A method and system for structured management and operational status analysis of electrical detection data

Abstract

The application discloses a kind of structured management and operating state analysis method and system of electrical detection data, it is related to power equipment operating state analysis technical field, comprising: based on time irreversible anchoring point is segmented to equipment operating time axis, directional drift analysis is executed to the state parameter in each segmented time interval, and state transition association is established, generates discrete state sequence;And establish state inertia retention relationship, generate continuous inertia state trajectory;Establish defect state inversion relationship to generate state disturbance characteristics, and correct continuous inertia state trajectory, obtain defect correction state trajectory;Deviation calibration is carried out using new observation data and defect correction state trajectory, inertia retention and defect inversion relationship are updated simultaneously, generate continuous deterioration state sequence covering whole operating cycle;The application solves the problem that electrical detection data in the prior art relies on discrete sampling, leading to operating state expression fracture, difficult to truly depict the continuous deterioration evolution process of equipment.

Description

Technical Field

[0001] This invention relates to the field of power equipment operation status analysis technology, and more specifically, to a structured management and operation status analysis method and system for electrical testing data. Background Technology

[0002] Electrical equipment (such as transformers, circuit breakers, cables, and GIS equipment) operates long-term in power systems, and their condition assessment typically relies on periodic test data, online monitoring data, and defect record data. Current technologies mostly determine the health status of equipment by performing time-sequencing or statistical analysis on the aforementioned data.

[0003] However, existing methods mainly construct state sequences based on discrete sampling points, forming state results only at the time of testing or monitoring. They lack the ability to depict the continuous changes in equipment state within adjacent time intervals, resulting in significant discontinuities in the expression of operating state.

[0004] Meanwhile, defect information is usually recorded using the "discovery time" rather than the actual deterioration start time, which causes a lag between the defect occurrence time and the actual deterioration process of the equipment, making it difficult to accurately recreate the equipment's state evolution process.

[0005] Therefore, existing technologies lack mechanisms for maintaining and reconstructing the state of electrical equipment during continuous degradation, resulting in a discontinuity in the operating state over time, which fails to accurately reflect the true degradation and evolution process of the equipment. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a structured management and operational status analysis method and system for electrical testing data, which solves the problems in the prior art where electrical testing data relies on discrete sampling, resulting in fragmented expression of operational status and difficulty in truly depicting the continuous deterioration and evolution process of equipment.

[0007] To achieve the above objectives, the present invention provides the following technical solution: Firstly, this application provides a structured management and operational status analysis method for electrical testing data. This method includes: segmenting the equipment's operating time axis based on irreversible time anchor points; performing directional drift analysis on the state parameters within each segmented time interval; establishing state transition correlations; generating discrete state sequences; establishing state inertia maintenance relationships based on the discrete state sequences; generating continuous inertial state trajectories; establishing defect state inversion relationships based on defect data to generate state disturbance characteristics; correcting the continuous inertial state trajectories; and performing deviation calibration using new observation data and the defect-corrected state trajectories, synchronously updating the inertia maintenance and defect inversion relationships, and generating a continuous deterioration state sequence covering the entire operating cycle.

[0008] In one embodiment, generating a discrete state sequence includes: identifying abnormal states that satisfy irreversible state conditions based on a standard state parameter set, generating a set of time-irreversible anchor points; segmenting the equipment operating time axis using the time-irreversible anchor points, and performing unidirectional state drift analysis to generate a continuous state trajectory in intervals; extracting interval state features and establishing state transition mapping relationships with anchor points, and performing forced transition connections on adjacent intervals to generate a segmented state evolution chain for the equipment; The continuous state trajectories of each interval are sequentially associated with the state evolution chain to generate a discrete state sequence of the device.

[0009] In one embodiment, establishing a state inertia maintenance relationship based on a discrete state sequence includes: acquiring a discrete state sequence of the device, establishing a state inertia analysis sequence in chronological order, and extracting state inertia feature parameters between adjacent discrete states; based on the state inertia feature parameters, identifying state intervals that continuously maintain the same direction of change, and generating a set of inertia state intervals; performing trend continuity analysis on each inertia state interval to generate a state inertia continuity trajectory; and for time intervals without new observation data, performing non-observation state compensation processing based on the state inertia continuity trajectory, and applying continuity constraints to establish a state inertia maintenance relationship.

[0010] In one embodiment, generating a continuous inertial state trajectory includes: generating a state inertial evolution baseline sequence based on the state inertia maintenance relationship; aligning and fusing running data and defect data according to a time window to generate a set of candidate options for reverse trend triggering; performing weighted fusion on deviation indicators in the candidate option set to obtain a reverse trend triggering intensity parameter, and determining whether a local state reversal mechanism is triggered accordingly; performing reverse trend state correction when triggered, fusing it with the non-triggered portion, and then performing continuity verification and smoothing processing to generate a continuous inertial state trajectory.

[0011] In one embodiment, establishing a defect state inversion relationship based on defect data to generate state disturbance features includes: performing time-series processing on the defect data to generate a defect event sequence; constructing an initial feasible state domain before the defect occurs, and introducing a defect constraint function to constrain and truncate the initial feasible state domain to generate a first-stage feasible state domain; performing backpropagation processing on the defect constraint conditions in reverse time order according to the occurrence time of the defect events, recursively updating the historical state space of the equipment, and generating a multi-stage feasible state domain sequence; calculating the state domain shrinkage rate based on the multi-stage feasible state domain sequence, and centering the feasible state domain at each time node to generate a state disturbance reference vector; and performing component decomposition on different defect types according to the state disturbance reference vector to generate a state disturbance feature vector.

