Multi-module integrated solenoid valve working state detection system

Abstract

The invention relates to the technical field of electromagnetic valve monitoring, and particularly discloses a multi-module integrated electromagnetic valve working state detection system which comprises an electromagnetic valve data acquisition module, a data processing module, a dynamic extraction module, a fault diagnosis module, a verification module and an early warning module. The electromagnetic valve data acquisition module is used for acquiring an action detection data set of the multi-module integrated electromagnetic valve; the data processing module is used for acquiring electromagnetic valve action characteristic data corresponding to each module integrated electromagnetic valve according to the action detection data set; and the dynamic extraction module is used for acquiring electromagnetic valve action pull-in data, electromagnetic valve action release data and electromagnetic valve action touch current data according to each piece of electromagnetic valve action characteristic data. According to the multi-module integrated solenoid valve working state detection system, through cooperative operation of the six modules, the problem of traditional response lag is solved, and it is ensured that emergency faults are handled in time. Full-period accurate monitoring and fault early warning are integrally realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electromagnetic valve monitoring, and in particular to a multi-module integrated electromagnetic valve working state detection system. BACKGROUND

[0002] As the core executive element for controlling fluid on-off in industrial automation systems, the stability of the working state of the electromagnetic valve directly affects the safety and efficiency of the entire industrial system. Multi-module integrated electromagnetic valves are widely used in chemical industry, energy, intelligent manufacturing and other fields because they can realize complex fluid control logic.

[0003] The working state sampling strategy of the multi-module integrated electromagnetic valve mainly focuses on the current signal in the steady state maintenance stage, and lacks coverage of the current rising characteristics in the power pre-absorption stage, the current dynamic change in the valve core opening movement stage, and the current decay process in the power-off acceleration release stage, etc. Key dynamic stages, resulting in early signals of valve core start-up delay, initial movement jam and other faults being missed; at the same time, for the multi-module integrated scene, independent marking and synchronous collection of data of each module are not realized, parameters of different modules are confused, it is difficult to distinguish individual performance differences, and fault positioning accuracy is affected. SUMMARY

[0004] The purpose of the present application is to provide a multi-module integrated electromagnetic valve working state detection system to solve the technical problems raised in the background.

[0005] To achieve the above purpose, the present application provides the following technical scheme: A multi-module integrated electromagnetic valve working state detection system, comprising: An electromagnetic valve data acquisition module, a data processing module, a dynamic extraction module, a fault diagnosis module, a verification module, and a warning module. The electromagnetic valve data acquisition module is used to obtain action detection data sets of the multi-module integrated electromagnetic valve. The data processing module is used to obtain electromagnetic valve action characteristic data corresponding to each module integrated electromagnetic valve according to the action detection data sets. The dynamic extraction module is used to obtain electromagnetic valve action absorption data, electromagnetic valve action release data, and electromagnetic valve action touch current data according to each electromagnetic valve action characteristic data. The fault diagnosis module is used to obtain electromagnetic valve judgment information according to the electromagnetic valve action absorption data, the electromagnetic valve action release data, and the electromagnetic valve action touch current data. The verification module is used to generate electromagnetic valve verification information according to the electromagnetic valve judgment information. The warning module is used to generate warning information according to the electromagnetic valve verification information.

[0006] Preferably, the electromagnetic valve data acquisition module acquires the first current signal and the starting current data of the energized pre-absorption stage, the second current signal and the spool absorption time data of the spool opening movement stage, the third current signal and the electromagnetic valve energized time of the steady-state open stage, and the fourth current signal and the steady-state current data of the de-energized accelerated release stage of each module integrated electromagnetic valve, generates the multi-stage signal integration information of each module integrated electromagnetic valve according to the first current signal, the second current signal, the third current signal, and the fourth current signal, generates the action characteristic data according to the starting current data, the steady-state open stage, the electromagnetic valve energized time, and the steady-state current data, and integrates the multi-stage signal integration information and the action characteristic data to obtain the action detection data set of each module integrated electromagnetic valve.

[0007] Preferably, the data processing module synchronizes the action structured data through time stamp synchronization multi-channel acquisition of the action detection data set, wherein the action structured data includes action division time stamp, action stage current value, and action stage label, acquires current feature point positioning-stage division information according to the action division time stamp, the action stage current value, and the action stage label, generates a current-stage curve according to the current feature point positioning-stage division information, acquires the dynamic stage characteristic data set of the energized pre-absorption stage, the spool opening movement stage, the steady-state open stage, and the de-energized accelerated release stage according to the current-stage curve, and acquires the electromagnetic valve action characteristic data according to the dynamic stage characteristic data set.

[0008] Preferably, the dynamic extraction module is used to acquire the energized absorption stage data segment, the action release stage data segment, and the touch point related data window according to the electromagnetic valve action characteristic data, acquire the absorption total time and the absorption current change rate according to the absorption stage data segment, acquire the absorption energy data according to the absorption total time and the absorption current change rate, and take the absorption energy data as the electromagnetic valve action absorption data, acquire the release total time and the release current change rate according to the action release stage data segment, acquire the release peak current data according to the release total time and the release current change rate, and take the release peak current data as the electromagnetic valve action release data, acquire the touch current and the touch delay time according to the touch point related data window, acquire the touch current to steady-state current ratio according to the touch current and the touch delay time, and take the touch current to steady-state current ratio as the electromagnetic valve action touch current data.

[0009] Preferably, the fault diagnosis module is used to associate the three groups of data of the electromagnetic valve action suction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data through the action cycle timestamp to generate a single cycle characteristic parameter, obtain a preliminary abnormal marker parameter according to the single cycle characteristic parameter, obtain a composite fault determination information according to the preliminary abnormal marker parameter, compare the composite fault determination information with a matching fault characteristic library and divide the levels to obtain fault level-type information, and obtain electromagnetic valve judgment information according to the fault level-type information.

[0010] Preferably, the fault diagnosis module is further used to extract the action cycle identifier and the action cycle timestamp corresponding to the electromagnetic valve action suction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data, align the action cycle identifier and the action cycle timestamp corresponding to the electromagnetic valve action suction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data to a sorting time axis, obtain a single cycle time range according to the sorting time axis, obtain the correlation characteristic parameter of the electromagnetic valve action suction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data according to the single cycle time range, obtain an abnormal correlation marker parameter based on the correlation characteristic parameter, and generate a single cycle characteristic parameter according to the correlation characteristic parameter and the abnormal correlation marker parameter.

[0011] Preferably, the fault diagnosis module is further used to obtain a basic information package according to the fault level and the fault type associated with the action cycle identifier and the action cycle timestamp, extract the specific parameter measured value and the standard value and the deviation rate of triggering the fault from a single cycle characteristic parameter set according to the basic information package to obtain a triggering parameter detail, obtain a structured judgment information according to the triggering parameter detail, compare and verify the structured judgment information with a historical similar fault record to obtain electromagnetic valve judgment information.

[0012] Preferably, the fault diagnosis module is further used to obtain a classification characteristic parameter according to a single cycle characteristic parameter, call a threshold library of a corresponding electromagnetic valve model according to the classification characteristic parameter, match the normal range of each parameter to obtain a parameter normal threshold, determine a single cycle characteristic parameter based on the parameter normal threshold to obtain a single abnormal marker, obtain an associated abnormal marker according to the single abnormal marker, and generate an abnormal level according to the single abnormal marker and the associated abnormal marker, and label the single cycle characteristic parameter to obtain an abnormal marker parameter according to the abnormal level.

