreads from the initial link to different links, and the changing patterns of the abnormal behavior under different operating conditions.

[0096] Step S137: Record the test data category, data sequence, collection scenario and abnormal behavior corresponding to the affected subsequent collaborative behavior, including the triggering method of the abnormality, the degree of response deviation and the proportion of amplitude linkage abnormality, and determine the link where the affected collaborative behavior is located as the intermediate transmission link.

[0097] Based on the above transmission sequence, record relevant information about subsequent affected coordinated behaviors, excluding the initial stage, including the corresponding test data category (e.g., media interaction data), data sequence (e.g., outlet pressure sequence), acquisition scenario (e.g., high load-high temperature condition), and abnormal behavior. The triggering method for anomalies is passive triggering (affected by the preceding stage), and the degree of response deviation is such as a delay deviation ratio of 20% or an amplitude linkage anomaly ratio of 15%. The stage where the aforementioned affected coordinated behaviors occur is identified as the intermediate transmission stage.

[0098] Step S138: Continuously track the spread of abnormal collaborative behavior in intermediate transmission links, analyze the changes in the transmission rate of abnormality under different operating conditions, until the scope of influence of abnormal collaborative behavior no longer expands, and determine the link where the final affected collaborative behavior is located as the terminal impact link.

[0099] The abnormal coordinated behavior of intermediate transmission links will continue to be tracked to observe whether it will further affect more coordinated links. The rate changes of abnormal transmission under different operating conditions will be analyzed; for example, the abnormal transmission rate is faster in low-viscosity media and slower in high-viscosity media. When the impact range of the abnormal coordinated behavior no longer expands, i.e., no new abnormalities appear in subsequent coordinated links, the link where the finally affected coordinated behavior occurs will be identified as the terminal impact link.

[0100] Step S139: Analyze the correlation between the initiation stage, intermediate transmission stage and terminal impact stage of the collaborative anomaly, and record the transmission path of the abnormal collaborative behavior from the initiation stage to the terminal stage, including direct transmission path and indirect transmission path.

[0101] Identify the connections between the initiating, intermediate, and terminal stages of a coordinated anomaly. A direct transmission path refers to the anomaly being transmitted directly from the initiating stage to an intermediate or terminal stage, such as an initiating stage (an abnormal electrical signal) directly leading to a terminal stage (abnormal structural vibration). An indirect transmission path refers to the anomaly being transmitted through multiple intermediate stages, such as initiating stage → intermediate stage one → intermediate stage two → terminal stage. Record these transmission paths in detail, clarifying the sequential influence relationships between each stage.

[0102] Step S1310: Record the abnormal collaborative behavior, triggering relationship, degree of influence and working condition association of each link in each transmission path to form a detailed description of abnormal collaborative behavior transmission. The detailed description of abnormal collaborative behavior transmission includes the abnormal parameter changes of each link, the association with the preceding link and the proportion of influence on the following link.

[0103] For each conduction path, record in detail the abnormal collaborative behavior of each link in the path, such as the specific range of changes in abnormal parameters (e.g., coil current changes from the normal 5A to 3A), the correlation with preceding links (e.g., abnormal pressure in preceding links leading to abnormal flow in this link), and the proportion of impact on subsequent links (e.g., the degree to which an abnormality in this link causes an abnormality in subsequent links accounts for 80% of the total abnormality in subsequent links). Simultaneously, record the correlation of operating conditions when each link's abnormality occurs, such as under what operating conditions the proportion of impact changes.

[0104] Step S1311: Integrate the information on the initiation, intermediate transmission, and terminal impact of collaborative anomalies, as well as the transmission paths, anomaly manifestations, and operating condition correlations between each link, to construct an anomaly collaboration chain.

[0105] The aforementioned identified initiating links, all intermediate transmission links, and terminal impact links of the collaborative anomaly, as well as the transmission paths, anomaly manifestations, and operational condition correlation information between each link, are integrated and organized according to the transmission sequence and correlation relationships to construct an anomaly collaboration chain. This anomaly collaboration chain demonstrates the entire process of an anomaly from its generation and transmission to its final impact, along with related information.

[0106] Step S140: Inversely deduce the core cause of the failure through the abnormal cooperative chain. The core cause of the failure includes the cause type, triggering conditions, cooperative interference mode and initial impact link.

[0107] Step S141: Analyze the terminal impact link of the abnormal coordination chain, extract the abnormal coordination performance characteristics of the terminal impact link, the test data categories involved, the details of data changes, and the combination of working conditions when the abnormality occurs. The abnormal coordination performance characteristics include the delay deviation value, the amplitude deviation ratio, and the form of coordination logic destruction.

[0108] A detailed analysis of the terminal impact links of the abnormal coordination chain is provided. For example, the test data category involved in the terminal impact link is structural vibration data, specifically the Z-axis vibration acceleration sequence. The abnormal coordination characteristics include the numerical delay deviation (vibration response delay increases by 0.2 seconds compared to normal), the amplitude deviation ratio (vibration acceleration amplitude increases by 25% compared to normal), and the form of coordination logic failure (the order of vibration and medium pressure changes is reversed). Data change details include the peak occurrence time and fluctuation frequency of the vibration acceleration sequence. The operating conditions at the time of the anomaly were high load, ambient temperature of 35℃, and medium viscosity of 200 cSt.

[0109] Step S142: Based on the transmission path of the abnormal collaboration chain, trace back from the terminal impact link to the next higher intermediate transmission link, analyze the specific manifestations of the abnormal collaboration behavior of the intermediate transmission link, the source of the trigger signal and the way it affects the terminal impact link, and record the way the abnormal is transmitted.

[0110] Based on the transmission path of the abnormal coordination chain, starting from the terminal impact link, we trace back to the intermediate transmission link preceding it. We analyze the specific manifestations of the abnormal coordination behavior in this intermediate transmission link, such as abnormal outlet pressure in the media interaction data, with pressure amplitude fluctuations exceeding the normal range. The trigger signal originates from abnormal valve core displacement in the preceding action response data. The effect on the terminal impact link is that the abnormal outlet pressure fluctuations lead to increased shell vibration, and the abnormality is transmitted directly through mechanical transmission.

[0111] Step S143: Extract the abnormal collaborative features of the intermediate transmission link of the previous level, compare them with the abnormal features of the terminal influence link, calculate the feature similarity between the two, and record the collaborative interference relationship between the two. The collaborative interference relationship is divided into direct interference and indirect interference.

[0112] Extract the abnormal collaborative features of the intermediate transmission link (abnormal outlet pressure of the medium), such as pressure fluctuation frequency and peak pressure value. Compare these with the abnormal features (vibration frequency, peak acceleration) of the terminal influencing link (abnormal structural vibration), and calculate the feature similarity, for example, by comparing whether the frequency components of the two are consistent and whether the peak occurrence time corresponds. If the fluctuation of the outlet pressure directly causes the change in vibration, the collaborative interference relationship between the two is direct interference; if it is caused indirectly by other factors, it is indirect interference.

[0113] Step S144: Continue to trace back the intermediate transmission links of the previous level, repeat the steps of abnormal collaborative feature extraction, feature comparison and interference relationship analysis, and trace upward step by step until the starting link of collaborative anomaly is traced back to form a complete reverse tracing chain.

[0114] Following the above method, continue tracing back from the current intermediate transmission link to the previous intermediate transmission link, extract its abnormal collaborative characteristics, compare them with the abnormal characteristics of the current link, analyze the interference relationship, and so on upwards until tracing back to the link where the collaborative anomaly originated. Connect the links and relationships in the above tracing process to form a complete reverse tracing chain.

[0115] Step S145: Integrate all abnormal collaborative features, interference relationships and operating condition information extracted during the reverse tracing process, construct a reverse transmission logic chain, and record the origin of abnormal collaborative behavior, transmission logic, changes in different links and the impact of operating conditions on the transmission process.

[0116] The abnormal collaborative characteristics of each link, the interference relationship between links, and the operating condition information when the abnormality occurs in each link are integrated during the reverse tracing process to construct a reverse transmission logic chain. This reverse transmission logic chain records the origin of abnormal collaborative behavior (the abnormality in the initiating link of the collaborative abnormality), the transmission logic between each link (such as the physical mechanism through which it is transmitted), the specific changes of the abnormality in different links (such as the increase or decrease of the degree of abnormality), and the impact of operating condition information on the transmission process (such as high temperature conditions accelerating the transmission of abnormality).

[0117] Step S146: Locate the exception triggering link in the reverse propagation logic chain. This exception triggering link is the initial source of all subsequent abnormal collaborative behaviors. Its exception directly triggers a chain reaction. This link is identified as the initial impact link.

[0118] In the reverse propagation logic chain, the first abnormal link is identified, i.e., the abnormal triggering link. This abnormal triggering link is the initial source of all subsequent abnormal coordinated behaviors; without the influence of preceding abnormal links, its own abnormality directly triggers the subsequent chain reaction. This abnormal triggering link is identified as the initial influencing link, such as an abnormal coil current link in electrical signal data.