[0012] In one embodiment, correcting the continuous inertial state trajectory to obtain a defect-corrected state trajectory includes: analyzing the disturbance intensity of the continuous inertial state trajectory based on the state disturbance feature vector, and constructing a state disturbance driving function field; mapping the continuous inertial state trajectory to the observation state space, and introducing observation consistency constraints to calculate the deviation sequence between the inertial state and the observation state; constructing a dual-constraint state projection optimization model based on the deviation sequence and disturbance intensity, performing projection correction processing on the inertial state trajectory, and performing enhancement correction or smooth preservation according to the disturbance driving intensity to obtain the defect-corrected state trajectory.

[0013] In one embodiment, deviation calibration is performed using new observation data and the defect-corrected state trajectory, and the inertial preservation and defect inversion relationships are updated synchronously. This includes: acquiring new observation data and generating an observation state sequence, aligning it with the defect-corrected state trajectory in time, and calculating a state deviation calibration sequence; performing adaptive feedback correction on the defect-corrected state trajectory based on the state deviation calibration sequence to generate an updated state-corrected trajectory; and updating the state inertial preservation relationship and the defect state inversion relationship based on the updated state-corrected trajectory, and performing joint reconstruction to generate an updated continuous state trajectory of the equipment.

[0014] In one embodiment, adaptive feedback correction is performed on the defect correction state trajectory based on the state deviation calibration sequence to generate an updated state correction trajectory. This includes: constructing an original error propagation sequence based on the state deviation calibration sequence, calculating the error propagation distance and establishing an error propagation damping kernel function to generate a damped error propagation sequence; normalizing the error propagation influence intensity in the damped error propagation sequence to generate an error propagation influence weight sequence; performing damped feedback correction on the defect correction state trajectory based on the error propagation influence weight sequence to generate an initial feedback correction trajectory; performing continuity verification on the initial feedback correction trajectory and performing local smooth transition processing on state windows that do not meet continuity constraints to generate a calibrated state correction trajectory.

[0015] In one embodiment, generating a continuous degradation state sequence covering the entire operating cycle includes: extracting a continuous state evolution feature sequence based on the calibrated continuous state trajectory of the equipment; dividing the continuous degradation stage based on the continuous state evolution feature sequence according to the state change gradient and the cumulative state offset; performing trend fitting on the state evolution features within each continuous degradation stage to generate a stage state evolution function, and smoothing the state connection boundary between adjacent stages to generate a full-cycle continuous state connection chain; calculating the cumulative state degradation index based on the full-cycle continuous state connection chain, and performing level mapping to generate a continuous degradation state label sequence; and temporally associating the label sequence with the connection chain to generate a continuous degradation state sequence covering the entire operating cycle of the equipment.

[0016] Secondly, this application provides a structured management and operational status analysis system for electrical testing data. This system includes: a discrete state sequence generation module, used to segment the equipment's operating time axis based on irreversible time anchor points, perform directional drift analysis on state parameters within each segmented time interval, establish state transition correlations, and generate a discrete state sequence; an inertia-maintaining evolution module, used to establish state inertia-maintaining relationships based on the discrete state sequence and generate a continuous inertia-maintaining state trajectory; a defect disturbance correction module, used to establish defect state inversion relationships based on defect data to generate state disturbance characteristics, and correct the continuous inertia-maintaining state trajectory to obtain a defect-corrected state trajectory; and a dynamic calibration and full-cycle degradation module, used to perform deviation calibration using new observation data and the defect-corrected state trajectory, synchronously update the inertia-maintaining and defect inversion relationships, and generate a continuous degradation state sequence covering the entire operating cycle.

[0017] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: By introducing irreversible time anchor points to physically constrain the segmentation of the equipment's operating time axis, the pseudo-continuity of the state caused by traditional uniform time slicing is avoided. Simultaneously, the continuous evolution trend of equipment degradation is characterized by state inertia maintenance relationships, and a defect state inversion mechanism transforms discrete defect information into structured perturbation constraints on the state trajectory, enabling state modeling to possess both "inertia continuation capability" and "defect-driven correction capability." Furthermore, dynamic calibration with new observation data enables synchronous iterative updates of the inertial model and the defect inversion relationship, thus forming a closed-loop adaptive state evolution system. Compared to existing technologies, this scheme can more realistically depict the nonlinear and irreversible degradation process of electrical equipment under the coupled effects of multiple factors such as insulation aging, partial discharge, and thermal stress, effectively reducing state inversion drift and error accumulation, and improving the continuity, physical consistency, and predictive reliability of the entire lifecycle state sequence. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic flowchart of a method for structured management and operational status analysis of electrical testing data provided in an embodiment of this application.

[0020] Figure 2 This is a schematic diagram of a structured management and operational status analysis system for electrical testing data provided in an embodiment of this application.

[0021] Figure 3 This is a schematic diagram showing the comparison of the state continuity of the missing observation interval provided in the embodiments of this application.

[0022] Figure 4 This is a comparison diagram of uncorrected and double-constraint state projection correction provided in the embodiments of this application. Detailed Implementation

[0023] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] Reference Figure 1 As shown in the diagram, the present invention provides a method for structured management and operational status analysis of electrical testing data, comprising the following steps: S1. Collect raw data of the target electrical equipment at different time points. The raw data includes test data, operation data and defect data. The equipment operation time axis is segmented based on the irreversible time anchor point. Oriented drift analysis is performed on the state parameters corresponding to each segment time interval. At the same time, the state transition correlation is established according to the state change amplitude between adjacent segment time intervals to generate the discrete state sequence of the equipment.

[0025] The test data includes insulation test data, partial discharge detection data, and condition-based maintenance detection data; the operating data includes operating condition parameters, load current, operating temperature, voltage fluctuations, and environmental parameters; and the defect data includes defect type, defect confirmation time, and defect handling records.