[0013] Preferably, the verification module is configured to acquire electromagnetic valve core verification elements according to the electromagnetic valve judgment information, wherein the electromagnetic valve core verification elements include electromagnetic valve key parameters, electromagnetic valve time dimension information and electromagnetic valve state identification, historical judgment information corresponding to a preset period is extracted based on the electromagnetic valve key parameters, the electromagnetic valve time dimension information and the electromagnetic valve state identification, and historical trend comparison result information is obtained by comparing the historical judgment information, the electromagnetic valve standard parameter database corresponding to the type is called according to the historical trend comparison result information, and the database matching degree is obtained by comparing the historical trend comparison result information with the electromagnetic valve standard parameter database, and the electromagnetic valve matching verification information is acquired according to the database matching degree.

[0014] Preferably, the early warning module is configured to acquire early warning trigger element information according to the electromagnetic valve verification information, wherein the early warning trigger element information includes a fault type and an original judgment level, a preliminary early warning level is generated according to the fault type and the original judgment level, a differential structured content is generated according to the preliminary early warning level, and early warning information is acquired according to the differential structured content.

[0015] The multi-module integrated electromagnetic valve working state detection system of the present application has the following beneficial effects: the data acquisition module realizes non-invasive full-cycle signal acquisition, solves the problems of traditional invasive installation interference and incomplete sampling, and guarantees data integrity; the data processing module eliminates multi-channel asynchronous error through time synchronization and accurate stage division, and improves the reliability of feature extraction; the dynamic extraction module accurately quantifies suction, release and triggering parameters, solves the problem of traditional parameter extraction isolation, and provides fine-grained features for fault analysis; the fault diagnosis module breaks through the limitations of single parameter misjudgment through multi-parameter association and level division, and realizes accurate identification of fault type and level; the verification module improves the reliability of diagnosis results by comparing historical trends and standard databases, and eliminates occasional interference; the early warning module pushes differential content by level, solves the problem of traditional response lag, and ensures timely handling of emergency faults. The whole realizes accurate monitoring and early warning of faults. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 FIG. 1 is a system structure schematic diagram of an embodiment of the present application.

[0017] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0018] It should be understood that the specific embodiments described herein are merely intended to explain the present application, and are not intended to limit the present application.

[0019] As Figure 1As shown, the application provides a multi-module integrated electromagnetic valve working state detection system, comprising: An electromagnetic valve data acquisition module, a data processing module, a dynamic extraction module, a fault diagnosis module, a verification module, and a warning module. The electromagnetic valve data acquisition module is used to obtain action detection data sets of the multi-module integrated electromagnetic valve. The data processing module is used to obtain electromagnetic valve action characteristic data corresponding to each module integrated electromagnetic valve according to the action detection data sets. The dynamic extraction module is used to obtain electromagnetic valve action attraction data, electromagnetic valve action release data, and electromagnetic valve action touch current data according to each electromagnetic valve action characteristic data. The fault diagnosis module is used to obtain electromagnetic valve judgment information according to the electromagnetic valve action attraction data, the electromagnetic valve action release data, and the electromagnetic valve action touch current data. The verification module is used to generate electromagnetic valve verification information according to the electromagnetic valve judgment information. The warning module is used to generate warning information according to the electromagnetic valve verification information.

[0020] Through the cooperation of the electromagnetic valve data acquisition module, the data processing module, the dynamic extraction module, the fault diagnosis module, the verification module, and the warning module, the application realizes current signal acquisition, feature extraction, fault diagnosis, result verification, and hierarchical warning of the full action cycle of the multi-module integrated electromagnetic valve, finally accurately monitors the working state, predicts potential faults in advance, and ensures the stable operation of the industrial system.

[0021] The electromagnetic valve data acquisition module collects the current signals and key parameters of the full action cycle of the multi-module integrated electromagnetic valve in a non-invasive manner, and generates complete action detection data sets. It solves the problems of traditional invasive installation affecting production continuity, incomplete sampling missing key features, weak anti-interference ability, etc. Through an open electromagnetic induction current sensor, the current signals are synchronously collected in four stages of power-on pre-attraction, valve core opening movement, steady-state open, and power-off accelerated release. Combined with pre-processing such as sliding average filtering and 3σ rule outlier rejection, characteristic parameters such as power-on initial current, attraction time, and steady-state current are extracted. Finally, the multi-stage signals and feature data are associated according to the time stamp to form a structured data set containing "time-current value-stage label-feature parameter", providing complete and reliable raw data support for subsequent analysis, and ensuring that the full action features of the electromagnetic valve from power-on to closing are covered.

[0022] The data processing module processes the action detection data set to generate action characteristic data reflecting the dynamic performance of the electromagnetic valve, solving problems such as traditional multi-channel asynchronization, rough feature point positioning, and fuzzy stage division. Through time stamp synchronization multi-channel data acquisition, the sampling delay of each module is eliminated to generate action structured data containing action division time stamp, corresponding stage current value and stage label; the second-order difference method is used to accurately locate the key feature points (such as power-on starting point, valve core in-place point, etc.) of the current curve; based on these feature points, four action stages are divided to generate current-stage curve graphs, and characteristic parameters such as the rise or fall slope, steady-state current, and fluctuation coefficient of each stage are extracted, and finally integrated into multi-module, multi-cycle electromagnetic valve action characteristic data, providing standardized and highly consistent input for subsequent parameter extraction and fault diagnosis, ensuring the comparability of parameters of different cycles and modules.

[0023] The dynamic extraction module accurately extracts the suction, release and touch current related key parameters from the action characteristic data, quantifies the dynamic response performance of the electromagnetic valve, and solves problems such as low automation, fuzzy definition and ignoring parameter correlation in traditional parameter extraction. By locating the data window related to the suction, release and touch points, suction data such as suction total time, suction current change rate and suction energy, release data such as release total time, release current change rate and release peak current, and touch current data such as touch current, touch delay time and touch current to steady-state current ratio are extracted; at the same time, parameter correlation verification is carried out to ensure the physical reasonableness of the parameters.

[0024] The fault diagnosis module generates fault judgment information based on the suction, release and touch current data, solving problems such as traditional dependence on single parameter, fixed threshold and lack of historical trend analysis. The three types of data (suction, release and touch current data) are associated by using the action cycle time stamp to generate a single cycle characteristic parameter set containing basic parameters and their associated characteristic parameters; the model adaptation threshold library is called to mark single parameter and associated abnormalities; based on the association rule, the compound fault (such as valve core sticking risk and coil aging trend) is determined; the fault characteristic library is matched to divide the fault level, and finally integrated into structured judgment information containing fault type, level, trigger parameter, processing suggestion and trend prediction. This process combines multi-parameter correlation analysis and historical trend to improve the accuracy of fault identification and provide basis for maintenance decision.

[0025] The verification module improves the reliability of fault judgment information through multi-dimensional verification, and solves the problems of traditional verification links, single method, and unquantized reliability. Key parameters, time dimension information, and state identifiers are extracted from the judgment information; historical data of a preset period are traced back, and the current fault is compared with the historical trend; the high-precision original data of the period in the historical record are reviewed to verify the data consistency; the same type of standard parameter database is called to compare the deviation of the current parameter from the group mean; finally, the historical trend consistency, secondary sampling consistency, and standard database matching are quantified to measure the reliability, and the electromagnetic valve verification information containing the verification conclusion and the correction suggestion is generated, which eliminates the misjudgment caused by accidental interference and ensures the reliability of the diagnosis result.

[0026] The early warning module generates graded early warning information based on the verification information, and solves the problems of traditional early warning level, limited push channel, and no closed-loop tracking. The fault type, original judgment level, reliability, and parameter deterioration rate are extracted from the verification information; the early warning level is dynamically determined; differentiated content is generated according to the level (level one contains emergency operation and position, level two contains trend and detection suggestion, and level three contains instantaneous abnormality); the differentiated content is pushed through multiple channels, and the early warning frequency and processing time are archived and counted.