[0119] Step S147: Extract the test data sequence, abnormal coordination characteristics, acquisition scenario and trigger signal details of the initial impact link, and analyze the specific manifestations of the disruption of the normal coordination mode of this link, including the satisfaction of coordination trigger conditions, deviation of response delay and abnormal changes in amplitude linkage ratio.

[0120] Extract the initial impact test data sequence (e.g., coil current change over time sequence), abnormal coordination characteristics (e.g., slow current rise, low peak current), acquisition scenario (medium load - normal temperature condition), and trigger signal details (control signal pulse width, voltage value). Analyze the coordination triggering conditions of this component under normal coordination mode (e.g., control signal pulse width ≥ 10ms), and whether the actual trigger signal meets these conditions; normal response delay (e.g., the time for current to rise from the initial value to the stable value is 0.1 seconds), and whether the actual response delay deviates from the expected value; normal amplitude linkage ratio (e.g., the ratio of peak current to peak valve core displacement is 5:1), and whether the actual ratio changes abnormally.

[0121] Step S148: Using an active collaborative disturbance simulation method, based on a data behavior collaborative model, simulate the disturbance effect of different potential inducements on the initial impact link, and monitor the fit between simulated anomalies and actual anomalies.

[0122] Based on a data-driven behavioral coordination model, different potential triggers are set, such as inter-turn short circuits, control signal interference, and core jamming. The model simulates the disturbance effects of these potential triggers on the initial influencing element (coil current). For example, simulating an inter-turn short circuit reduces coil resistance, leading to an increase in current; simulating control signal interference introduces noise into the control signal, causing current fluctuations. The degree of fit between the simulated abnormal coordination characteristics (such as current change curves and peak values) and the actual abnormal coordination characteristics of the initial influencing element is monitored and calculated.

[0123] Step S149: Associate with a preset list of solenoid valve fault causes. The list of fault causes includes various types of causes that may lead to synergistic anomalies, typical triggering conditions, common interference methods, and corresponding synergistic anomaly manifestations. The types of causes are divided into mechanical, electrical, media, and environmental categories.

[0124] Step S1491: Collect all fault cause information from historical fault cases of the solenoid valve. The fault cause information covers different types such as mechanical structure wear, electrical component aging, changes in medium composition, and abnormal ambient temperature and humidity.

[0125] We extensively collected historical failure cases of this type of solenoid valve, obtaining information on the causes of failure from equipment maintenance records, fault reports, technical documents, and other sources. Mechanical structural wear includes valve core wear, aging and damage to sealing rings, and weakened spring elasticity; electrical component aging includes coil aging, contact oxidation, and sensor failure; changes in media composition include increased impurity content in the media, oil emulsification, and viscosity changes; and abnormal environmental temperature and humidity include excessively high or low ambient temperatures and excessive humidity leading to component corrosion.

[0126] Step S1492: Analyze the complete record of each historical fault case, extract the operating conditions, running time, previous operation records and abnormal signs before the fault occurred, and form a typical trigger condition description. The operating conditions include operating load, medium parameters, environmental parameters and electrical parameters.

[0127] Detailed analysis was conducted on the records of each historical fault case. For example, in a coil aging fault case, the operating conditions at the time of the fault were: the operating load was consistently above 90% of the rated load; the medium parameters were: medium temperature 50℃, viscosity 150cSt; the environmental parameters were: ambient temperature 30℃, humidity 60%; and the electrical parameters were: power supply voltage fluctuation within ±5%. The operating duration was 2000 hours. Previous operation records included several recent frequent start-stop operations. The abnormal signs before the fault occurred were increased coil heating and a gradually slowing response. Integrating the above information formed a typical description of the triggering conditions for coil aging.

[0128] Step S1493: Analyze how each historical fault cause interferes with the collaborative behavior of multidimensional test data. Through experimental simulation and data analysis, record the process by which the cause affects the trigger response, amplitude linkage, and time synchronization of the data. The impact of the trigger response includes extending the response delay and changing the response threshold. The impact of amplitude linkage includes disrupting the amplitude ratio and changing the phase synchronization. The impact of time synchronization includes disrupting the collaborative triggering sequence.

[0129] Taking oil emulsification caused by changes in medium composition as an example, this study simulates the interference of oil emulsification on the coordinated behavior of multidimensional test data through experiments. In the experiment, emulsified oil was introduced into a solenoid valve, and relevant test data were collected. Analysis revealed that oil emulsification increases the viscosity of the medium, thus affecting the trigger response of the action response data and prolonging the response delay of the valve core. Simultaneously, the amplitude linkage ratio between the medium flow rate and the valve core displacement is disrupted, and the phase synchronization deteriorates (the time lag between flow rate changes and displacement changes increases). The coordinated triggering sequence of action response data and medium interaction data is disrupted, and some changes that originally occurred simultaneously exhibit time differences. Through data analysis, the specific manifestations and extent of these interference methods are recorded.

[0130] Step S1494: Record the abnormal collaborative behavior caused by each historical fault, including the type of abnormal collaborative behavior, transmission path, scope of influence, degree of abnormality in each link and changes in the rate of abnormal transmission. The types of abnormal collaborative behavior are divided into surface and deep, the transmission path is divided into direct and indirect, and the scope of influence is divided into single link and multiple links.

[0131] For the cause of oil emulsification, the resulting synergistic anomalies were recorded: the types of abnormal synergistic behaviors were mainly surface anomalies (such as deviations in flow rate and pressure amplitude); the transmission path was media interaction data → action response data → structural vibration data (indirect transmission); the scope of influence was multi-stage (involving three types of test data); the degree of anomaly in each stage was as follows: high degree of anomaly in media interaction data (flow rate deviation 30%), medium degree of anomaly in action response data (displacement deviation 15%), and low degree of anomaly in structural vibration data (vibration deviation 10%); the rate of anomaly transmission was slower in high-viscosity media.

[0132] Step S1495: Classify and organize the collected fault cause information, establish a classification directory according to the cause type, and further subdivide each directory into subdirectories according to the similarity of the triggering conditions. Each subdirectory contains the corresponding typical triggering conditions, interference methods and collaborative abnormal manifestations.

[0133] The collected fault cause information is categorized into mechanical, electrical, media, and environmental categories. Within the electrical category, it is further subdivided based on the similarity of triggering conditions, including subcategories for control signal anomalies, coil faults, and sensor faults. Each subcategory contains specific fault causes; for example, the coil fault subcategory includes coil turn-to-turn short circuits, coil open circuits, and coil aging. Each cause corresponds to and records typical triggering conditions, interference patterns, and associated abnormal behaviors.

[0134] Step S1496: Supplement the relevant information in publicly available research data, technical standards and manufacturer maintenance manuals on the causes of solenoid valve failures in the industry, and improve the details of the triggering conditions, the description of the interference methods and the quantitative indicators of abnormal performance for various causes. The details of the triggering conditions include the specific parameter range.

[0135] Consult publicly available research papers on the causes of solenoid valve failures, relevant technical standards (such as ISO standards and national industry standards), and maintenance manuals provided by solenoid valve manufacturers to obtain more information. For example, supplement the maintenance manual with details of the triggering conditions for coil aging, such as operating time exceeding 3000 hours or ambient temperature consistently above 40°C; improve the description of interference methods, such as coil aging leading to increased resistance and decreased current; and supplement quantitative indicators of abnormal performance, such as a 10% to 20% reduction in peak current.

[0136] Step S1497: Standardize the processed fault cause information and use a unified expression method to describe the cause type, triggering condition, interference method and abnormal manifestation. The triggering condition should specify the parameter type and manifestation form, the interference method should specify the target and method of interference, and the abnormal manifestation should specify the characteristic parameters and the range of change.

[0137] All fault cause information is standardized and presented in a unified manner. For example, the cause type is standardized as "electrical - coil aging"; the triggering conditions specify the parameter type (running time, ambient temperature) and manifestation (running time > 3000 hours, ambient temperature > 40℃ duration > 500 hours); the interference method specifies the affected object (coil resistance) and the method (increased resistance); the abnormal manifestation specifies the characteristic parameters (peak coil current, response delay) and the range of variation (peak current decreases by 10%-20%, response delay increases by 0.1-0.2 seconds).

[0138] Step S1498: Assign a unique identifier to each fault cause and establish an index linking the cause information with historical fault cases. The index includes information such as case number, fault occurrence time, and processing result.

[0139] Assign a unique identifier to each fault cause, such as "Electrical-Coil-001" representing an electrical coil aging fault. Establish an index linking the cause information with historical fault cases. Through this index, you can query information such as the historical fault case number corresponding to the cause, the specific time the fault occurred, the handling measures at that time, and the handling results (such as the fault being resolved after replacing the coil).

[0140] Step S1499: Construct a structured list of fault causes. The list includes fields such as cause identifier, cause type, subcategory, typical triggering conditions, cooperative interference methods, typical abnormal manifestations, related case index, and update time. Typical triggering conditions include operating parameters, and cooperative interference methods include the mode of action.

[0141] The standardized information is organized in a structured format to construct a fault cause list. The list includes the following fields: cause identifier (e.g., "electrical-coil-001"), cause type (electrical), subcategory (coil fault), typical triggering conditions (including operating parameters such as running time and ambient temperature), cooperative interference method (including the affected object and the method of action), typical abnormal manifestations (characteristic parameters and range of variation), related case index (case number list), and update time (the time when the cause information was last revised).