[0026] Preprocessing of the raw data includes: Based on the acquired raw data, time tags are extracted from data from different sources, and time alignment is performed according to a unified time benchmark to establish corresponding time mapping relationships and generate a multi-source time dataset for the device. The time alignment process is used to eliminate time discrepancies between different data sources, ensuring that different types of data correspond to the same time coordinate system. This time alignment process includes: aggregating and statistically analyzing continuous operational data according to a preset time window to generate operational status parameters for that time window; mapping experimental data to corresponding time windows according to the nearest-time matching principle; and mapping defect data to corresponding time windows according to the defect occurrence time or defect confirmation time. This time mapping relationship is used to form joint status data within the same time window. Based on the equipment's multi-source time dataset, test data, operational data, and defect data are standardized to generate a standard state parameter set. The data standardization process includes anomaly noise filtering, missing value completion, and parameter normalization to ensure that different types of data meet the requirements of a unified state representation. Missing value completion includes using the average of adjacent time windows or historical data of the same type for completion; anomaly noise filtering includes removing abnormal sampled values ​​that exceed the rated operating range of the equipment; and parameter normalization includes converting parameters of different dimensions into a unified state evaluation interval.

[0027] In this embodiment, generating a discrete state sequence of the device includes: S101, based on the standard state parameter set, calculates the magnitude, direction and duration of parameter changes between adjacent time windows, and generates corresponding state change parameters; The state change parameters include the parameter change rate, the cumulative change amount of the parameter, and the parameter continuous change period; The rate of change of the parameter is determined by the ratio of the difference in parameters between adjacent time windows to the time interval; The cumulative change in the parameter is obtained by summing the parameter change values ​​over a continuous time window. The period of continuous change of the parameter is used to characterize the length of time that the parameter continuously deviates from the stable state.

[0028] S102, Based on the state change parameters, the data that meets the preset mutation conditions are screened, the state mutation characteristics are extracted, and the data of continuous cumulative change are statistically analyzed to extract the damage accumulation characteristics and generate a candidate abnormal state set. The preset mutation conditions include the parameter change rate exceeding a preset multiple of the historical average change rate of the corresponding parameter, or the parameter change direction continuously changing in one direction within a preset time range. The damage accumulation characteristics include insulation aging accumulation, thermal stress accumulation, and discharge intensity accumulation.

[0029] S103, Based on the candidate abnormal state set, perform state backtracking analysis on the historical time interval corresponding to each candidate abnormal state, determine whether the corresponding abnormal state meets the irreversible state condition, and generate a set of time irreversible anchor points. The state backtracking analysis includes: Analyze whether the equipment status returns to the historical stable parameter range after the abnormal state occurs; After an abnormal state occurs, does the corresponding parameter continue to deviate from the historical stable parameter range? If the parameter corresponding to the abnormal state continues to deviate from the range of historical stable parameters for more than a preset duration threshold, then the corresponding abnormal state is determined to meet the conditions for an irreversible state. The irreversible time anchor points include insulation degradation critical points, partial discharge transition points, thermal aging abrupt change points, and the first confirmation point of defects.

[0030] S104, Based on the generated set of irreversible time anchor points, the equipment operation time axis is segmented according to the chronological order, and the time interval between adjacent irreversible time anchor points is divided into the corresponding quasi-steady state interval. Each quasi-stable state interval corresponds to a continuous slowly changing state stage. The quasi-stable state interval is used to characterize the time range in which the device state is in a continuous slowly changing process. The quasi-stable state interval does not contain any new irreversible time anchor points.

[0031] S105, based on the quasi-stable state interval, performs unidirectional state drift analysis on the standard state parameter set corresponding to each quasi-stable state interval to generate a continuous state trajectory within the interval. The unidirectional state drift analysis includes: Determine the dominant degradation parameter corresponding to the current quasi-stable state interval; Establish the state drift direction according to the direction of change of the dominant degradation parameter; The state parameters corresponding to subsequent time windows are constrained to change continuously along the state drift direction; When subsequent state parameters change in the opposite direction to the state drift, a smoothing correction is performed on the reverse-changing parameters. The smoothing correction process includes replacing the reverse-variable parameter with the mean of adjacent time windows.

[0032] S106. Based on the continuous state trajectory of the interval, extract the interval state features corresponding to each quasi-stable state interval, and establish the state transition mapping relationship between each interval state feature and the corresponding irreversible anchor point in time, and generate the anchor point state transition model. The interval state characteristics include the interval average degradation rate, the interval cumulative damage value, and the interval state drift direction. The state transition mapping relationship is used to characterize the magnitude and direction of state change when the device state crosses the irreversible anchor point in time.

[0033] S107, Based on the generated anchor point state transition model, force transition connection is performed on the state changes between adjacent quasi-stable state intervals to generate a segmented state evolution chain of the equipment. The forced transition connection includes: The initial state of the subsequent quasi-stable state interval is adjusted according to the magnitude of the state change corresponding to the irreversible anchor point in time. Establish the starting point for the state evolution of the subsequent quasi-stable state interval based on the state transition direction corresponding to the irreversible anchor point in time; The device segmented state evolution chain is used to characterize the irreversible evolution relationship between device states in different quasi-stable state intervals.

[0034] S108, based on the segmented state evolution chain of the equipment, performs time-series correlation on the continuous state trajectory within each quasi-stable state interval and the state transition results corresponding to each irreversible time anchor point, and generates a discrete state sequence of the equipment.

[0035] The discrete state sequence of the equipment includes normal state, slowly deteriorating state, abruptly deteriorating state, and severely deteriorating state. The discrete state sequence of the equipment is used to characterize the phased evolution process of the target electrical equipment from a stable state to a continuously deteriorating state under the action of irreversible state events.