[0027] In one embodiment, the electromagnetic valve data acquisition module acquires the first current signal and the starting current data in the power pre-absorption stage, the second current signal and the solenoid valve core absorption time data in the valve core opening movement stage, the third current signal and the solenoid valve power time in the steady-state open stage, and the fourth current signal and the steady-state current data in the power-off acceleration release stage of each module integrated electromagnetic valve, generates the multi-stage signal integration information of each module integrated electromagnetic valve according to the first current signal, the second current signal, the third current signal, and the fourth current signal, generates the action characteristic data according to the starting current data, the steady-state open stage, the solenoid valve power time, and the steady-state current data, and integrates the multi-stage signal integration information and the action characteristic data to obtain the action detection data set of each module integrated electromagnetic valve.

[0028] As described above, for each module integrated solenoid valve, accurate data is collected in four stages of action cycle, and current characteristics and key parameters covering the whole process of "power on - open - maintain - close" are obtained. In the power pre-absorption stage, the first current signal reflecting the initial excitation characteristics of the coil and the power-on starting current are collected, the current rising process is captured by non-intrusive sensor to avoid interference with the normal work of the circuit; in the valve core opening movement stage, the second current signal reflecting the conversion of electric energy and mechanical energy and the valve core absorption time are collected, the starting point, valley and action time of current drop are recorded, and the key time points are identified by algorithm to reflect the valve core movement state; in the steady-state maintaining open stage, the third current signal reflecting the energy consumption stability and the solenoid valve power-on time are collected, the effective fluctuation characteristics are retained after filtering processing, and the state of the coil maintaining open is reflected; in the power-off accelerated release stage, the fourth current signal reflecting the reset performance and the steady-state current data are collected, the power-off instantaneous change is captured by hardware trigger mechanism, and the steady-state current is taken as the reference to judge the release stage abnormality. The whole process marks the data separately according to the modules to ensure that the data can be distinguished among multiple modules, and solves the problems of missing dynamic process in traditional collection and chaotic data among multiple modules.

[0029] Then, the current signals collected in stages are integrated into continuous full-cycle characteristic curves. The first to fourth current signals of the four stages are spliced in time sequence based on time stamp to form a coherent current-time curve; clear stage labels (such as "pre-absorption", "opening movement", etc.) and key time points are added to the curve to clearly define the time boundaries of each stage; at the same time, the integrated signals are marked according to the module number to ensure that the full-cycle characteristics of different modules can be analyzed separately. Through this process, the originally dispersed stage signals are associated as complete action cycle records, solving the analysis deviation problem caused by stage separation in traditional collection, and enabling the current signal to intuitively reflect the whole process change of the solenoid valve from power-on to closing, providing continuous and traceable basic data for subsequent stage division and feature extraction.

[0030] After that, the original current signals are converted into quantifiable and comparable feature parameters to accurately describe the performance of the solenoid valve. Key parameters are extracted from the feature data collected in stages, including power-on starting current, valve core absorption time, solenoid valve power-on time, steady-state current average, etc., which directly reflect the electrical and mechanical performance of the solenoid valve; at the same time, derived features such as current rising slope in pre-absorption stage and current drop rate in release stage are obtained; all feature parameters are associated with corresponding module number and cycle number to form structured data, ensuring that each parameter can be corresponded to a specific module and action cycle. Through this process, the abstract current signal is converted into specific performance indicators, solving the problem of only signal existing and lacking of quantitative parameters in traditional collection, and enabling the performance of different cycles and different modules to be directly compared through parameters.

[0031] The multi-stage signal integration information is bound with the action feature data to form a complete and correlated structured data set. The correlation index is established through the module number and the cycle number to ensure that the current signal curves and the feature parameters of the same module and the same cycle can be synchronously called to realize the bidirectional correlation of "signal tracing parameter and parameter corresponding signal". The standardized format is used for storage to facilitate the database import and quick query. The redundancy check is performed. For example, if the stage signal is missing or the feature parameter is abnormal, it is marked as "to be supplemented" and triggers the re-collection to guarantee the data integrity. Through this integration, the problem of traditional signal and parameter separation is solved. The generated data set contains not only the detailed information of the original signal but also the quantitative feature parameters, which provides comprehensive and reliable basic data for the subsequent links such as stage division of the data processing module and abnormal identification of the fault diagnosis module, and ensures that the entire detection system can carry out analysis based on complete information.

[0032] In one embodiment, the data processing module obtains action structured data by time stamp synchronizing multi-channel collection on the action detection data set, wherein the action structured data includes action division time stamp, action stage current value and action stage label, acquires current feature point positioning-stage division information according to the action division time stamp, action stage current value and action stage label, generates current-stage curve diagram according to the current feature point positioning-stage division information, acquires dynamic stage characteristic data set of the power pre-absorption stage, the valve core opening movement stage, the steady-state open stage and the power-off acceleration release stage according to the current-stage curve diagram, and acquires electromagnetic valve action characteristic data according to the dynamic stage characteristic data set.

[0033] As described above, the time stamp synchronizing multi-channel collection generates the action structured data, which is used to solve the time asynchronization problem of the multi-module collection data, ensures the alignment of the channel data on the time axis, and lays a foundation for cross-module performance analysis. First, the multi-module data in the action detection data set is imported, and the time stamp and the module identifier of each data point are extracted. The time deviation of other modules is obtained based on the action start time stamp of the main module. If the deviation exceeds the threshold, the time synchronization error is controlled within a very small range through linear interpolation correction. Finally, the corrected time stamp, the current value and the stage label at the corresponding moment are integrated to form the action structured data including "action division time stamp, action stage current value and action stage label". This process eliminates the time deviation caused by hardware delay, ensures that the same action stage of different modules can be directly compared, avoids the analysis error caused by time asynchronization in the traditional processing, and provides a reliable time reference for subsequent cross-module performance evaluation.

[0034] Then, the problems of traditional positioning roughness and stage boundary ambiguity are solved by locating the feature points of the current curve and dividing the action stages. First, the current signal is filtered to eliminate high-frequency noise interference; the inflection point of the current curve is identified by the second-order difference method (the second-order difference extreme point corresponds to the maximum curvature change), and the current change trend (rising or falling) is determined by the first-order difference to accurately locate the key feature points such as the power-on starting point, the current inflection point, and the steady-state end point; based on these feature points, the power-on pre-absorption, valve core opening movement, steady-state open, and power-off accelerated release stages are clearly divided, and the positioning-division information containing the feature point timestamp and stage label is output. Through algorithmic positioning, the feature point recognition accuracy is significantly improved, avoiding the subjective error of manual interpretation or fixed threshold method, ensuring clear stage boundaries and providing accurate time range basis for subsequent stage-based parameter extraction.

[0035] After that, the relationship between current change and stage division is visualized to solve the problem of abstract original data, which is convenient for manual review and intuitive analysis. With time as the horizontal axis and current value as the vertical axis, the "timestamp-current value" in the action structured data is mapped into a curve; the stage labels (such as different colors to distinguish stages) and key feature points (such as inflection points marked with specific symbols) are labeled on the curve; multiple module data can be distinguished by different line types or drawn separately; finally, the curve graph is stored in vector format, supporting detailed zoom-in viewing. The curve graph intuitively presents the current change trend of the electromagnetic valve during the whole action cycle, making the correspondence between feature points and stages clear at a glance, which not only facilitates technicians to review the accuracy of algorithmic positioning (such as correcting feature points with large deviations), but also provides intuitive evidence for fault diagnosis (such as directly observing the current fluctuation in the steady-state stage from the graph).