[0142] Step S14910: Establish a dynamic update method for the list, regularly collect new failure cases and industry research results, add new failure cause information to the list, optimize the description of existing causes, adjust the classification method to adapt to new failure types, and optimize the description of existing causes by refining the triggering conditions and supplementing interference methods.

[0143] Establish a regular update mechanism, such as updating the list of failure triggers quarterly. Collect new failure cases and new research findings published in the industry during the quarter. If new types of failure triggers are discovered, add them to the list and create corresponding classification directories. For existing triggers, optimize the descriptions based on new cases and research findings, such as refining trigger conditions (adding more influencing factors) and adding newly discovered interference methods. If new failure types emerge, adjust the classification method of the list to ensure accurate coverage of all failure triggers.

[0144] Step S1410: Compare the abnormal collaborative characteristics, triggering scenarios, operating conditions and simulated disturbance results of the initial influencing link with the items in the fault cause list, calculate the degree of fit, and select the candidate cause with the highest degree of fit.

[0145] The abnormal collaborative characteristics of the initial influencing factors (such as slow rise in coil current and low peak value), triggering scenario (medium load - normal temperature), operating condition information (running time of 1800 hours), and simulation results of active collaborative disturbances (such as the simulation of abnormal current characteristics during inter-turn short circuits being similar to the actual situation) are compared with the various factors in the fault cause list. For each factor, scores are given based on abnormal feature matching degree, triggering scenario similarity, operating condition information conformity, and simulation result fit, and a weighted fit value is calculated. The factor with the highest fit value is selected as a candidate factor, such as "inter-turn short circuit in coil".

[0146] Step S1411: Analyze the correlation between the candidate cause and all abnormal collaborative behaviors in the reverse propagation logic chain, verify that the candidate cause can explain all abnormal collaborative behaviors, including the abnormal differences, propagation rate and impact range of each link, and match with the operating condition information, determine that the candidate cause is the core cause of the fault, record its cause type, triggering conditions, collaborative interference mode and initial impact link, and form the final cause tracing result. The collaborative interference mode refers to the specific way of destroying collaborative behavior.

[0147] For example, step S1411-1: Extract typical collaborative anomaly manifestations, collaborative interference methods and applicable operating conditions corresponding to candidate causes. Typical collaborative anomaly manifestations include delay deviation range, amplitude linkage ratio anomaly, and collaborative logic destruction form. Collaborative interference methods include acting on specific collaborative links and destruction triggering methods.

[0148] Extract typical coordinated anomalies corresponding to candidate causes (inter-turn short circuits in the coil) from the fault cause list, such as delay deviation range (current response delay increases by 0.1-0.3 seconds), abnormal amplitude linkage ratio (current to displacement linkage ratio becomes 3:1), and coordinated logic disruption (valve core only starts to move after current stabilizes). The coordinated interference mode acts on the electrical signal data link, disrupting the triggering method (by changing the coil resistance, causing abnormal current changes, thus affecting the effective transmission of the trigger signal). Applicable operating conditions include medium to high loads and ambient temperatures above 25℃.

[0149] Step S1411-2: Compare the typical collaborative anomalies of the candidate causes with the actual performance of each abnormal collaborative link in the abnormal collaborative chain, calculate the performance fit of each link, and record the fit points and differences. Fit points refer to the same anomaly type, and differences refer to different anomaly degrees.

[0150] The typical coordinated anomalies of candidate triggers are compared one by one with the actual performance of each link in the anomaly coordination chain. For example, the electrical signal link (initial influence link) is actually characterized by a slow current rise and a low peak value, which is consistent with the typical performance of the candidate trigger, showing a high degree of fit; the action response link is actually characterized by a smaller valve core displacement and increased delay, which is consistent with the linkage anomaly caused by the candidate trigger, showing a high degree of fit; the media interaction link is actually characterized by a lower flow rate, which is linked to the displacement anomaly, showing a relatively high degree of fit. The points of fit (such as the anomaly type of each link being consistent with the typical performance) and the points of difference (such as the amplitude deviation ratio of a certain link being slightly larger than the typical performance) are recorded.

[0151] Step S1411-3: Compare the cooperative interference methods of candidate inducements with the cooperative characteristics of the initial influencing link, and determine the matching degree between the interference methods and cooperative characteristics. Cooperative characteristics include cooperative strength and disturbance sensitivity.

[0152] The cooperative interference mode of the candidate cause is to act on the electrical signal data link, changing the coil resistance. The cooperative characteristics of the initial influencing link are high cooperative strength (level two in terms of cooperative strength with the action response data) and high disturbance sensitivity (sensitive to current changes). To determine the matching degree between the interference mode and the cooperative characteristics, the change in coil resistance directly affects the current, and since this link has high cooperative strength and disturbance sensitivity, it is easy to trigger cooperative anomalies due to current abnormalities, indicating a high matching degree.

[0153] Step S1411-4: Based on the data behavior collaboration model, simulate the action process of candidate triggers, starting from the initial impact link, reproduce the formation and transmission process of the abnormal collaboration chain, and generate simulated transmission paths and simulated abnormal performance of each link.

[0154] Based on the data behavior collaborative model, parameters for candidate causes (inter-turn short circuits in coils) are set (e.g., a 20% reduction in coil resistance), and the simulation begins from the initial influencing element (electrical signal data). According to the collaborative logic, the model sequentially simulates how current anomalies affect action response data, how action response anomalies affect medium interaction data, and how medium interaction anomalies affect structural vibration data, generating simulated conduction paths and simulated abnormal behaviors of each element (e.g., delay deviations, amplitude deviations, etc.).

[0155] Step S1411-5: Compare the simulated transmission path, the simulated abnormal behavior of each link with the actual abnormal collaborative chain transmission path and the actual abnormal behavior of each link, and calculate the path consistency and behavior consistency.

[0156] Compare the simulated conduction path (electrical → action → medium → structure) with the actual abnormal conduction path to calculate the path consistency (100% if completely consistent). Compare the simulated abnormal behavior of each link (such as simulated displacement delay, flow deviation) with the actual abnormal behavior to calculate the behavior consistency (high consistency if the difference is within the allowable range).

[0157] Step S1411-6: Verify the interference method of the candidate cause. Through the anomaly of the initial influencing link, according to the transmission path, coordination strength identifier and operating conditions of the abnormal coordination chain, it triggers the process of all subsequent intermediate transmission links to become abnormal, and the difference in the abnormal performance is consistent with the actual situation. The difference in the abnormal performance includes the degree and rhythm.

[0158] After verifying that the interference mode of the candidate cause (change in coil resistance) leads to the initial influencing link (abnormal current), it is determined whether the abnormality is triggered in subsequent intermediate transmission links according to the transmission path of the abnormality coordination chain (electrical → action → dielectric → structure), based on the coordination strength of each link (e.g., high coordination strength between electrical and action, fast transmission) and operating conditions (medium load, 35℃). Simultaneously, the differences in the abnormal behavior of each link obtained from the simulation (e.g., the abnormality degree of the dielectric link is lower than that of the action link) are consistent with the differences in the actual abnormality coordination chain, and the abnormal rhythm (e.g., the abnormal transmission rate is initially fast and then slows down) is also consistent with reality.

[0159] Step S1411-7: Compare the typical synergistic abnormal manifestations of candidate causes with the differences in abnormal manifestations of each link in the abnormal synergistic chain. The differences in abnormal manifestations include the strength of the abnormality, the speed of change, and the size of the impact range. The matching process combines the transmission path and changes in operating conditions.

[0160] In the typical coordinated anomaly manifestations of candidate causes, the degree of anomaly gradually decreases from the electrical signal link to the structural vibration link, with a rhythm of rapid transmission in the early stage and slow transmission in the later stage. The scope of influence covers four types of data: electrical, motion, medium, and structure. The actual differences in the anomaly manifestations of each link in the anomaly coordination chain are as follows: the electrical link exhibits the strongest anomaly, followed by motion, then medium, and the structure is the weakest. The transmission rhythm is fast in the early stage and slow in the later stage, with the same scope of influence. The two match well.

[0161] Step S1411-8: Match the typical triggering conditions of candidate causes with the triggering scenarios and operating condition change trends of the abnormal coordination chain, and verify the consistency between the typical triggering conditions of the causes and the actual operating scenarios when the abnormality occurs. The actual operating scenarios include load, medium parameters, and environmental conditions.

[0162] Typical triggering conditions for candidate causes are runtime exceeding 1500 hours, medium to high load, and ambient temperature above 25℃. The triggering scenario for the abnormal coordination chain is runtime of 1800 hours, medium load, and ambient temperature of 35℃, with a recent trend of gradually increasing load. The actual operating scenario is consistent with the typical triggering conditions, and the verification is successful.

[0163] Step S1411-9: Calculate the fit value between all candidate causes and the abnormal synergy chain, and select the candidate cause with the highest fit value that can cover all abnormal synergy links.

[0164] If multiple candidate causes exist, their degree of fit with the abnormal coordination chain is calculated for each, and the candidate cause with the highest degree of fit that can explain all abnormal coordination links is selected. In this embodiment, the inter-turn short circuit of the coil has the highest degree of fit and can cover all links.