[0036] It should be noted that by using "time-irreversible anchor points" to segment and reconstruct the equipment's operating timeline, and combining this with unidirectional state drift analysis within the quasi-steady state interval to generate a discrete state sequence, the "pseudo-continuous state" problem caused by directly splicing state data based solely on time sequence in conventional techniques can be effectively avoided. This makes the generated state sequence more consistent with the actual physical degradation laws of electrical equipment. Simultaneously, by introducing an irreversible state identification mechanism, abrupt states generated by the equipment during insulation degradation, thermal aging, and partial discharge can be accurately captured, solving the problems of traditional methods failing to identify state abrupt boundary boundaries and distinguish between gradual and abrupt degradation. Furthermore, this scheme establishes cross-interval state transition relationships through an anchor point state transition model, avoiding state reversal fluctuations and spurious recovery phenomena caused by conventional interpolation methods. This solves the technical problems of distorted equipment state evolution paths, lack of real degradation constraints, and inability to accurately reflect the accumulation process of irreversible damage in existing technologies, improving the authenticity, continuity, and evolutionary reliability of the equipment state sequence.

[0037] S2 establishes a state inertia preservation relationship based on the discrete state sequence of the equipment, so that the equipment state continues to change according to the existing state trend when new observation data is lacking. Based on the state inertia preservation relationship, the continuous state evolution association processing is performed on the time interval with missing data in the discrete state sequence of the equipment to generate a continuous inertial state trajectory.

[0038] The state inertia preservation relation is used to constrain the device state to maintain continuous evolution during non-observation time intervals.

[0039] In this embodiment, establishing a state inertia preservation relationship based on the discrete state sequence of the device includes: S201, obtain the state parameters, state change direction and state duration corresponding to each discrete state in the discrete state sequence of the equipment, and establish the state time sequence correlation relationship according to the time sequence to generate the state inertial analysis sequence. The state parameters include equipment degradation degree parameters, operational stability parameters, and defect impact parameters. The state change direction is used to characterize the trend of equipment state change over time, and the state duration is used to characterize the duration of maintenance of the corresponding discrete state.

[0040] S202, based on the state inertia analysis sequence, performs statistical analysis on the state change amplitude, state change rate and state change continuity between adjacent discrete states, and extracts state inertia characteristic parameters. The state inertial characteristic parameters include state duration period, state change rate, state drift direction, state change stability coefficient, and state decay rate. The state duration period is used to characterize the time range within which the current state remains stable; the state change stability coefficient is used to characterize the degree of fluctuation of the device state within a continuous time interval; and the state change rate is calculated by the ratio of the difference between the corresponding state parameters in adjacent time windows to the corresponding time interval.

[0041] S203, based on state inertia characteristic parameters, identifies state intervals that continuously maintain the same direction of change in the discrete state sequence of the equipment, and generates a set of inertial state intervals; The inertial state interval is a continuous time interval in which the direction of state change remains consistent and the rate of state change does not exceed a preset fluctuation threshold. The preset fluctuation threshold is used to limit the abnormal fluctuation amplitude of the state during the inertial continuation process. It can be obtained by statistically analyzing the state change parameters in the historical operating cycle and calculating based on the average value and standard deviation of the state change parameters according to a preset multiple relationship. The preset multiple coefficient can be an empirical value between 1 and 3.

[0042] S204, based on the set of inertial state intervals, performs trend continuity analysis on the corresponding state change parameters within each inertial state interval to generate the corresponding state inertial continuity trajectory; The trend continuation analysis includes: The direction of state continuation is determined based on the state drift direction corresponding to the current inertial state interval; The magnitude of subsequent state changes is determined based on the state decay rate. The range of subsequent state change fluctuations is constrained by the stability coefficient of the state change. The state inertial continuation trajectory is used to characterize the natural continuation and change process of the device state within a continuous time interval.

[0043] S205, Based on the generated state inertia continuation trajectory, perform non-observation state compensation processing on the time interval where no new observation data exists, and generate the predicted state data corresponding to the non-observation time interval; The non-observation state compensation process includes: Use the device state corresponding to the most recent time node as the initial state; Perform state continuation deduction according to the corresponding state inertial continuation trajectory; The abnormal state transitions that occur during the state deduction process are limited by the stability coefficient of the state change.

[0044] S206, Based on the predicted state data, perform continuity constraint processing on the corresponding equipment state within the non-observation time interval to generate a state inertia maintenance relationship; The continuity constraint processing includes: The direction of equipment status change during non-observation time intervals must be consistent with the historical direction of status change. Limit the rate of change of equipment state within the non-observation time interval to not exceed the historical maximum rate of change of the corresponding inertial state interval; The range of equipment status changes during non-observation time intervals is limited to the fluctuation range corresponding to the stability coefficient of the corresponding status change.

[0045] Furthermore, from Figure 3 It can be seen that: When using the traditional linear interpolation method, the device state exhibits a distinct broken-line connection characteristic within the missing interval. Although numerical completion is achieved, local slope abrupt changes occur in the state evolution trend, resulting in non-physical "inflection points" in the state trajectory, thereby weakening the physical consistency of the continuous state evolution.

[0046] After adopting the state inertia preservation relationship constructed by this invention, the state trajectory still maintains continuous nonlinear evolution characteristics in the missing observation interval. Its change process is dominated by the state inertia constraint term, so that the direction and rate of state change are consistent with the historical evolution trend, thereby avoiding the path breakage and trend reversal problems common in traditional interpolation methods.

[0047] Meanwhile, through statistical analysis of the state error at the end of the missing interval, the method of this invention reduces the state reconstruction error by approximately 38% to 52% and improves the state trend consistency index by approximately 25% to 40% compared with the traditional method, indicating that it has higher stability and reliability in expressing the continuous degradation process of equipment. The statistical state fluctuation stability index is shown in Table 1: Table 1 Comparison of state fluctuations before and after inertia maintenance

[0048] Furthermore, based on the state inertia preservation relationship, continuous state evolution correlation processing is performed on time intervals with missing data in the discrete state sequence of the equipment to generate continuous inertial state trajectories, including: S2071, obtain the initial inertial state trajectory generated based on the state inertia maintenance relationship, and analyze the state change direction, state change rate and state change amplitude corresponding to each time window in the initial inertial state trajectory to generate a state inertia evolution baseline sequence. The state inertial evolution baseline sequence is used to characterize the continuous inertial change trend of the device state under uncorrected conditions.