[0036] Then, stage-based quantitative parameters are extracted to solve the problem of traditional parameter fragmentation and inability to reflect stage dynamic characteristics. For the four divided stages, the key characteristics are extracted: the current rising slope and touch current in the pre-absorption stage reflect the initial magnetizing capacity of the coil; the current falling slope and absorption time in the opening movement stage represent the valve core movement performance; the steady-state current mean and fluctuation coefficient in the steady-state maintenance stage reflect the energy consumption and stability; the current falling rate and release time in the release stage represent the reset performance. These parameters are integrated according to "module number-cycle number-stage label" to form a dynamic stage characteristic dataset. Through stage-based extraction, each parameter is bound to a specific physical meaning (such as the fluctuation coefficient reflecting the coil stability), achieving multi-dimensional quantification of the electromagnetic valve during the whole action cycle and providing fine-grained feature basis for subsequent fault diagnosis.

[0037] Finally, by integrating parameters from each stage, the problem of parameter dispersion is solved, providing structured input for subsequent modules. Characteristic parameters of each stage within the same module and cycle are associated through cycle numbers to ensure temporal consistency. The logical relationships between parameters (such as the sum of total time and the time of each stage) are verified; if the deviation exceeds a threshold, an anomaly is marked and the parameters are re-extracted. Ultimately, solenoid valve action characteristic data containing "module identifier, cycle identifier, characteristic parameters of each stage, and verification status" is generated.

[0038] In one embodiment, the dynamic extraction module is used to obtain a data segment for the energized engagement stage, a data segment for the action release stage, and a data window related to the contact point based on the solenoid valve's action characteristic data; to obtain the total engagement time and the rate of change of engagement current based on the engagement stage data segment; to obtain engagement energy data based on the total engagement time and the rate of change of engagement current; and to use the engagement energy data as the solenoid valve's action engagement data; to obtain the total release time and the rate of change of release current based on the action release stage data segment; to obtain release peak current data based on the total release time and the rate of change of release current; and to use the release peak current data as the solenoid valve's action release data; to obtain the contact current and the contact delay time based on the data window related to the contact point; to obtain the ratio of contact current to steady-state current based on the contact current and the contact delay time; and to use the ratio of contact current to steady-state current as the solenoid valve's action contact current data.

[0039] As described above, extracting three key stage data segments from the solenoid valve's action characteristic data is crucial for accurately locating the key data segments corresponding to the engagement, release, and actuation processes, thus solving the problem of distorted physical meaning of parameters caused by confusing stage boundaries in traditional methods. Specifically, the module calls the "current feature point positioning - stage division information" output by the data processing module, and separates three key data segments based on the timestamp boundaries (such as energization start, engagement end, de-energization start, etc.) and stage labels: the energization engagement stage (covering the time range from energization to full valve core opening), the action release stage (covering the time range from de-energization to full valve core closing), and the actuation point related data window (covering the time range from energization to valve core activation). Each data segment contains the current time sequence of the corresponding interval, stage labels, and feature point timestamps. Simultaneously, by verifying the temporal continuity of each data segment (e.g., the end time of the engagement stage must be earlier than the start time of the release stage), timing errors are marked and corrected to ensure that the data segments strictly match the physical process of "engagement-release-actuation".

[0040] Based on the data segment of the attraction stage, the electromagnetic valve action attraction data is extracted, and the key parameters of quantifying the attraction performance are extracted from the data segment of the attraction stage, solving the problem of one-sidedness of traditional parameter extraction and inability to reflect the multi-dimensional characteristics of the attraction process. First, the total attraction time is obtained according to the time stamp of the attraction stage, that is, the total time from the start of power supply to the complete opening of the valve core, reflecting the overall response speed of the attraction process. Second, for the key sub-stage of the valve core movement in the attraction stage, the attraction current change rate is obtained through linear fitting algorithm, reflecting the conversion efficiency of electric energy to mechanical energy. Finally, the total energy consumption of the attraction process is quantified by combining the coil resistance and the current time sequence of the attraction stage. The total attraction time, attraction current change rate and attraction energy jointly depict the attraction performance from the response speed, energy conversion efficiency and overall energy consumption. For example, when the attraction time is too long and the current change rate is low, the energy is high, it is more likely to be valve core jam rather than coil excitation deficiency. Finally, these parameters are integrated into structured electromagnetic valve action attraction data, associated with module number and cycle number, providing multi-dimensional and associated attraction features for fault diagnosis.

[0041] Based on the data segment of the action release stage, the electromagnetic valve action release data is extracted, and the characteristic parameters reflecting the release performance are extracted from the data segment of the release stage, solving the problem of roughness of traditional release process parameter acquisition and inability to distinguish between circuit and mechanical faults. First, the total release time is obtained according to the time stamp of the release stage, reflecting the actual reset time of the valve core from power-off to complete closing; second, the release current change rate is obtained by least square fitting of the current curve in the release stage, reflecting the freewheeling performance of the release circuit; finally, the release peak current is extracted from the current sequence in the release stage by peak detection algorithm, reflecting the current fluctuation characteristics in the release process. These three parameters cooperatively reflect the circuit and mechanical performance of the release stage: the total release time reflects the mechanical reset speed, the current change rate reflects the circuit freewheeling efficiency, and the peak current reflects the circuit stability. For example, when the release current change rate is low and the peak current is abnormally high, it is more likely to be an accelerated release circuit fault rather than a spring problem.

[0042] The solenoid action touch current data is extracted based on a touch point related data window, and key parameters that characterize the starting performance are extracted from the touch point related data window, so as to solve the problem of the traditional neglect of the correlation of the touch process parameters and the inability to judge the matching of the magnetic force and the spring force. First, the current value at the touch point time is extracted as the touch current, which reflects the critical magnetic force for driving the valve core to start. Second, the touch delay time, that is, the time consumed from power-on to the start of the valve core, is obtained, which reflects the speed of the coil generating sufficient magnetic force. Finally, the touch current and the steady-state current ratio are obtained by combining the average current in the steady-state maintenance stage, which reflects the matching degree of the magnetic force and the spring pre-tightening force. The three parameters characterize the touch performance from three dimensions of starting threshold, time and matching degree: the touch current is the minimum magnetic force threshold for starting, the delay time reflects the speed of the magnetic force establishment, and the ratio reflects the balance state of the magnetic force and the spring force. For example, when the touch current and the steady-state current ratio are abnormal and the delay time is long, it may be caused by the over-tightening of the spring or the aging of the coil.

[0043] In one embodiment, the fault diagnosis module is configured to associate the solenoid action attraction data, the solenoid action release data and the solenoid action touch current data through the action cycle timestamp to generate single-cycle characteristic parameters, obtain preliminary abnormal marker parameters according to the single-cycle characteristic parameters, obtain composite fault determination information according to the preliminary abnormal marker parameters, compare the composite fault determination information with a matching fault characteristic library and divide the composite fault determination information into grades to obtain fault grade-type information, and obtain solenoid valve judgment information according to the fault grade-type information.

[0044] As described above, the three groups of data are associated to generate single-cycle characteristic parameters, the solenoid action attraction data, the release data and the touch current data are associated through the action cycle timestamp to generate characteristic parameters reflecting the complete performance of a single cycle, and the one-sidedness problem caused by the traditional isolated data analysis is solved. Specifically, first, three groups of data (including attraction total time, release current change rate, touch current and the like) are imported, and the action cycle identifier and the timestamp of each group of data are extracted; the time axis is aligned based on the power-on start timestamp, the deviation is corrected, and the logical relationship that the release start time is later than the attraction end time is verified, and the timing error is marked; the correlation characteristics (such as time proportion, slope ratio and energy-current ratio) are obtained based on the aligned parameters, which reflect the internal relationship between the parameters; the abnormality is marked through the verification of the physical logic of the parameters; finally, the basic parameters, the correlation characteristics and the abnormality markers are integrated to form a single-cycle characteristic parameter set. This process converts the dispersed parameters into a characteristic set from the perspective of the whole cycle, avoiding missing potential problems due to isolated analysis.