[0165] Step S1411-10: Based on the simulation results and comparative verification results, generate a correlation assessment report between candidate causes and abnormal collaborative chains. The report includes data on path consistency, performance consistency, and trigger condition matching degree.

[0166] Based on data such as the consistency between the simulated transmission path and the actual path (90%), the consistency between the simulated performance and the actual performance (85%), and the matching degree of the triggering conditions (95%), a correlation assessment report is generated to evaluate the degree of correlation between the candidate triggers and the abnormal collaborative chain.

[0167] Step S1411-11: If the performance of some components does not meet the expected threshold, the interference effect of the candidate cause under the corresponding working condition is re-simulated to verify whether it can explain the abnormality of the component. The causes include changes in working conditions and other minor interferences.

[0168] If the fit of a certain step is lower than the expected threshold (e.g., 80%), the interference effect of the candidate cause under the corresponding operating conditions of that step is re-simulated, considering the influence of changes in operating conditions or other minor interference factors (e.g., slight media contamination), to verify whether it can explain the anomaly of that step. If the fit is satisfactory after the re-simulation, continue; otherwise, consider other candidate causes.

[0169] Step S1411-12: Based on the correlation assessment report and re-verification results, select the candidate cause with the highest fit value, the highest path consistency, and that can cover all abnormal collaborative links as the core cause of the failure. Record its cause type, specific triggering conditions, collaborative interference method, and detailed information on the initial impact link. The specific triggering conditions include the parameter range, and the collaborative interference method includes the way it disrupts collaborative behavior and the path of action, forming the final cause tracing result.

[0170] Based on the comprehensive correlation assessment report and re-verification results, the candidate cause (inter-turn short circuit in the coil) was identified as the core cause of the fault. The cause type was recorded as electrical, with specific triggering conditions including 1800 hours of operation, medium load (50%-80% of rated load), and an ambient temperature of 35°C. The cooperative interference method involved changing the coil resistance through the inter-turn short circuit, causing abnormal current changes in the electrical signal data (damage mode). This then affected the action response data through the electrical-action cooperative path, subsequently affected the medium interaction data through the action-medium cooperative path, and finally affected the structural vibration data through the medium-structure cooperative path (action path). The initial influencing element was the coil current sequence in the electrical signal data. This resulted in the final cause tracing result.

[0171] Step S150: Based on the transmission characteristics of the core cause of the fault and the abnormal coordination chain, generate an executable instruction sequence to control the maintenance system to perform corresponding fault handling operations.

[0172] Step S151: Analyze the type, triggering conditions, cooperative interference methods, and characteristics of the initial influencing links of the core causes of the fault. Combine the transmission path of the abnormal cooperative chain and the operating condition correlation information to determine the key intervention points that can prevent the causes from continuing to play a role. Record the test data links, operating conditions, and intervention timing windows corresponding to the intervention points.

[0173] The core cause of the fault is electrical (inter-turn short circuit in the coil), triggered by medium load and high temperature. The co-interference mechanism involves altering the coil resistance and affecting the current, with the initial impact occurring at the electrical signal data stage. Based on the transmission path of the abnormal co-interference chain (electrical → action → dielectric → structure) and the associated operating conditions (high temperature exacerbates the anomaly), the critical intervention point is determined to be the electrical signal data stage (coil replacement). The corresponding test data for the intervention point are coil current and voltage data. The operating condition is shutdown (to ensure safety), and the intervention window is the current production break, with an estimated 2 hours of maintenance time.

[0174] Step S152: Based on key intervention points, generate a targeted set of instructions for blocking the triggers. The set of instructions for blocking the triggers includes the specific content of the intervention, the execution order, the required tools and equipment, the operating environment requirements, and the collaborative monitoring indicators during the operation. The specific content of the intervention includes adjusting parameters, replacing components, and isolating interference sources. The execution order is sorted according to the priority of impact. The operating environment requirements include temperature, humidity, and safety protection conditions.

[0175] The specific content of the cause-blocking operation instruction set is to replace the faulty coil (replace the component). The execution sequence is as follows: 1. Disconnect the power supply to the solenoid valve; 2. Disassemble the solenoid valve coil housing; 3. Remove the old coil; 4. Install the new coil; 5. Connect the coil wiring; 6. Restore the housing. Required tools and equipment include screwdrivers, wire strippers, a new coil (model-matching), and insulating tape. The operating environment requirements are an ambient temperature of 15-30℃ and humidity ≤60%. Operators must wear insulating gloves and safety goggles (safety protection conditions). The co-monitoring indicators during the operation are coil resistance (measured after installation, should be within the rated range) and insulation resistance (≥10MΩ).

[0176] Step S153: Analyze the transmission characteristics of the abnormal cooperative chain, including the transmission path length, the degree of cooperative dependence of each link, the abnormal propagation speed, the impact of the operating conditions on the transmission, and the reversibility of the abnormality, and determine the priority order of cooperative recovery.

[0177] For example, step S1531: Extract all links in the abnormal collaboration chain, arrange them in the order of abnormal transmission, count the total number of links to determine the transmission path length, and record the test data category and collaborative behavior type corresponding to each link. The abnormal transmission order refers to the sequence from the starting link to the terminal link.

[0178] The links in the abnormal coordination chain are arranged in the following order of transmission: electrical signal link → action response link → medium interaction link → structural vibration link, with a total of 4 links, i.e., a transmission path length of 4. The test data categories corresponding to each link are electrical signal data, action response data, medium interaction data, and structural vibration data, respectively, and the coordination behavior type is cross-category coordination.

[0179] Step S1532: Analyze the collaborative dependency relationship between each link and the preceding link. Based on the collaborative strength identifier and transmission logic in the data behavior collaborative model, calculate the degree of dependency parameter of the abnormal performance of the link on the abnormality of the preceding link.

[0180] Taking the action response stage as an example, its coordination strength with the preceding electrical signal stage is identified as Level 2 (higher), and the transmission logic is current-driven valve core action. The dependency parameter is calculated by analyzing the probability of electrical signal anomalies causing action response anomalies in historical data, the proportion of action response anomalies caused by electrical signal anomalies, etc., and the dependency parameter is obtained as 0.8 (range 0-1, the higher the value, the greater the dependency).

[0181] Step S1533: Analyze the collaborative dependency relationship between each link and subsequent links. Based on the collaborative strength identifier and transmission logic in the data behavior collaborative model, calculate the impact parameter of the abnormal performance of this link on the abnormality of subsequent links.

[0182] The coordination strength between the action response stage and the subsequent media interaction stage is identified as Level 1 (High). The transmission logic is that the valve core action changes the flow area, affecting the media parameters. The influence degree parameter is calculated by analyzing the probability that abnormal action response leads to abnormal media interaction, and the proportion of abnormal media interaction caused by abnormal action response, resulting in an influence degree parameter of 0.9.

[0183] Step S1534: Combining the dependency parameter and the influence parameter, calculate the collaborative importance value of each link using the collaborative importance logical weight calculation method.

[0184] The synergy importance value is calculated as follows: Dependency parameter × 0.4 + Influence parameter × 0.6. For example, the synergy importance value for the action response stage is 0.8 × 0.4 + 0.9 × 0.6 = 0.86. Similarly, the values ​​for other stages are calculated: Electrical signal stage: Dependency parameter 0 (no preceding stages), Influence parameter 0.9 (significant impact on action response), Synergy importance value = 0 × 0.4 + 0.9 × 0.6 = 0.54; Medium interaction stage: Dependency parameter 0.85 (dependent on action response), Influence parameter 0.7 (impact on structural vibration), Synergy importance value = 0.85 × 0.4 + 0.7 × 0.6 = 0.76; Structural vibration stage: Dependency parameter 0.75 (dependent on medium interaction), Influence parameter 0 (no subsequent stages), Synergy importance value = 0.75 × 0.4 + 0 × 0.6 = 0.3.

[0185] Step S1535: Record the time interval for abnormal collaborative behavior to be transmitted from one link to the next, exclude abnormal intervals caused by sudden changes in operating conditions, calculate the effective transmission time between adjacent links, and form a transmission time series.

[0186] The time interval from the abnormality to the electrical signal link to the action response link is 0.5 seconds, the time interval from the action response to the medium interaction link is 0.3 seconds, and the time interval from the medium interaction to the structural vibration link is 0.4 seconds. All of these are effective transmission times (without the influence of sudden changes in operating conditions). The transmission time sequence is [0.5, 0.3, 0.4].

[0187] Step S1536: Calculate the average conduction time based on the conduction time series, analyze the trend of conduction time changes, determine the change law of abnormal spread speed, and analyze the reasons for speed changes in conjunction with changes in operating conditions.

[0188] The average conduction time is (0.5 + 0.3 + 0.4) / 3 = 0.4 seconds. The trend of conduction time variation is that the time from electrical response to action is relatively long, the time from action to medium interaction is short, and the time from medium to structural vibration is slightly longer. The abnormal propagation speed changes from slow to fast and then slows down again. In related operating conditions, the medium temperature is high and the viscosity is low at the point of action-to-medium interaction, which may lead to a faster conduction speed.