[0049] S2072, based on the state inertia evolution baseline sequence, align and fuse equipment operation data and defect data according to time windows, construct multi-source state deviation indicators, and generate a set of reverse trend triggering candidates; The multi-source state deviation indices include: operational deviation index, temperature rise deviation index, partial discharge deviation index, and defect mutation deviation index; all deviation indices are obtained by normalizing the difference between the current observation value and the corresponding historical inertial evolution baseline.

[0050] S2073, based on the generated set of counter-trend triggering candidates, the deviation indicators corresponding to each candidate are weighted and summed to obtain the counter-trend triggering strength parameter; The reverse trend triggering intensity parameter is used to characterize the degree of deviation of the state from the inertial evolution baseline sequence within the current time window. In the weighted summation, the weight of each deviation index is obtained by calculating the Pearson correlation coefficient between it and the degree of equipment state deterioration and normalizing it to obtain the initial weight. Then, the final weight is determined by linear weighted fusion and normalization after combining it with the frequency of its occurrence in historical abnormal samples.

[0051] S2074, Based on the reverse trend trigger strength parameter, construct a reverse trend trigger indicator function, and determine whether the local state reversal mechanism is triggered according to the indicator function, and generate a local reversal trigger result; The reverse trend trigger indicator function is defined as follows:

[0052] In the formula, This is a counter-trend trigger indicator function. This is the counter-trend trigger strength parameter. To preset the trigger threshold, This is an indicator function.

[0053] The reverse trend trigger indicator function is used to characterize whether the current state meets the reverse trend correction condition. When the trigger signal is output, it is activated.

[0054] S2075, based on the counter-trend trigger signal, performs counter-trend state correction processing on the corresponding time window to generate a local reversal state correction sequence; The reverse trend correction process includes: Determine the direction of inertial state change corresponding to the current time window; Extract the reverse trend correction direction that is opposite to the direction of change in the inertial state; The magnitude of the counter-trend correction is determined based on the counter-trend trigger strength parameter. The specific formula for calculating the counter-trend correction magnitude is as follows:

[0055] In the formula, The reverse trend correction magnitude represents the amount of reverse adjustment applied to the device state at time t. This is the parameter for the maximum counter-trend trigger strength. The state scaling parameter represents the typical fluctuation scale of the state variable and can be defined as the historical standard deviation. This is a correction factor used to control the correction strength. To avoid excessive reversal leading to system oscillation, This is a counter-trend direction function, used to indicate the direction of counter-trend correction. .

[0056] The reverse trend correction magnitude is superimposed on the inertial state change process to achieve local state reverse adjustment.

[0057] S2076, based on the local reversal state correction sequence, maintains the original state inertial evolution baseline sequence for time windows that do not trigger reverse trend correction, and replaces the state results for time windows that trigger reverse trend correction with the corrected state results to generate a fused state sequence. The fusion state sequence is used to simultaneously include the hybrid evolution results of inertial continuation states and local reversal states.

[0058] S2077, based on the fused state sequence, performs continuity verification on state changes between adjacent time windows and generates state continuity verification results; The continuity check includes: whether the rate of change of state exceeds the maximum allowable rate of change and whether the sudden change in state direction exceeds the direction angle threshold. If the rate of change of state does not exceed the maximum allowable rate of change and the direction angle of adjacent state changes does not exceed the preset direction angle threshold, the continuity constraint is satisfied. The direction angle threshold is determined by statistically analyzing the direction angles of adjacent state changes in the historical normal operation state sequence and taking the 95th percentile.

[0059] S2078, Based on the state continuity verification result, perform smooth transition processing on state windows that do not meet the continuity constraints to generate continuous inertial state trajectories; The smooth transition process includes linear interpolation correction using a sliding window; the continuous inertial state trajectory is used to characterize the continuous evolution of the equipment state after introducing local reverse trend correction during the state inertia continuation process.

[0060] It should be noted that the continuous inertial state trajectory generated by the above technical solution introduces a reverse trend triggering and local reversal correction mechanism based on multi-source deviation index on the basis of maintaining state inertia. This enables the equipment state to not only maintain a continuous evolution trend in the missing observation interval, but also to perform directional correction for abnormal mutations and defect-driven non-inertial changes. This effectively avoids the path locking and over-smoothing problems in traditional inertial extrapolation methods, improves the fitting ability of the state trajectory to the real degradation process and the response ability to sudden deterioration events, and makes the generated continuous state trajectory more consistent with the nonlinear and non-uniform degradation evolution law of actual operation of electrical equipment.

[0061] S3. Based on the defect data, establish the defect state inversion relationship and generate the corresponding state disturbance features to characterize the continuous impact of defects on the equipment state trajectory.

[0062] In this embodiment, a defect state inversion relationship is established based on defect data, and corresponding state perturbation features are generated, including: S301, Perform time-series processing on the defect data to generate a defect event sequence; The defect event sequence includes defect occurrence time, defect type parameter, and defect severity parameter; The defect severity parameter is obtained through a standardized defect scoring model:

[0063] In the formula, The standardized defect severity parameter is used to characterize the relative severity of the i-th defect event. This represents the original defect score corresponding to the i-th defect event. The minimum original defect score in the historical defect sample is used as the lower bound for standardization. The maximum original defect score in the historical defect sample is used as the standardization upper limit.

[0064] The defect event sequence is used to characterize the discrete evolution process of defects in the time dimension.