[0045] According to the single cycle characteristic parameter, the preliminary abnormal marker parameter is acquired, the abnormal parameter is marked through the preset threshold and the correlation analysis, the problem of traditional threshold rigidity and isolated parameter misjudgment is solved, and the basic abnormal characteristics are provided for fault judgment. First, the single cycle characteristic parameter is divided into time type, current / energy type and correlation characteristic type; the threshold library classified by model is called, the measured value of the parameter is compared with the threshold one by one, the single abnormality is marked; the correlation abnormality marker is superimposed based on the parameter correlation; the grade is divided according to the deviation rate and the number of abnormalities, and the serious abnormality needs immediate secondary verification. Through the classification verification and the correlation analysis, the abnormality is accurately marked, the accidental fluctuation misjudgment is reduced, and the reliability of the abnormality marking is ensured.

[0046] According to the preliminary abnormal marker parameter, the compound fault judgment information is acquired, the fault root is located by analyzing the abnormal marker combination through the correlation rule, and the problem that the traditional single parameter cannot distinguish the fault type is solved. First, the correlation rule library based on the fault mechanism and historical data is constructed; the preliminary abnormal marker parameter is matched with the rule, for example, the above-mentioned combined marker appears for 3 consecutive cycles, and it is judged as “valve core sticking risk”; for the matched fault type, the triggering parameter and the deviation degree are recorded to acquire the confidence. The abnormal combined marker that does not match the rule is marked as “unidentified fault type”, and is included in the manual analysis. This process traces back from “parameter abnormal phenomenon” to “fault nature”, for example, whether the long attraction time is caused by valve core sticking or coil excitation deficiency, and the accuracy of fault positioning is improved.

[0047] The fault characteristic library is matched to divide the grade, the fault grade-type information is obtained, the traditional fault response without priority is solved, and hierarchical maintenance is realized. First, the characteristic library containing the fault influence degree and development speed is constructed (for example, the valve core sticking needs immediate shutdown for the first-class emergency fault, the coil aging needs limited maintenance for the second-class early warning fault, and the transient fluctuation can be continuously monitored for the third-class slight fault); the compound fault judgment information is compared with the characteristic library, for example, the “valve core sticking risk” matches the first-class fault, and the “coil aging trend” matches the second-class fault; the grade is dynamically adjusted in combination with the fault development trend and historical data. Finally, the structured information containing the fault type, the grade and the confidence is output, the maintenance resources are reasonably distributed, and the waste of resources or fault omission is avoided.

[0048] According to the fault grade-type information, the electromagnetic valve judgment information is acquired, the fault details, the processing suggestions and the trend prediction are integrated, the executable judgment information is generated, and the problem that the traditional diagnosis result is not convenient for landing is solved. First, the key parameters triggering the fault and the historical trend are extracted; the targeted processing suggestion is generated based on the grade and the type; the fault development and the system influence are predicted; the structured information containing the fault type, the grade, the triggering parameter, the suggestion and the trend is integrated by comparing and verifying with the historical fault.

[0049] In one embodiment, the fault diagnosis module is further configured to extract the action period identifier and the action period timestamp corresponding to the solenoid valve action suction data, the solenoid valve action release data, and the solenoid valve action touch current data, align the action period identifier and the action period timestamp corresponding to the solenoid valve action suction data, the solenoid valve action release data, and the solenoid valve action touch current data to a sorting time axis, obtain a single period time range according to the sorting time axis, obtain a correlation characteristic parameter of the solenoid valve action suction data, the solenoid valve action release data, and the solenoid valve action touch current data according to the single period time range, obtain an abnormal correlation marker parameter based on the correlation characteristic parameter, and generate a single period characteristic parameter according to the correlation characteristic parameter and the abnormal correlation marker parameter.

[0050] As described above, the action period identifier and the action period timestamp are extracted, and the unique identifier and the key time point are extracted from the solenoid valve action suction data, the release data, and the touch current data, which lays a foundation for subsequent data correlation. Specifically, the action period identifier is extracted from the meta information of the three groups of data, ensuring that the three groups of data of the same period can be accurately matched; at the same time, the key time stamps of each process are extracted, including the suction start and end time, the release start and end time, and the touch time, etc. These time stamps are the core coordinates reflecting the action timing. By checking the consistency of the time stamps under the same period identifier, the period matching error is marked and corrected, and the confusion of data of different periods is avoided.

[0051] The time stamps are aligned to the sorting time axis, the time deviation of the multi-source data is corrected by constructing a unified time axis, and the timing accuracy of the parameter correlation is ensured. First, the sorting time axis is constructed with the power-on start timestamp as the reference, and the deviation of the three groups of data timestamps from the reference is obtained; if the deviation is small, the timestamp is directly mapped to the unified axis; if the deviation is large, the time sequence is corrected by linear interpolation to ensure the time continuity. After correction, the basic timing logic is verified, and the timing contradiction is marked and processed. The problem of time asynchronization caused by traditional data due to sensor delay and transmission time difference is solved, and the time deviation in the same period is controlled in a very small range.

[0052] The single period time range is obtained according to the sorting time axis, the time boundary of the single period is clearly defined, and the confusion of adjacent period data is avoided. The power-on start time is taken as the starting point of the period, and the release end time is taken as the end point of the period to form the single period time range, and the period duration is obtained as a reference parameter. By comparing with the standard period length of the same type of solenoid valve, the range validity is checked, and the period length abnormality is marked; at the same time, whether the adjacent periods overlap is checked to avoid boundary confusion.

[0053] According to the single cycle time range to obtain the associated characteristic parameters, the derived index is obtained to reflect the synergy of the attraction, release and triggering process, and the limitation of single parameter is made up. From the time correlation, current rate correlation and current-time correlation dimensions, the characteristics are extracted, which reflect the matching and rationality of multiple processes. The physical rationality of the characteristics is checked (such as whether the ratio is within the normal range), and the acquisition error is marked and corrected.

[0054] Based on the associated characteristic parameter acquisition abnormal association marker parameter, the threshold comparison and time sequence check are used to identify the association abnormality, and the systemic problem that cannot be detected by single parameter detection is captured. The association characteristic threshold library preset according to the model is called to mark the characteristics beyond the range; at the same time, the time sequence logic is checked to mark the contradiction that the triggering is later than the attraction and the release is before the attraction, and the abnormal level is divided according to the influence degree. This process solves the problem of isolation of traditional abnormal detection, such as through the "attraction energy consumption abnormality" marker, the implicit fault of valve core friction increase can be identified, and through the "time sequence error" marker, the data acquisition or circuit triggering problem can be found.

[0055] According to the associated characteristic parameters and the abnormal association marker parameters, the single cycle characteristic parameters are generated, the multi-dimensional information is integrated, and the structured single cycle characteristic parameter set is formed to provide "one-stop" data support for subsequent diagnosis. The integration content includes basic information, original parameters, associated characteristics, abnormal markers, and the data consistency is checked to ensure that there is no field missing or logical contradiction.

[0056] In one embodiment, the fault diagnosis module is further configured to associate an action cycle identifier and an action cycle timestamp according to a fault level and a fault type, obtain a basic information package, extract specific parameter measured values and standard values and deviation rates of a trigger parameter from the single cycle characteristic parameter set according to the basic information package, obtain trigger parameter details, obtain structured judgment information according to the trigger parameter details, and compare and check the structured judgment information with historical similar fault records to obtain electromagnetic valve judgment information.