[0189] Step S1537: Set a conduction speed threshold and an impact number threshold, and mark conduction links with a conduction speed faster than the speed threshold and an impact number of subsequent links greater than the number threshold as high priority handling links.

[0190] The transmission speed threshold was set at 0.35 seconds (90% of the average transmission time), and the threshold for the number of affected links was set at 2. The transmission time from the action response link to the medium interaction link was 0.3 seconds, which was faster than the speed threshold, and it affected two subsequent links: medium interaction and structural vibration (more than the number threshold). Therefore, the action response link was marked as a high-priority handling link.

[0191] Step S1538: Analyze the reversibility of anomalies in each step. Based on historical data and the collaborative model, determine whether the anomaly in the step can be restored to normal by adjusting parameters and reconstructing the collaborative logic.

[0192] Abnormalities in the electrical signal link are due to coil failure, and can be restored after replacing the coil (high reversibility); abnormalities in the action response link are caused by abnormal electrical signals, and can be restored after the electrical signal is restored (high reversibility); abnormalities in the medium interaction link are caused by abnormal action response, and can be restored after the action is restored (medium reversibility); abnormalities in the structural vibration link are caused by abnormal medium interaction, and can be restored after the medium is restored (medium reversibility).

[0193] Step S1539: Standardize the collaborative importance value, abnormal spread speed impact, and reversibility of each link, convert them into dimensionless evaluation scores, and combine them with the high-priority handling link identifiers to establish a multi-dimensional collaborative recovery priority evaluation standard, and implement the weight allocation of each dimension recorded in the evaluation standard.

[0194] The importance of collaboration, the impact of anomaly spread speed (shorter transmission time, higher score), and reversibility (higher reversibility, higher score) are standardized and converted into an evaluation score of 0-1. The weights are: importance of collaboration 0.4, impact of anomaly spread speed 0.3, reversibility 0.2, and high-priority handling step identification 0.1 (add 0.1 points if it is high priority). The evaluation criterion is the sum of the scores for each dimension multiplied by their respective weights.

[0195] Step S15310: Prioritize all abnormal steps according to the evaluation criteria. The priority order corresponds to the order of handling. The evaluation index value and the basis for ranking each step are recorded in the ranking result.

[0196] Calculate the evaluation scores for each stage: Electrical signal stage: Coordination importance score 0.54, propagation speed impact score 0.6 (conduction time 0.5 seconds, after standardization), reversibility score 1.0, high priority indicator 0 points, total score = 0.54×0.4+0.6×0.3+1.0×0.2+0=0.216+0.18+0.2=0.596; Action response stage: Coordination importance score 0.86, propagation speed impact score 0.9 (conduction time 0.3 seconds), reversibility score 1.0, high priority indicator 0.1, total score = 0.86×0.4+0.9×0.3+1.0×0.2+0 =0.344+0.27+0.2+0.1=0.914; Medium interaction link: Coordination importance score 0.76, propagation speed influence score 0.7 (propagation time 0.4 seconds), reversibility score 0.7, total score = 0.76×0.4+0.7×0.3+0.7×0.2=0.304+0.21+0.14=0.654; Structural vibration link: Coordination importance score 0.3, propagation speed influence score 0.5, reversibility score 0.7, total score = 0.3×0.4+0.5×0.3+0.7×0.2=0.12+0.15+0.14=0.41. Priority ranking is: Action response link (0.914) → Medium interaction link (0.654) → Electrical signal link (0.596) → Structural vibration link (0.41). The ranking is based on the total score of each stage of the evaluation.

[0197] Step S15311: Combine the priority ranking results with the transmission path and operating condition association information of the abnormal collaborative chain to form a priority list that includes the order of handling, the importance of the link, the transmission characteristics, the reversibility status and the operating condition requirements.

[0198] The priority list includes: 1. Action response stage (handling order 1, high importance, fast transmission speed, high reversibility, operating condition requires normal temperature); 2. Medium interaction stage (handling order 2, medium importance, medium transmission speed, medium reversibility, operating condition requires normal medium viscosity); 3. Electrical signal stage (handling order 3, medium importance, slow transmission speed, high reversibility, operating condition requires power off); 4. Structural vibration stage (handling order 4, low importance, slow transmission speed, medium reversibility, operating condition requires no special conditions).

[0199] Step S154: Using a collaborative adaptive recovery method, for each link in the abnormal collaborative chain, based on its abnormality, collaborative dependency, and reversibility, generate corresponding collaborative recovery operation instructions, record the specific methods of adjusting the response parameters of the test data and reconstructing the collaborative logic. Adjusting the response parameters of the test data includes correcting the delay and adjusting the amplitude ratio, and reconstructing the collaborative logic includes repairing the trigger relationship.

[0200] For the action response stage (priority 1), the anomaly level is moderate, the degree of coordination dependence is high, and the reversibility is high. The coordination recovery operation command is to adjust the response parameters, correct the valve core action delay (by adjusting the pulse width of the control signal, increasing it from the original 8ms to 12ms), and adjust the amplitude ratio (calibrating the valve core displacement sensor to ensure that the displacement-current linkage ratio is restored to 5:1). The coordination logic is reconstructed to repair the triggering relationship between the electrical signal and the action response (ensuring that the valve core action is triggered immediately after the current reaches the threshold).

[0201] For the media interaction stage (priority 2), the anomaly is mild, the degree of coordination dependence is moderate, and the reversibility is moderate. The operation instructions are to adjust the response parameters, correct the media pressure response delay (by cleaning impurities at the valve port to reduce flow resistance), and adjust the amplitude ratio (to ensure that the flow rate and displacement linkage ratio returns to normal).

[0202] For the electrical signal link (priority 3), the anomaly is severe, the degree of coordination dependence is low, and the reversibility is high. The operation instruction is to replace the coil, adjust the response parameters (measure and record the resistance and inductance parameters of the new coil, input them into the model for calibration), and reconstruct the coordination logic (ensure that the triggering relationship between the control signal and the current is correct).

[0203] For the structural vibration component (priority 4), the anomaly is mild, the degree of synergy is moderate, and the reversibility is moderate. The operating instruction is to adjust the response parameters (monitor the vibration acceleration, observe whether it recovers on its own after the medium parameters recover, and check the fixing bolts if it does not recover).

[0204] Step S155: Based on the priority order of collaborative recovery, integrate the collaborative recovery operation instructions of each stage to form an ordered sequence of collaborative recovery operation instructions. The instruction sequence records the order of recovery of each stage, the operation connection requirements, and the collaborative matching standards during the recovery process.

[0205] The sequence of coordinated recovery operation instructions is arranged in priority as follows: 1. Recovery instruction for the action response stage; 2. Recovery instruction for the media interaction stage; 3. Recovery instruction for the electrical signal stage (combined with the cause blocking operation); 4. Recovery instruction for the structural vibration stage. The operation connection requirement is that the next stage can only be performed after the previous stage's recovery operation is completed and verified by the coordinated monitoring indicators. The coordinated matching standard is that the time response characteristics and amplitude correlation characteristics of each stage after recovery should be within the normal coordinated mode parameter range, such as action response delay ≤ 0.1 seconds and amplitude linkage ratio deviation ≤ ±5%.

[0206] Step S156: Based on the impact range of the core cause of the failure, the transmission stage of the abnormal collaborative chain, and the trend of changes in operating conditions, the handling process is divided into an emergency handling stage, an in-depth handling stage, and a consolidation stage. The impact range refers to the number of links and data types involved, and the transmission stage is divided into initial, spread, and stable stages.

[0207] The core cause of the failure affects four stages and four types of data. The abnormality chain is in a stable transmission phase (the scope of impact is no longer expanding), and the operating condition trend is currently at the production gap, after which high-load operation will resume. The handling process is divided into: emergency handling phase (current production gap, handling key stages to prevent the abnormality from spreading), in-depth handling phase (after production resumes and low-load operation is resumed, fully restoring all stages), and consolidation phase (long-term operation, monitoring and maintenance).

[0208] Step S157: Generate an emergency response operation instruction set. This emergency response operation instruction set shall preferentially include the cause blocking operation instruction set and the collaborative recovery operation instruction set for the higher priority links in the priority ranking, and associate each instruction with an execution time window, operation parameter constraints and collaborative monitoring indicator thresholds.

[0209] The emergency response operation instruction set includes: cause-blocking operation instructions (coil replacement) and coordinated recovery instructions for the action response process. The execution time window is 0-1.5 hours. Operational parameter constraints include that the new coil model must be identical to the original model, and the control signal pulse width adjustment range is 8-15ms. The coordinated monitoring index thresholds are: coil resistance 30±2Ω, insulation resistance ≥10MΩ, and valve core action delay ≤0.15 seconds (slightly higher than normal standards are permissible during emergency situations).

[0210] Step S158: Generate a deep treatment operation instruction set, which includes collaborative recovery operation instructions for the remaining links in the priority ranking, potential hidden danger detection instructions based on collaborative disturbance sensitivity analysis, and removal instructions for residual effects of the inducing factors.