[0065] S302, Based on the defect event sequence, construct the initial feasible state domain of the defect state and represent the device state as an initial state set; The initial feasible state domain is used to characterize the possible range of equipment states before a defect occurs; the initial state set is used to characterize the range of state probability distribution under defect-free conditions.

[0066] S303, based on the initial feasible state domain, introduces a set of defect constraint functions and performs constraint truncation processing on the initial feasible state domain to generate the first stage feasible state domain; The defect constraint function is defined as follows:

[0067] In the formula, Let be the state constraint function corresponding to the i-th type of defect, used to characterize whether the equipment state at time t satisfies the current defect constraint range. The current device status variable is used to represent the status parameter value of the device at the current point in time. The defect reference evolution state is used to represent the reference center position of the equipment state under the current defect condition. The defect reference evolution state is obtained by statistically averaging the state evolution trajectories corresponding to historical samples of the same type of defect. The Euclidean distance between the current equipment state and the defect reference evolution state characterizes the degree to which the current state deviates from the defect reference state. The constraint radius corresponding to the i-th type of defect is used to represent the allowable range of state fluctuations under the current defect conditions.

[0068] S304, based on the first stage feasible state domain, performs backpropagation processing on the defect constraints in reverse chronological order of the occurrence time of the defect events, and recursively updates the historical state space of the equipment according to the defect constraint set corresponding to each time node, generating a multi-stage feasible state domain sequence. The recursive state domain update is as follows:

[0069] In the formula, Let be the updated feasible state domain corresponding to time t. The set of constraints generated from the defect constraints. Let be the feasible state domain at time t-1.

[0070] The backpropagation process is used to perform reverse constraint shrinking of the equipment state range corresponding to the time interval before the defect formation based on the current defect state; the multi-stage feasible state domain sequence is used to characterize the historical gradual degradation process of the equipment state under the influence of the defect.

[0071] S305, based on a multi-stage feasible state domain sequence, generates an evolution chain of the intersection of adjacent state domains and calculates the shrinkage rate of the state domains; The specific formula for calculating the shrinkage rate is as follows:

[0072] In the formula, Let be the feasible region shrinkage rate corresponding to time t, used to characterize the degree of state space reduction. Let t be the volume of the feasible state domain. Let be the volume of the initial feasible state domain.

[0073] S306, based on the state domain intersection evolution chain, centralizes the feasible state domain at each time node to generate a state perturbation reference vector. The centralization process includes obtaining the desired state point of the feasible state domain and using the desired state point as the state perturbation reference. The state perturbation reference vector is used to characterize the offset direction and offset magnitude of the current feasible state domain relative to the initial state.

[0074] S307. Based on the state disturbance reference vector, component decomposition is performed on different defect types to generate state disturbance feature vectors. The state disturbance feature vectors include defect intensity components (obtained by weighted summation of defect severity parameters corresponding to each defect event within the current time window), feasible region shrinkage rate components, and state offset components.

[0075] It should be noted that by using defect constraints to inversely shrink the historical state space of the equipment and constructing a feasible state domain that gradually shrinks over time, the problem of inversion drift accumulation caused by the lack of constraints in the traditional single state trajectory inversion process can be avoided. At the same time, by continuously limiting the uncertainty range of the equipment state through the state domain intersection evolution chain, the generated state disturbance characteristics can not only reflect the gradual deterioration trend before the formation of defects, but also maintain consistency with the actual physical evolution process of the equipment, thereby improving the stability, reliability and continuous evolution expression ability of the defect state inversion results.

[0076] S4, based on the continuous inertial state trajectory and state disturbance characteristics, corrects the equipment state trajectory to generate a defect-corrected state trajectory, including: S401, based on the state perturbation feature vector, performs perturbation intensity analysis on the continuous inertial state trajectory in the time dimension, and constructs the state perturbation driving function field; The state perturbation driving function field is used to characterize the intensity distribution of the influence of defects on the state trajectory within different time windows; The state perturbation driving function field is specifically:

[0077] In the formula, Let be the disturbance driving strength at time t. For state offset components, For the defect strength component, , , These are the normalized weighting coefficients.

[0078] S402, based on the state perturbation driving function field, maps the continuous inertial state trajectory to the observation state space, and constructs the mapping relationship between the inertial state space and the observation state space; The mapping relationship is used to establish the correspondence between the inertial state and the actual observed state.

[0079] S403, Based on the mapping relationship, an observation consistency constraint function is introduced to calculate the consistency deviation of the continuous inertial state trajectory and generate a state deviation sequence; The observation consistency constraint function is specifically as follows:

[0080] The state deviation sequence is used to characterize the degree of deviation between the inertial state trajectory and the actual observed state.

[0081] S404, based on the state deviation sequence and combined with the intensity distribution of disturbance influence, constructs a dual-constraint state projection optimization model; The dual-constraint state projection optimization model is as follows:

[0082] In the formula, This is the observation consistency deviation function value corresponding to time t, used to characterize the degree of inconsistency between the inertial state trajectory and the actual observed state. The larger the value, the more severe the deviation between the model's predicted state and the actual observed state. Let be the inertial state variable corresponding to time t, used to represent the predicted equipment state value generated based on the inertial maintenance relationship. Let be the observed state variable corresponding to time t, used to represent the equipment state measurement value obtained based on the actual collected operational data. Let be the corrected state variable after optimized projection over time t, used to represent the final equipment state that balances inertial consistency and observation consistency. This is an operator used to optimize the solution of variable x, and is used to find the state solution that minimizes the objective function. This is an inertial consistency constraint term, used to constrain the degree of deviation between the corrected state and the inertial state. This is an observation consistency constraint term, used to constrain the degree of deviation between the corrected state and the actual observed state. The balancing coefficient is used to adjust the weight relationship between the inertial consistency constraint and the observation consistency constraint. It can be determined by calculating the relative mean square error ratio of the inertial prediction error and the observation error respectively on historical sample data, and then taking the normalized value or empirical ratio value of the error energy ratio of the two.