[0057] As described above, the basic information package is obtained by associating the fault core attributes with the cycle space-time information according to the fault level and the fault type associated with the action cycle identifier and the timestamp, the precise positioning of the fault is realized, and the problem that the fault is disconnected with the occurrence scene in the traditional diagnosis is solved. Specifically, the "fault level-type information" and the "single cycle characteristic parameter set" generated in the early stage are called to establish the association according to the rules: the fault triggered by a single cycle is one-to-one bound with the cycle identifier and the timestamp, and the fault triggered by continuous cycles is associated with the first abnormal cycle and the number of continuous cycles. Finally, the basic information package containing the fault level, type, associated cycle identifier, starting timestamp and number of continuous cycles is formed, and the information integrity (such as whether the number of continuous cycles is missed) is checked.

[0058] According to the basic information package, trigger parameter details are extracted from the single-cycle characteristic parameter set, which focuses on quantifying the severity of the fault. By extracting the specific values of key parameters, the problem of traditional fault description being fuzzy and unquantifiable is solved. Based on the associated cycle identifier in the basic information package, the single-cycle characteristic parameter set of the corresponding cycle is called, and the trigger parameters (such as the pull-in time associated with the risk of spool sticking, the touch current ratio, etc., excluding irrelevant parameters) are selected according to the fault type. For the selected parameters, the measured values are extracted, the threshold library of the corresponding electromagnetic valve type is called to obtain the standard values, the deviation rate (reflecting the deviation between the measured value and the standard value) is obtained, the trigger parameter details including parameter name, measured value, standard value, deviation rate and unit are formed, and the rationality of the deviation rate is verified. This process converts the fault description from qualitative to quantitative, providing objective data support for subsequent fault classification and processing recommendations, so that faults of different severity can be responded differently.

[0059] According to the trigger parameter details, structured judgment information is obtained, which integrates multi-dimensional information to form standardized judgment information, solving the problem of unstructured traditional fault information and the inconvenience of subsequent processing. The integration content includes basic information (fault level, type, associated cycle identifier, timestamp), trigger parameter details (measured value, standard value, deviation rate), and processing recommendations are generated based on the fault level and type matching the preset recommendation library, supplemented by trend prediction based on continuous cycle parameter changes. Finally, it is stored in a standardized format, including core fields and verifying integrity.

[0060] The structured judgment information is compared and verified with historical similar fault records to obtain electromagnetic valve judgment information, which corrects abnormalities by comparing with historical data to improve the reliability of the judgment information, solving the problem of high misdiagnosis rate caused by the lack of verification in traditional diagnosis. From the fault log database, the near preset similar fault records (same type, same type electromagnetic valve, excluding misdiagnosis records) are called, and the comparison is made from the dimensions of parameter deviation rate, trend consistency, and processing recommendation matching degree: if the current deviation rate is significantly different from the historical average, mark "parameter abnormal fluctuation"; if the trend is inconsistent with the history, mark "trend inconsistency"; if the suggestion is not matched, it is automatically corrected. According to the verification result, the prediction is supplemented or adjusted, and finally the electromagnetic valve judgment information containing the verification state is formed and stored according to the "model-fault type-timestamp" index.

[0061] In one embodiment, the fault diagnosis module is also configured to obtain classification characteristic parameters according to the single-cycle characteristic parameters, call the threshold library of the corresponding electromagnetic valve type according to the classification characteristic parameters, match the normal range of each parameter to obtain the parameter normal threshold, determine the single-cycle characteristic parameters based on the parameter normal threshold to obtain the single-item abnormal mark, obtain the associated abnormal mark according to the single-item abnormal mark, and generate the abnormal level according to the single-item abnormal mark and the associated abnormal mark, and label the single-cycle characteristic parameters to obtain the abnormal mark parameters according to the abnormal level.

[0062] As described above, according to the single-cycle characteristic parameter acquisition classification characteristic parameter, the single-cycle characteristic parameter is classified, which lays a foundation for subsequent targeted matching threshold, and solves the problem of poor targeting of threshold matching caused by traditional parameter confusion. From the single-cycle characteristic parameter set, parameters such as suction time, release current change rate, and touch current proportion are extracted, which are divided into three categories according to the performance dimensions reflected: time, current / energy, and correlation characteristic. Through verification, the parameter classification is accurate, and finally the structured storage is classified according to the category, and each category of parameters is attached with physical units.

[0063] According to the classification characteristic parameter, the threshold library of the corresponding electromagnetic valve type is called, the parameter normal threshold is matched, and the parameter normal range is matched by calling the type-specific threshold library, which solves the problem of traditional general threshold ignoring the differences between electromagnetic valve types and ensures that the threshold fits the actual performance. Specifically, the electromagnetic valve type is extracted from the meta-information of the single-cycle characteristic parameter set, and the corresponding threshold library is called as an index. The threshold library is classified by type, dynamically updated based on manufacturer technical manuals and historical normal data statistics of the same type, and contains the normal range of each classification parameter: time class parameters such as the standard range of suction time, current class parameters such as the reasonable interval of touch current proportion, and correlation characteristic class parameters such as the minimum requirement of slope ratio. If the threshold of a certain type is missing, the threshold of the same series type is automatically called and marked as "reference value".

[0064] Based on the parameter normal threshold, the single-cycle characteristic parameter is determined, and the single-item abnormality mark is obtained, which marks the single parameter abnormality by comparing the parameter with the normal threshold, and solves the problem of potential problem missed detection caused by lack of clear abnormality identification in the traditional. Specifically, the classified characteristic parameters are compared with the matched normal threshold one by one: time class parameters such as suction time exceeding the upper limit or being lower than the lower limit, marked as "suction time over limit"; current class parameters such as release peak current exceeding the limited proportion of steady-state current, marked as "release peak current over limit"; correlation characteristic class parameters such as slope ratio being lower than the standard, marked as "release-suction efficiency ratio not up to standard". The measured value, threshold range and deviation direction are recorded for each abnormality, and slight fluctuations with very small deviation rate are filtered out to avoid excessive sensitivity. Finally, the structured single-item abnormality mark set is formed according to the parameter category.

[0065] According to the single abnormality label, the associated abnormality label is obtained, and the abnormality is cooperated by analyzing the association between parameters. The problem that the traditional isolated label cannot identify the complex fault is solved, and the specificity of abnormality detection is improved. Specifically, the association rule library is constructed based on the fault mechanism, for example, "suction time limit + touch delay ratio too high" corresponds to "suction stage timing abnormality", "release peak current limit + efficiency ratio not up to standard" corresponds to "release circuit efficiency abnormality", and three single abnormalities are labeled as "multi-parameter cooperative abnormality". The single abnormality label is compared with the rule library, the matching rule is triggered to generate the associated abnormality label, and the association strength is evaluated according to the number of parameters (for example, 2 parameters are matched as weak association, and 3 or more are matched as strong association). The associated abnormality label is bound and stored with the single abnormality, which is convenient for tracing the root cause.

[0066] According to the single abnormality label and the associated abnormality label, the abnormality level is generated, the level is divided according to the abnormality influence degree, the problem that the traditional unified processing of all abnormalities leads to unreasonable allocation of maintenance resources is solved, and the processing priority is clear. Specifically, the level division standard is: critical abnormality, significant abnormality, and serious abnormality. The level is determined by the maximum deviation rate of single abnormality and the strength of associated abnormality, for example, serious deviation or strong association directly determines serious abnormality, and only moderate deviation or weak association determines significant abnormality. Logical verification is performed on the level result to avoid contradictions (such as multi-parameter cooperative abnormality should not be determined as critical).