[0211] The in-depth handling operation instruction set includes: collaborative recovery instruction for media interaction links, collaborative recovery instruction for structural vibration links, potential hidden danger detection instruction (checking whether there is poor contact in the coil power supply line; based on collaborative disturbance sensitivity analysis, poor line contact may cause similar collaborative anomalies), and instruction to remove residual effects of inducing factors (cleaning up metal powder that may be generated inside the solenoid valve due to coil failure).

[0212] Step S159: Generate a long-term monitoring parameter set, which includes the frequency of routine monitoring of collaborative behavior, key indicator thresholds, and the periodic verification cycle of the data behavior collaborative model, and is used to configure the monitoring system for continuous operation.

[0213] Long-term monitoring parameter set: Collaborative behavior is monitored routinely with complete multi-dimensional test data collected hourly for real-time analysis. Key indicator thresholds include the allowable range of response delay for each stage, the amplitude linkage ratio deviation threshold (±5%), and the normal sequence of collaborative logic. The data behavior collaboration model is periodically validated monthly using the normal operating data of that month.

[0214] Step S1510: Sort and encapsulate the cause-blocking operation instruction set, the collaborative recovery operation instruction sequence, the emergency treatment operation instruction set, the deep treatment operation instruction set, and the long-term monitoring parameter set according to the execution logic to generate an executable instruction sequence; associate preset operation constraints, safety interlock conditions, and effect evaluation index thresholds with the key operation steps in the executable instruction sequence; the operation constraints define parameter adjustment limits and operation order, the safety interlock conditions determine whether the operation is allowed to be executed based on environmental sensor data, and the effect evaluation index thresholds include a collaborative recovery rate threshold and an abnormal recurrence rate threshold, which are used to trigger automatic judgment of the treatment effect.

[0215] The instruction sets are ordered and packaged according to execution logic: Emergency response operation instruction set (0-1.5 hours) → Deep response operation instruction set (within 4 hours after production resumption) → Long-term monitoring parameter set (continuous execution). Among them, the cause blocking operation instruction set and the action response link recovery instruction set belong to emergency response, while the media interaction, structural vibration link recovery, hidden danger detection, and residue removal belong to deep response.

[0216] Key operational steps (such as coil replacement and control signal parameter adjustment) are subject to the following constraints: Coil replacement must be performed with the power off first; control signal pulse width adjustments must not exceed 15ms. Safety interlock conditions: Power-off operations must be performed only after confirming the supply voltage is 0V (based on voltage sensor data); ambient humidity during coil replacement must be ≤60% (based on ambient humidity sensor data). Effectiveness evaluation thresholds: Synergistic recovery rate threshold ≥90% (number of synergistic behaviors restored to normal / total number of abnormal behaviors); abnormal recurrence rate threshold ≤5% (number of abnormal synergistic behavior recurrences within 24 hours / total number of monitoring sessions). Upon completion, the monitoring system automatically determines the treatment effect based on these threshold indicators. If the targets are met, the treatment is considered complete; otherwise, a secondary treatment process is triggered.

[0217] In one exemplary embodiment, an intelligent fault diagnosis system for solenoid valve test data is provided. This system can be a terminal, server, etc., and its internal structure diagram can be as follows: Figure 2 As shown, the intelligent fault diagnosis system for solenoid valve test data includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computational and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, near-field communication, or other technologies. When the computer program is executed by the processor, it implements an intelligent fault diagnosis method for solenoid valve test data. The display unit is used to generate a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, or a button, trackball, or touchpad set on the housing of an intelligent fault diagnosis system for solenoid valve test data, or an external keyboard, touchpad, or mouse, etc.

[0218] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. An intelligent fault diagnosis method for solenoid valve test data, characterized in that, The method includes: Collect multidimensional test data of the solenoid valve under various operating conditions. The multidimensional test data includes action response data, media interaction data, structural vibration data, environmental impact data, and electrical signal data. A data behavior collaboration model is constructed based on the behavioral collaboration relationship between multidimensional test data. The data behavior collaboration model reflects the synchronous response rules, mutual triggering mechanisms and collaborative change patterns of different types of test data during the operation process. From the data behavior collaboration model, we can discover abnormal collaboration chains that deviate from the normal collaboration mode. The abnormal collaboration chain includes the collaboration abnormality initiation link, intermediate transmission link, terminal impact link and abnormal collaboration behavior between each link. The core cause of the failure is deduced by reverse engineering the abnormal collaborative chain. The core cause of the failure includes the cause type, triggering conditions, collaborative interference mode and initial impact link. Based on the transmission characteristics of the core causes of failures and the abnormal coordination chain, an executable instruction sequence is generated to control the maintenance system to perform corresponding fault handling operations. The method of reverse-engineering the core causes of failure through abnormal collaborative chains includes: The analysis focuses on the terminal impact link of the abnormal coordination chain, extracting the abnormal coordination performance characteristics of the terminal impact link, the types of test data involved, the details of data changes, and the combination of working conditions when the abnormality occurs. The abnormal coordination performance characteristics include the delay deviation value, the amplitude deviation ratio, and the form of coordination logic failure. Based on the transmission path of the abnormal collaboration chain, we trace back from the terminal impact link to the next higher intermediate transmission link, analyze the specific manifestations of the abnormal collaboration behavior of the intermediate transmission link, the source of the trigger signal and the way it affects the terminal impact link, and record the way the abnormal is transmitted. Extract the abnormal collaborative features of the intermediate transmission link at the previous level, compare them with the abnormal features of the terminal influence link, calculate the feature similarity between the two, and record the collaborative interference relationship between the two. The collaborative interference relationship is divided into direct interference and indirect interference. Continue to trace back to the previous intermediate transmission link, repeat the steps of abnormal collaborative feature extraction, feature comparison and interference relationship analysis, and trace upward step by step until the starting link of collaborative anomaly is traced, forming a complete reverse tracing chain; Integrate all abnormal collaborative features, interference relationships and operating condition information extracted during the reverse tracing process, construct a reverse transmission logic chain, and record the origin of abnormal collaborative behavior, transmission logic, changes in different links and the impact of operating conditions on the transmission process; Locate the exception triggering link in the reverse propagation logic chain. This exception triggering link is the initial source of all subsequent abnormal collaborative behaviors. Its exception directly triggers a chain reaction. This exception triggering link is identified as the initial impact link. Extract the test data sequence, abnormal coordination characteristics, acquisition scenarios and trigger signal details of the initial impact stage, and analyze the specific manifestations of the disruption of the normal coordination mode in this stage, including the satisfaction of coordination trigger conditions, deviation of response delay and abnormal changes in amplitude linkage ratio. An active collaborative perturbation simulation method is adopted, based on a data behavior collaborative model, to simulate the perturbation effect of different potential inducements on the initial impact stage, and to monitor the fit between simulated anomalies and actual anomalies. The system is associated with a pre-defined list of causes for solenoid valve failures. This list includes various types of causes that may lead to abnormal coordination, typical triggering conditions, common interference methods, and corresponding abnormal coordination manifestations. The types of causes are categorized as mechanical, electrical, media, and environmental. By comparing the abnormal collaborative characteristics, triggering scenarios, operating conditions, and simulated disturbance results of the initial impact phase with the items in the fault cause list, the degree of fit is calculated, and the candidate cause with the highest degree of fit is selected. Analyze the correlation between the candidate cause and all abnormal collaborative behaviors in the reverse propagation logic chain, verify that the candidate cause can explain all abnormal collaborative behaviors, including the abnormal differences, propagation rate and impact range of each link, and match with the operating condition information, determine that the candidate cause is the core cause of the fault, record its cause type, triggering conditions, collaborative interference mode and initial impact link, and form the final cause tracing result. The collaborative interference mode refers to the specific way of destroying collaborative behavior.