[0083] S405. Based on the optimized solution obtained from the dual-constraint state projection optimization model, perform time-window projection correction processing on the continuous inertial trajectory to generate a local corrected state sequence. The projection correction process is used to adjust the inertial trajectory to the optimal compromise between the inertial space and the observation space.

[0084] S406, Based on the local corrected state sequence, perform enhanced correction processing on the time window when the disturbance driving intensity exceeds the preset threshold, and perform smooth preservation processing on the time window when it does not exceed the preset threshold, to generate the final corrected state sequence; The enhanced correction processing refers to adjusting the state trajectory with stronger constraints by increasing the state offset correction magnitude or introducing a higher-weighted disturbance response term within a time window when the disturbance intensity exceeds a preset threshold. This strengthens the representation of the influence of the defect-dominated region on state changes. The smoothing and preservation processing refers to weakening the state changes by using moving averages or low-pass filtering within a time window when the disturbance driving intensity is below a preset threshold. This maintains the continuity and stability of the state trajectory and avoids unnecessary drastic fluctuations. The preset threshold can be directly determined by statistically analyzing the state deviation indicators in historical normal operating state data and taking their empirical quantile values ​​(e.g., the 95th percentile or the mean plus several times the standard deviation).

[0085] S407 generates a defect correction state trajectory based on the final corrected state sequence, which is used to characterize the actual state evolution process of the equipment after combining inertial evolution with observation consistency constraints.

[0086] To verify the effectiveness of the state perturbation driving function field and the double-constraint projection optimization model in suppressing error accumulation, a time series comparison analysis of the state deviation before and after correction was performed, such as... Figure 4 As shown, in the uncorrected method, due to the lack of suppression of historical errors during the time series extension process, the error term exhibits a cumulative amplification trend under the effect of inertial propagation, and its growth curve shows an approximately exponential increase. However, the method of this invention introduces a joint optimization mechanism of observation consistency constraint term and inertial consistency constraint term, and performs local projection correction on the deviation state after the disturbance trigger time point, so that the error propagation path is damped and constrained, thereby gradually converging in subsequent time windows.

[0087] Furthermore, to quantify the error suppression effect, statistical analysis was performed on the state deviation values ​​within each time window, an error propagation sequence was constructed, and the mean square error (MSE) values ​​under different time spans were calculated. The results are shown in Table 2. Table 2. Suppression effect of state error propagation

[0088] As can be seen from Table 2, as the time span gradually increases, the error growth rate of the method of the present invention is significantly lower than that of the uncorrected method, and it can still maintain a low error level under long-term window conditions (such as t=50). This indicates that the method of the present invention can effectively suppress the cumulative diffusion effect in the error propagation process and improve the stability and consistency of the equipment state trajectory in the long-term dimension.

[0089] It should be noted that by introducing observation consistency constraints and disturbance-driven hierarchical correction mechanisms on the basis of inertial state trajectories, the continuity and authenticity of state evolution are optimized in a coordinated manner, thereby effectively reducing the cumulative deviation caused by a single inertial prediction or a single observation correction and improving the stability and reliability of the expression of equipment defect state trajectories.

[0090] S5, acquire new observation data, calibrate the deviation between the corresponding observation state and the defect correction state trajectory, and simultaneously correct the state inertia maintenance relationship and the defect state inversion relationship to update the continuous state trajectory of the equipment, including: S501, acquire new observation data of the target electrical equipment within the current time window, and perform multi-source fusion processing on the new observation data to generate a corresponding observation state sequence; The new observation data includes operational data, experimental data, and defect data.

[0091] S502, perform time alignment matching between the observed state sequence and the state data corresponding to the time window based on the defect correction state trajectory to generate a set of state alignment pairs; S503, based on the set of state alignment pairs, quantifies the deviation between the observed state and the defect correction state trajectory, and generates a state deviation calibration sequence. The specific calculation formula for the state deviation calibration sequence is as follows:

[0092] In the formula, The state deviation calibration sequence for time t. The state corresponding to the defect correction state trajectory.

[0093] S504, Based on the generated state deviation calibration sequence, perform adaptive feedback correction processing on the defect correction state trajectory to generate an updated state correction trajectory; S505, based on the updated state correction trajectory, performs parameter update processing on the state inertia preservation relationship to generate the updated state inertia preservation relationship; The state inertia preservation relationship is expressed as:

[0094] In the formula, To maintain the relationship in the updated state inertia, The inertia retention coefficient is used to characterize the weight of the influence of historical states on the continuation of the current state. The correction influence coefficient is used to characterize the weight of the influence of the current corrected state on the final state.

[0095] Perform parameter update processing, specifically as follows:

[0096]

[0097] In the formula, Let be the inertia retention coefficient at time t. , These are the preset adjustment coefficients. , These are preset threshold parameters. The amplitude of the state fluctuation at time t is used to characterize the rate of state change between adjacent time windows.

[0098] S506, Based on the updated state correction trajectory, perform reverse consistency correction on the defect state inversion relationship to generate an updated defect state inversion relationship; The reverse consistency correction process is used to refit the pre-defect evolution path based on the latest observation bias, so that the inversion relationship satisfies the bidirectional consistency constraint between the historical trajectory and the current observation. Specifically:

[0099] In the formula, Let be the inversion and reconstruction state variable at time t, used to represent the equipment state derived from the defect state inversion relationship.

[0100] S507, based on the updated state correction trajectory and the state inertia maintenance relationship, performs joint reconstruction processing on the continuous state trajectory of the equipment to generate the updated continuous state trajectory of the equipment.