[0067] According to the abnormality level, the single cycle characteristic parameter is labeled to obtain the abnormality labeled parameter, all abnormality information is integrated to generate the structured labeled parameter, the problem that the traditional information fragmentation leads to low subsequent diagnosis efficiency is solved, and complete input is provided for fault determination. The abnormality labeled parameter includes basic information, classified characteristic parameter measured value, single and associated abnormality label, abnormality level and determination basis, adopts standardized format, and is stored according to "model-cycle" index, which is associated with the single cycle characteristic parameter set.

[0068] In one embodiment, the verification module is configured to obtain electromagnetic valve core verification elements based on the electromagnetic valve judgment information, wherein the electromagnetic valve core verification elements include electromagnetic valve key parameters, electromagnetic valve time dimension information, and electromagnetic valve state identification. Based on the electromagnetic valve key parameters, the electromagnetic valve time dimension information, and the electromagnetic valve state identification, the historical judgment information corresponding to the preset period is extracted and compared with the historical judgment information to obtain historical trend comparison result information. According to the historical trend comparison result information, the electromagnetic valve standard parameter database corresponding to the model is called, and the historical trend comparison result information is compared with the electromagnetic valve standard parameter database to obtain a database matching degree. The electromagnetic valve matching verification information is obtained based on the database matching degree.

[0069] As mentioned above, according to the electromagnetic valve judgment information, the key elements of the electromagnetic valve core verification are extracted from the electromagnetic valve judgment information, the verification core is focused, and the foundation for subsequent multi-dimensional comparison is laid, solving the problem of lack of pertinence caused by traditional verification due to complex information. From the judgment information output by the fault diagnosis module, three types of core elements are extracted: electromagnetic valve key parameters (such as triggering fault closing time, triggering current ratio and its deviation rate, reflecting the direct manifestation of the fault), electromagnetic valve time dimension information (such as the starting period of the fault, the number of continuous periods, and the timestamp range, reflecting the timing characteristics of the fault), and electromagnetic valve state identification (such as fault type, level, trend prediction, and clear fault attribute). Through verification, the elements are ensured to be complete (such as supplementing the missing continuous period number), and finally stored in a structured manner according to the "key parameter-time dimension-state identification" structure.

[0070] Based on the core verification elements, historical judgment information is extracted and compared, and historical trend comparison result information is obtained. By comparing and analyzing the parameter change trend with historical data, the difference between occasional abnormality and real fault is distinguished, solving the problem of easy misjudgment in traditional methods that only rely on single-period data. Based on the model, module identification and state identification in the core verification elements, the same type of historical judgment information in the near preset period is extracted from the database, and the comparison is made from three dimensions: parameter change trend, fault persistence, and correlation parameter consistency. Quantitative comparison results such as "trend consistency", "duration score" and "correlation matching degree" are formed into a structured report. For example, if the "closing time is out of limits" and the same type of abnormality occurs in the last 3 periods, the trend is consistent and the correlation parameters match, it is determined as a persistent fault; if only single-period abnormality and historical parameters are normal, it is marked as a suspected occasional interference.

[0071] According to the historical trend comparison result information, the electromagnetic valve standard parameter database of the corresponding model is called, and the database matching degree is obtained by comparison. By calling the model-specific standard database, it is verified whether the current fault conforms to the typical characteristics of the model, solving the problem of poor adaptability of traditional general standards. The standard database is classified by model and contains typical fault types, characteristic parameter ranges, and associated abnormality combinations, and is dynamically updated to optimize accuracy. Based on the model in the core elements, the corresponding database is accurately called, and the matching degree is obtained from three aspects: whether the fault type is in the typical library of the model, whether the current parameter deviation falls within the characteristic range, and whether the associated abnormality combination is consistent with the typical combination. For example, the "valve core sticking" of the electromagnetic valve is usually accompanied by double abnormalities of closing time and triggering current ratio. If the current fault conforms to this feature, the matching degree is high; if a parameter deviation exceeds the typical range of the model, the matching degree is reduced.

[0072] According to the database matching degree, electromagnetic valve matching verification information is obtained, the reliability of the quantitative diagnosis result is based on the matching degree, a targeted suggestion is generated, a decision basis is provided for the early warning module, and the problem of no clear reliability index of the traditional verification result is solved. According to the matching degree, the reliability is divided into three levels: high, medium and low. High reliability suggestion directly triggers early warning; medium reliability suggestion secondary sampling verification; low reliability suggestion to investigate the interference source. The reliability rating, matching degree, historical trend conclusion and correction suggestion are integrated to form standardized verification information. For example, high reliability "valve core sticking risk" directly triggers early warning; low reliability abnormality suggests reviewing the sensor.

[0073] In one embodiment, the early warning module is configured to obtain early warning trigger element information according to the electromagnetic valve verification information, wherein the early warning trigger element information includes a fault type and an original judgment level, generate a preliminary early warning level according to the fault type and the original judgment level, generate a differentiated structured content according to the preliminary early warning level, and obtain early warning information according to the differentiated structured content.

[0074] As described above, according to the electromagnetic valve verification information, early warning trigger element information is obtained, the key elements that determine the early warning response are extracted from the electromagnetic valve verification information, accurate basis is provided for subsequent level division and content generation, and the problem of insufficient targeting caused by fuzzy elements in traditional early warning is solved. From the matching verification information output by the verification module, three types of core elements are extracted: fault type, original judgment level, and reliability rating. Through verification, the element logic is ensured to be consistent, and if there is a contradiction, the information with the highest reliability is used as the reference, for example, a structured trigger element of "fault type: XX; original judgment level: XX; reliability: XX" is formed.

[0075] According to the fault type and the original judgment level, a preliminary early warning level is generated, a preliminary early warning level matched with the actual risk is generated through a dynamic adjustment mechanism, and the problem of rigid traditional early warning level and disconnection with the actual risk is solved. Based on the constructed level adjustment rule library, the original judgment level is taken as the reference, the risk weight of the fault type and the reliability of the verification information are combined for adjustment: even if the original level is low, the high risk type will not be downgraded or even upgraded under high reliability; the original judgment level is maintained for the medium risk type, and the early warning level is downgraded if the reliability is low; the early warning level is appropriately downgraded for the low risk type combined with the reliability. For example, "valve core sticking risk" (high risk) + original level one + high reliability → preliminary level "one"; "coil aging trend" (medium risk) + original level two + low reliability → preliminary level "three (need to review)". At the same time, the level logic is verified to ensure that there is no invalid level.

[0076] According to the preliminary warning level, generate differentiated structured content, customize differentiated information content for different warning levels, solve the problem of traditional warning content homogeneity and key information missing, and ensure that maintenance personnel quickly obtain action guidelines. Based on the preliminary warning level, design three types of template library: first-level warning highlights emergency operation, impact range and real-time data link; second-level warning specifies maintenance time limit, detection project and trend prediction; third-level warning only needs to record abnormal parameters and monitoring requirements (such as "review every 10 cycles"). Extract data to fill in templates from trigger elements and verification information, for example, first-level warning automatically associates real-time monitoring screen, second-level warning inserts maintenance manual link. Structured content is presented in "title + points" to ensure that key information is clear at a glance.

[0077] According to the differentiated structured content, obtain warning information, generate executable warning information through multi-channel push and closed-loop tracking, solve the problem of traditional warning transmission lag and no follow-up tracking. According to the preliminary warning level, match the push channel: first-level warning synchronously triggers local sound and light alarm, remote pop-up window, mobile phone message and other multi-channels, ensuring that emergency failures are received in time; second-level warning is pushed to the operation and maintenance APP and monitoring board, taking into account timeliness and planning; third-level warning is only recorded in the log to avoid information redundancy. At the same time, add a unique ID to the warning information, associate it with the operation and maintenance work order system, track the processing status, and automatically upgrade the push if the first-level warning is not processed within a certain time limit, and archive the results after processing is completed. For example, first-level warning is reminded through "red strobe + message + pop-up window" multi-dimensional, with emergency operation steps; third-level warning is only reflected in the weekly report.