2. The intelligent fault diagnosis method for solenoid valve test data according to claim 1, characterized in that, The data behavior collaboration model constructed based on the behavioral collaboration relationships between multidimensional test data includes: The multidimensional test data is classified into types, and the original data sequences and data acquisition scenarios corresponding to action response data, media interaction data, structural vibration data, environmental action data and electrical signal data are recorded. The data acquisition scenarios include the operating load level of the solenoid valve, the physical characteristics of the medium, the ambient temperature and humidity conditions and the electrical power supply status. Extract the time response features of each type of original data sequence. The time response features reflect the response delay of the data after receiving the trigger signal and the change pattern in the response process, including the delay period from the moment the trigger signal is received to the start of data change, the duration of the data reaching a stable state, and the rate fluctuation performance in the change process. The amplitude correlation features of each type of original data sequence are extracted. The amplitude correlation features reflect the correspondence and linkage between the amplitude changes of the data and the amplitude changes of other types of data, including the proportional relationship of amplitude changes, the degree of phase synchronization, and the time difference of the amplitude peak. Analyze the collaborative behavior between different original data sequences within the same type of test data, and record the process by which a change in one data sequence triggers synchronous adjustments in other data sequences of the same category, including the trigger threshold for adjustment, the proportional relationship of the adjustment magnitude, and the duration of the stable state after adjustment. Analyze the cross-category collaborative behavior between different types of test data, and record the mutual triggering behavior between action response data and medium interaction data, structural vibration data and electrical signal data, and environmental action data and action response data, including the transmission path of trigger signals, response delay after triggering, and linkage law of response amplitude. Based on time response characteristics and amplitude correlation characteristics, establish collaborative behavior benchmark rules for various test data, record the collaborative boundaries and response standards between data under normal operating conditions, including the range of collaborative triggering conditions, the allowable range of response delay, and the amplitude linkage ratio standard under different operating conditions. Analyze the sensitivity of various test data to cooperative perturbation, and record the perception threshold of different data types to cooperative anomalies, that is, the specific manifestations of subsequent chain anomalies caused by data under a certain degree of cooperative deviation. The frequency and stability of different collaborative behaviors are quantified by statistical analysis methods. Each collaborative behavior is assigned a collaborative strength label to reflect the tightness of the collaborative relationship. The strength label is associated with the transmission ability and scope of influence of collaborative anomalies. Based on the collaborative behavior benchmark rules, collaborative disturbance sensitivity threshold, and collaborative strength identifier, a basic framework for a data behavior collaborative model is constructed. The nodes and edges of the data behavior collaborative model are defined. The nodes include data type, acquisition scenario, and feature parameters, while the edges include collaborative triggering conditions, response delay parameters, and amplitude linkage ratio. Cross-class and same-class collaborative behaviors are integrated into the basic framework, and the details of the collaborative triggering conditions, the dynamic adjustment range of the response delay parameter, and the fluctuation threshold of the amplitude linkage ratio corresponding to each edge are recorded to improve the collaborative logic of the data behavior collaborative model. The historical normal operation sequence of multidimensional test data is input into the model for collaborative verification to simulate the collaborative behavior under different working conditions. Based on the verification results, the collaborative parameters of the data behavior collaborative model are adjusted, including the weight allocation of the collaborative strength identifier and the calibration of the collaborative disturbance sensitivity threshold, to form a data behavior collaborative model that reproduces the normal collaborative mode.

3. The intelligent fault diagnosis method for solenoid valve test data according to claim 2, characterized in that, The method quantifies the frequency and stability of different collaborative behaviors through statistical analysis, assigns a collaborative strength label to each collaborative behavior to reflect the tightness of the collaborative relationship, including: Extract all instances of collaborative behavior from the historical, normally functioning multidimensional test data, and distinguish between instances of collaborative behavior of the same type and instances of collaborative behavior across different types. Instances of collaborative behavior of the same type refer to the collaboration of different sequences within the same data category, while instances of collaborative behavior across different types refer to the collaboration between different data categories. Count and count the instances of each type of collaborative behavior, record the total number of times the type of collaborative behavior occurs within a preset time period, and record the number of times the solenoid valve's operating condition changes and the type of operating condition within the time period, forming a frequency statistics result of collaborative behavior that includes operating condition association information; Analyze the duration of each type of collaborative behavior instance, calculate the time from trigger to end for each occurrence of this type of collaborative behavior, exclude abnormal durations caused by sudden changes in operating conditions, and form an effective duration sequence; The average duration and duration fluctuation range of this type of cooperative behavior are calculated based on the effective duration sequence. The average duration reflects the normal maintenance level of the cooperative behavior, and the duration fluctuation range reflects the temporal stability of the cooperative behavior. Analyze the consistency of data response in each instance of collaborative behavior, calculate the fit of each data sequence participating in the collaboration in terms of change magnitude and change rhythm, including the difference ratio of change magnitude and the sum of time differences in change rhythm, to form response consistency parameters; The statistical results of the frequency of collaborative behavior, average duration, duration fluctuation range, and response consistency parameters are standardized. Combined with the working condition correlation information, a quantitative logic for collaborative intensity is established, and the role ratio of each parameter in the quantification process is recorded. Among them, the frequency and consistency parameters are assigned preset main weights in the quantification logic, which directly determine the level of collaborative intensity. The strength of each collaborative behavior is calculated using quantitative logic, and the magnitude of the value reflects the tightness and stability of the collaborative relationship. Set the classification criteria for the synergy strength indicator, divide the indicator into different levels according to the distribution range of the synergy strength value, each level corresponds to a fixed value range, and each level is associated with the corresponding synergy anomaly transmission rate and impact range prediction. A dynamic adjustment method for the coordination strength is adopted, taking into account the impact of the solenoid valve's running time on the coordination relationship. That is, an increase in running time may lead to a decrease in coordination strength. An attenuation coefficient is set to correct the coordination strength value at different operating stages. The modified collaborative strength value of each collaborative behavior is compared with the classification standard, and a corresponding collaborative strength label is assigned to it. The label is bound to the triggering condition and response delay parameter of the collaborative behavior. By binding the coordination strength identifier with the corresponding coordination behavior, a list of coordination characteristics is formed, which includes the type of coordination behavior, frequency of occurrence, working condition association information, stability parameters, coordination strength identifier and attenuation coefficient.

4. The intelligent fault diagnosis method for solenoid valve test data according to claim 1, characterized in that, The method of mining abnormal collaboration chains that deviate from the normal collaboration pattern from the data behavior collaboration model includes: Extract normal collaboration mode parameters from the data behavior collaboration model, record the trigger condition thresholds, response delay standard ranges, amplitude linkage standard ratios, and collaboration disturbance sensitivity thresholds for various collaborative behaviors, and form a normal collaboration parameter set; The real-time collected multi-dimensional test data is input into the data behavior collaboration model to simulate the real-time collaborative behavior process and generate a real-time collaborative behavior parameter sequence that includes the trigger condition satisfaction status, the actual value of the response delay, and the actual proportion of amplitude linkage. By comparing the real-time collaborative behavior parameter sequence with the normal collaborative mode parameters, abnormal collaborative behaviors in real-time collaborative behavior are identified, such as trigger conditions not being met, response delays exceeding the standard range, amplitude linkage ratios deviating from the standard, or exceeding the collaborative disturbance sensitivity threshold. A cross-level collaborative anomaly identification method is adopted to distinguish between surface-level collaborative anomalies and deep-level collaborative anomalies. Surface-level collaborative anomalies refer to anomalies that are directly manifested as parameter deviations, while deep-level collaborative anomalies refer to anomalies where the parameters are not deviated but the collaborative logic is broken. Mark the test data category, data sequence and collection scenario corresponding to the first occurrence of abnormal collaborative behavior, determine the link where the abnormal collaborative behavior is located as the starting link of the collaborative anomaly, and record the time of occurrence of the anomaly, the source of the trigger signal and the initial abnormal behavior of the starting link; Track the impact of abnormal collaborative behavior at the initiation stage on subsequent related collaborative behaviors, and monitor the process by which the abnormal collaborative behavior causes other collaborative behaviors to deviate from the normal pattern based on the collaborative strength identifier and transmission logic in the data behavior collaborative model. Record the test data categories, data sequences, collection scenarios, and abnormal behaviors corresponding to the affected subsequent collaborative behaviors, including the triggering method of the abnormality, the degree of response deviation, and the proportion of amplitude linkage abnormalities, and determine the link where the affected collaborative behavior is located as the intermediate transmission link; Continuously track the spread of abnormal collaborative behavior in intermediate transmission links, analyze the changes in the transmission rate of abnormality under different operating conditions, until the scope of influence of abnormal collaborative behavior no longer expands, and determine the link where the final affected collaborative behavior is located as the terminal impact link. Analyze the relationships between the initiation, intermediate transmission, and terminal impact stages of collaborative anomalies, and record the transmission path of abnormal collaborative behavior from the initiation stage to the terminal stage, including direct and indirect transmission paths; Record the abnormal collaborative behavior, triggering relationship, impact degree and working condition association of each link in each transmission path to form a detailed description of abnormal collaborative behavior transmission. The detailed description of abnormal collaborative behavior transmission includes the abnormal parameter changes of each link, the association with the preceding link and the proportion of impact on the following link. By integrating information on the initiation, intermediate transmission, and terminal impact of collaborative anomalies, as well as the transmission paths, anomaly manifestations, and operational conditions between each stage, an anomaly collaboration chain is constructed.