[0078] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiment methods. Any reference to memory, storage, value library or other medium provided by the present application and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0079] It should be noted that in this document, the terms "comprising", "including", or any other variant thereof are intended to cover a non-exclusive inclusion, so that a process, device, article or method that includes a list of elements not only includes those elements, but also includes other elements not expressly listed or inherent to such process, device, article or method. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, device, article or method that includes the element.

[0080] The above only describes the preferred embodiments of the present application, and does not limit the patent scope of the present application. Any equivalent result or equivalent process transformation obtained by using the content of the present application specification and drawings, or direct or indirect application in other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A multi-module integrated electromagnetic valve working state detection system applied to multi-module integrated electromagnetic valve test, characterized in that, The application relates to a multi-module integrated electromagnetic valve data acquisition system. The application comprises the following modules: An electromagnetic valve data acquisition module, a data processing module, a dynamic extraction module, a fault diagnosis module, a verification module and a pre-warning module. The electromagnetic valve data acquisition module is used for acquiring action detection data sets of multi-module integrated electromagnetic valves. The data processing module is used for acquiring electromagnetic valve action characteristic data corresponding to each module integrated electromagnetic valve according to the action detection data sets. The dynamic extraction module is used for acquiring electromagnetic valve action attraction data, electromagnetic valve action release data and electromagnetic valve action touch current data according to each electromagnetic valve action characteristic data. The fault diagnosis module is used for acquiring electromagnetic valve judgment information according to the electromagnetic valve action attraction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data. The verification module is used for generating electromagnetic valve verification information according to the electromagnetic valve judgment information.

2. The multi-module integrated solenoid valve operating state detection system according to claim 1, characterized by The pre-warning module is used for generating pre-warning information according to the electromagnetic valve verification information.

3. The multi-module integrated solenoid valve operating state detection system according to claim 1, characterized by The electromagnetic valve data acquisition module acquires first current signals and power-on initial current data of a pre-attraction stage of each module integrated electromagnetic valve, second current signals and valve core attraction time data of a valve core opening movement stage, third current signals and electromagnetic valve power-on time of a steady-state keeping open stage and fourth current signals and steady-state current data of a power-off acceleration release stage, generates multi-stage signal integrated information of each module integrated electromagnetic valve according to the first current signals, the second current signals, the third current signals and the fourth current signals, generates action characteristic data according to the power-on initial current data, the steady-state keeping open stage, the electromagnetic valve power-on time and the steady-state current data, and integrates the multi-stage signal integrated information and the action characteristic data to obtain the action detection data set of each module integrated electromagnetic valve. The data processing module acquires action structured data through time stamp synchronization multi-channel collection of the action detection data set, wherein the action structured data comprises action division time stamps, action stage current values and action stage labels, acquires current feature point positioning-stage division information according to the action division time stamps, the action stage current values and the action stage labels, generates a current-stage curve diagram according to the current feature point positioning-stage division information, acquires dynamic stage characteristic data sets of the power-on pre-attraction stage, the valve core opening movement stage, the steady-state keeping open stage and the power-off acceleration release stage according to the current-stage curve diagram, and acquires electromagnetic valve action characteristic data according to the dynamic stage characteristic data sets.

4. The multi-module integrated solenoid valve operating condition detection system according to claim 1, characterized by The dynamic extraction module is configured to acquire energization attraction stage data segment, action release stage data segment and touch point related data window according to the electromagnetic valve action characteristic data, acquire attraction total time and attraction current change rate according to the attraction stage data segment, acquire attraction energy data according to the attraction total time and attraction current change rate, and take the attraction energy data as electromagnetic valve action attraction data, acquire release total time and release current change rate according to the action release stage data segment, acquire release peak current data according to the release total time and release current change rate, and take the release peak current data as electromagnetic valve action release data, acquire touch current and touch delay time according to the touch point related data window, acquire touch current to steady current ratio according to the touch current and touch delay time, and take the touch current to steady current ratio as electromagnetic valve action touch current data.

5. The multi-module integrated solenoid valve operating condition detection system according to claim 1, characterized by, The fault diagnosis module is configured to associate the three groups of data by action cycle timestamp to generate single cycle characteristic parameter according to the electromagnetic valve action attraction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data, acquire preliminary abnormality marking parameter according to the single cycle characteristic parameter, acquire composite fault determination information according to the preliminary abnormality marking parameter, compare the composite fault determination information with a matching fault characteristic library and divide the composite fault determination information into grades to obtain fault grade-type information, and acquire electromagnetic valve judgment information according to the fault grade-type information.

6. The multi-module integrated solenoid valve operating condition detection system according to claim 5, characterized by The fault diagnosis module is further configured to extract action cycle identifier and action cycle timestamp corresponding to the electromagnetic valve action attraction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data, align the action cycle identifier and the action cycle timestamp corresponding to the electromagnetic valve action attraction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data to a sorting time axis, acquire single cycle time range according to the sorting time axis, acquire correlation characteristic parameter of the electromagnetic valve action attraction data, the electromagnetic valve action release data and the electromagnetic valve action touch current data according to the single cycle time range, acquire abnormality correlation marking parameter based on the correlation characteristic parameter, and generate single cycle characteristic parameter according to the correlation characteristic parameter and the abnormality correlation marking parameter.

7. The multi-module integrated solenoid valve operating condition detection system according to claim 5, characterized by The fault diagnosis module is further configured to acquire basic information package according to fault grade and fault type correlation action cycle identifier and action cycle timestamp, acquire trigger parameter details according to the basic information package, extract specific parameter measured value and standard value and deviation rate of triggering the fault from a single cycle characteristic parameter set, acquire structured judgment information according to the trigger parameter details, and obtain electromagnetic valve judgment information by comparing and verifying the structured judgment information with historical similar fault records.

8. The multi-module integrated solenoid valve operating condition detection system according to claim 5, characterized by The fault diagnosis module is further configured to acquire a classification characteristic parameter according to the single-cycle characteristic parameter, call a threshold library of a corresponding electromagnetic valve model according to the classification characteristic parameter, match a normal range of each parameter to obtain a parameter normal threshold, determine a single-item abnormal mark based on the parameter normal threshold and the single-cycle characteristic parameter, acquire a correlation abnormal mark according to the single-item abnormal mark, generate an abnormal level according to the single-item abnormal mark and the correlation abnormal mark, and label the single-cycle characteristic parameter according to the abnormal level to obtain an abnormal mark parameter.

9. The multi-module integrated solenoid valve operating condition detection system according to claim 1, characterized by, The verification module is configured to acquire an electromagnetic valve core verification element according to the electromagnetic valve judgment information, wherein the electromagnetic valve core verification element includes an electromagnetic valve key parameter, electromagnetic valve time dimension information and an electromagnetic valve state identifier, extract historical judgment information of a corresponding preset period based on the electromagnetic valve key parameter, the electromagnetic valve time dimension information and the electromagnetic valve state identifier, and compare the historical judgment information to obtain historical trend comparison result information, call an electromagnetic valve standard parameter database of a corresponding model according to the historical trend comparison result information, and compare the historical trend comparison result information with the electromagnetic valve standard parameter database to obtain a database matching degree, and acquire electromagnetic valve matching verification information according to the database matching degree.

10. The multi-module integrated solenoid valve operating condition detection system according to claim 1, characterized by, The early warning module is configured to acquire early warning trigger element information according to the electromagnetic valve verification information, wherein the early warning trigger element information includes a fault type and an original judgment level, generate a preliminary early warning level according to the fault type and the original judgment level, generate a differential structured content according to the preliminary early warning level, and acquire early warning information according to the differential structured content.

Images