5. The intelligent fault diagnosis method for solenoid valve test data according to claim 4, characterized in that, The process of tracking the impact of abnormal collaborative behavior at the initiation stage of abnormal collaboration on subsequent related collaborative behaviors, and monitoring the process by which this abnormal collaborative behavior causes other collaborative behaviors to deviate from the normal pattern, includes: Extract the abnormal collaborative behavior characteristics of the initiation link of the collaborative anomaly, record the triggering method, response deviation, amplitude linkage anomaly, and operating conditions when the anomaly occurs. The triggering method is divided into active triggering and passive triggering, and the response deviation is divided into delay deviation and amplitude deviation. Based on the data behavior collaboration model, query all related collaboration links that have a direct collaboration relationship with the starting link of the collaboration anomaly, sort them according to the collaboration strength identifier, give priority to the related links corresponding to the collaboration strength identifier, and identify potential collaboration objects that may be affected. Obtain the real-time test data sequence corresponding to the potential collaborative object, and analyze the changes in the trigger signal of the potential collaborative object after the abnormal collaborative behavior occurs in the initiation stage of the collaborative anomaly, including the amplitude change, frequency change and phase change of the trigger signal; Monitor the response process of potential collaborating objects after the trigger signal changes, compare the difference between their response delay and amplitude changes and the normal collaborating mode, calculate the delay deviation ratio and amplitude deviation ratio, and quantify the abnormal situation. Record the collaborative links that exhibit abnormal response delays or amplitude changes among potential collaborative objects, mark them as initially affected collaborative links, and record the time difference of the abnormality and the changes in operating conditions. The time difference refers to the time interval between the abnormality and the initial link. Analyze the abnormal collaborative behavior characteristics of the initially affected collaborative links, compare them with the abnormal characteristics of the collaborative abnormal initiation link, and determine whether there is a causal relationship between them and the abnormal collaborative behavior of the collaborative abnormal initiation link. The relationship judgment is based on the transmission logic and collaborative strength in the collaborative model. Based on the preset collaborative logic rules in the data behavior collaborative model, it is determined whether the causal relationship conforms to the transmission path and triggering conditions defined by the model. Abnormal collaborative behaviors that do not conform to the preset rules are marked as isolated anomalies and are not included in the transmission sequence of the current abnormal collaborative chain. Track the impact of abnormal collaborative behavior of initially affected collaborative links on their own subsequent collaborative links, repeat the steps of potential collaborative object query, trigger signal analysis, response process monitoring and anomaly marking to form a multi-level affected chain; Record the abnormal behavior of each affected collaborative link, the association method with the abnormal collaborative behavior of the previous level, and the time interval of the impact transmission. At the same time, record the operating parameters corresponding to the abnormal collaborative behavior of this level. Abnormal behavior includes delay deviation, amplitude deviation, and collaborative logic destruction. The association method is divided into direct triggering and indirect triggering. All affected collaborative links are arranged in the order of impact transmission, and the degree of abnormality, correlation, and working condition correlation information of each link are marked to form a transmission sequence of abnormal collaborative behavior, reflecting the spread process of the abnormality from the initial link and the changing pattern under different working conditions.

6. The intelligent fault diagnosis method for solenoid valve test data according to claim 1, characterized in that, The pre-defined list of potential causes for solenoid valve malfunctions includes various types of causes that may lead to abnormal coordination, typical triggering conditions, and common interference methods, including: Collect all fault cause information from historical failure cases of solenoid valves. The fault cause information covers different types such as mechanical structure wear, electrical component aging, changes in media composition, and abnormal environmental temperature and humidity. Analyze the complete records of each historical failure case, extract the operating conditions, running time, previous operation records and abnormal signs before the failure occurred, and form a typical trigger condition description. The operating conditions include operating load, medium parameters, environmental parameters and electrical parameters. The study analyzes how each historical fault cause interferes with the collaborative behavior of multidimensional test data. Through experimental simulation and data analysis, it records the process by which the causes affect the trigger response, amplitude linkage, and time synchronization of the data. The impact of the trigger response includes extending the response delay and changing the response threshold. The impact of amplitude linkage includes disrupting the amplitude ratio and changing the phase synchronization. The impact of time synchronization includes disrupting the collaborative triggering sequence. Record the abnormal collaborative behavior caused by each historical fault, including the type of abnormal collaborative behavior, transmission path, scope of impact, degree of abnormality in each link, and changes in the rate of abnormal transmission. The types of abnormal collaborative behavior are divided into surface and deep, the transmission path is divided into direct and indirect, and the scope of impact is divided into single link and multiple links. The collected fault cause information is classified and organized, and a classification directory is established according to the cause type. Each directory is further subdivided into subdirectories according to the similarity of the triggering conditions. Each subdirectory contains the corresponding typical triggering conditions, interference methods and collaborative abnormal manifestations. Supplement the publicly available research data, technical standards and manufacturer maintenance manuals on the causes of solenoid valve failures in the industry, improve the details of the triggering conditions, the description of the interference methods and the quantitative indicators of abnormal performance for various causes, and include specific parameter ranges in the details of the triggering conditions. The collected fault cause information is standardized, and a unified expression method is used to describe the cause type, triggering condition, interference method and abnormal manifestation. The triggering condition must specify the parameter type and manifestation form, the interference method must specify the target and method of interference, and the abnormal manifestation must specify the characteristic parameters and the range of change. Assign a unique identifier to each fault cause and establish an index linking the cause information with historical fault cases. The index includes information such as case number, fault occurrence time, and handling result. Construct a structured list of fault causes, which includes fields such as cause identifier, cause type, subcategory, typical triggering conditions, cooperative interference methods, typical abnormal manifestations, related case index, and update time. Typical triggering conditions include operating parameters, and cooperative interference methods include the mode of action. Establish a dynamic update mechanism for the list, regularly collect new failure cases and industry research results, add information on newly emerging failure causes to the list, optimize the description of existing causes, adjust the classification method to adapt to new failure types, and optimize the description of existing causes by refining trigger conditions and supplementing interference methods.

7. The intelligent fault diagnosis method for solenoid valve test data according to claim 1, characterized in that, Based on the transmission characteristics of the core fault cause and the abnormal coordination chain, an executable instruction sequence is generated to control the maintenance system to perform corresponding fault handling operations, including: Analyze the type, triggering conditions, cooperative interference methods, and characteristics of the initial impact links of the core causes of the failure. Combine the transmission path of the abnormal cooperative chain and the operating condition correlation information to determine the key intervention points that can prevent the causes from continuing to play a role. Record the test data links, operating conditions, and intervention timing windows corresponding to the intervention points. Based on key intervention points, a set of targeted trigger blocking operation instructions is generated. The trigger blocking operation instruction set includes the specific content of the intervention operation, the execution order, the required tools and equipment, the operating environment requirements, and the collaborative monitoring indicators during the operation. The specific content of the intervention operation includes adjusting parameters, replacing components, and isolating interference sources. The execution order is sorted according to the priority of impact. The operating environment requirements include temperature, humidity, and safety protection conditions. The transmission characteristics of the abnormal collaborative chain are analyzed, including the transmission path length, the degree of collaborative dependence of each link, the anomaly propagation speed, the impact of the operating conditions on the transmission, and the reversibility of the anomaly. The priority order of collaborative recovery is determined. The transmission path length refers to the number of links, the degree of collaborative dependence is based on the collaborative strength indicator, and the anomaly propagation speed refers to the transmission time between each link. A collaborative adaptive recovery approach is adopted. For each link in the abnormal collaborative chain, corresponding collaborative recovery operation instructions are generated based on its abnormality, collaborative dependency, and reversibility. The specific methods of adjusting the response parameters of the test data and reconstructing the collaborative logic are recorded. Adjusting the response parameters of the test data includes correcting the delay and adjusting the amplitude ratio. Reconstructing the collaborative logic includes repairing the trigger relationship. Based on the priority order of collaborative recovery, the collaborative recovery operation instructions of each stage are integrated to form an ordered sequence of collaborative recovery operation instructions. The sequence of collaborative recovery operation instructions records the order of recovery of each stage, the operation connection requirements, and the collaborative matching standards during the recovery process. Based on the impact range of the core cause of the failure, the transmission stage of the abnormal collaborative chain, and the trend of changes in operating conditions, the handling process is divided into an emergency handling stage, an in-depth handling stage, and a consolidation stage. The impact range refers to the number of links and data types involved, and the transmission stage is divided into initial, spread, and stabilization stages. An emergency response operation instruction set is generated, which prioritizes the cause blocking operation instruction set and the collaborative recovery operation instruction set for the higher priority steps. Each instruction is associated with an execution time window, operation parameter constraints and collaborative monitoring indicator thresholds. Generate a set of deep treatment operation instructions, which includes collaborative recovery operation instructions for the remaining links in the priority ranking, potential hidden danger detection instructions based on collaborative disturbance sensitivity analysis, and instructions for clearing residual effects of the causes. Generate a long-term monitoring parameter set, which includes the frequency of routine monitoring of collaborative behavior, key indicator thresholds, and the periodic verification cycle of the data behavior collaborative model, and is used to configure the monitoring system for continuous operation. The instruction sets for blocking the cause, the sequence of instructions for coordinated recovery, the instruction sets for emergency treatment, the instruction sets for deep treatment, and the long-term monitoring parameter set are sorted and encapsulated according to execution logic to generate an executable instruction sequence. Pre-set operation constraints, safety interlock conditions, and effect evaluation index thresholds are associated with the key operation steps in the executable instruction sequence. The operation constraints define parameter adjustment limits and operation order. The safety interlock conditions determine whether the operation is allowed to be executed based on environmental sensor data. The effect evaluation index thresholds include a coordinated recovery rate threshold and an abnormal recurrence rate threshold, which are used to trigger automatic judgment of the treatment effect.

8. An intelligent fault diagnosis system for solenoid valve test data, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the intelligent fault diagnosis method for solenoid valve test data as described in any one of claims 1 to 7 by executing the machine-executable instructions.

9. A computer program product, characterized in that, The computer program product includes machine-executable instructions stored in a computer-readable storage medium. The processor of the intelligent fault diagnosis system for solenoid valve test data reads the machine-executable instructions from the computer-readable storage medium and executes the machine-executable instructions, causing the intelligent fault diagnosis system for solenoid valve test data to perform the intelligent fault diagnosis method for solenoid valve test data as described in any one of claims 1 to 7.

Images