[0101] Furthermore, based on the generated state deviation calibration sequence, adaptive feedback correction processing is performed on the defect correction state trajectory to generate an updated state correction trajectory, including: S5041, construct the original error propagation sequence based on the state deviation calibration sequence corresponding to each time window; The original error propagation sequence is arranged in chronological order and is used to characterize the temporal propagation relationship of errors during the evolution of equipment state.

[0102] S5042, based on the original error propagation sequence, calculates the error propagation distance corresponding to each time window, establishes the error propagation damping kernel function, and generates a damped error propagation sequence; The specific formula for calculating the error propagation distance is as follows:

[0103] In the formula, For the error propagation distance, For the current time window, This represents the time window in which the error first occurs.

[0104] The error propagation damping kernel function satisfies:

[0105] In the formula, For error passing through The residual error after propagation within a time window Let be the initial error intensity at time t, used to represent the original impact of the error within the initial time window. This is the error propagation damping coefficient. The error propagation damping coefficient is determined by statistically analyzing the error decay rate in a historical normal state sequence and taking the average decay rate.

[0106] S5043, based on the damped error propagation sequence, normalizes the error propagation influence intensity corresponding to each time window to generate an error propagation influence weight sequence; The weight of the error propagation effect is defined as follows:

[0107] In the formula, Here is the error propagation impact weight corresponding to time t, used to characterize the proportion of the current time window error affecting the state correction process. This is the cumulative sum of residual error intensities across all time windows, used to normalize the impact of errors. N is the total number of time windows.

[0108] The error propagation impact weight is used to characterize the degree of influence of historical errors on the current state window.

[0109] S5044, Based on the generated error propagation influence weight sequence, perform damped feedback correction processing on the defect correction state trajectory to generate the initial feedback correction trajectory; The damping feedback correction process includes: S5045, calculate the state correction increment based on the error propagation impact weight corresponding to the current time window, and add the state correction increment on the original defect correction state to obtain the updated state result corresponding to the current time window. The specific calculation formula for the state correction increment is as follows:

[0110] In the formula, The state correction increment is the value corresponding to time t. This represents the state deviation value.

[0111] S5046, based on the initial feedback correction trajectory, performs continuity verification on the state changes between adjacent time windows, generates continuity verification results, and when there is a state window that does not meet the continuity constraint, performs local smooth transition processing on the corresponding state window to generate the calibrated state correction trajectory.

[0112] It should be noted that by introducing an error propagation damping kernel function, the propagation process of state error in the time chain is subject to exponential decay constraint, so that the influence of historical error on subsequent state windows gradually weakens as the propagation distance increases. This effectively suppresses the problem of error accumulation and diffusion in the feedback correction process and improves the stability and authenticity of the long-term evolution of the equipment state trajectory.

[0113] S6 generates a continuous degradation state sequence covering the entire operating cycle of the equipment based on the calibrated continuous state trajectory of the equipment, which is used to characterize the continuous evolution process of the equipment state from normal state to degradation state.

[0114] In this embodiment, based on the calibrated continuous state trajectory of the equipment, a continuous degradation state sequence covering the entire operating cycle of the equipment is generated, including: S601: Obtain the calibrated continuous state trajectory of the equipment, and extract the state parameters, state change rate, state disturbance intensity, and state stability parameters corresponding to each time window according to the preset time window to generate a continuous state evolution parameter set. S602, based on the continuous state evolution parameter set, analyzes the continuity of state changes between adjacent time windows, and extracts the state change gradient, state cumulative offset and state evolution duration period to generate a continuous state evolution feature sequence. The state change gradient is used to characterize the direction and magnitude of the change in the device state between adjacent time windows, the cumulative state offset is used to characterize the cumulative deviation of the device state from the initial stable state, and the state evolution duration is used to characterize the duration of the current state's continuous evolution.

[0115] S603, based on the continuous evolution feature sequence of the state, identifies the stage of the continuous change trend of the equipment state in the whole operation cycle, and divides the continuous deterioration stage according to the state change gradient and the cumulative state offset, and generates a set of state stage intervals. Specifically, when the state change gradient is lower than the first preset threshold and the cumulative state offset is lower than the second preset threshold, it is determined to be a stable operation stage; when the state change gradient continues to rise and the cumulative state offset continues to increase, it is determined to be a slow deterioration stage or a continuous deterioration stage; when the state change gradient exceeds the third preset threshold, it is determined to be a sudden deterioration stage; when the cumulative state offset exceeds the fourth preset threshold, it is determined to be a severe instability stage.

[0116] The first preset threshold, the second preset threshold, the third preset threshold, and the fourth preset threshold are determined by statistical analysis of historical normal operation data and according to the mean and standard deviation of the corresponding parameters in a preset multiple relationship.

[0117] S604, based on the set of state stage intervals, performs trend fitting processing on the state evolution characteristics corresponding to each continuous deterioration stage to generate a stage state evolution function. The trend fitting process includes continuously fitting the state change trajectory using a sliding time window and determining the state evolution trend function parameters within the corresponding time window using the least squares method.

[0118] The stage state evolution function is:

[0119] In the formula, The state value corresponding to time window t. The slope parameter represents the state evolution at the corresponding stage. This refers to the state offset parameter for the corresponding stage.

[0120] S605, based on the stage state evolution function, performs smooth transition processing on the state connection boundary between adjacent continuous deterioration stages to generate a full-cycle continuous state connection chain. The smooth transition process includes constraining the rate of change of state at the stage connection boundary and limiting the magnitude of state abrupt changes between adjacent stages to no more than a preset change threshold, so as to avoid non-physical jumps in the state sequence.

[0121] S606, based on the continuous state connection chain throughout the entire cycle, cumulatively calculates the state degradation intensity of the equipment state within the entire operating cycle and generates a state degradation cumulative index sequence. The cumulative state degradation index is obtained by weighting and accumulating the state offset, defect disturbance intensity and state change rate corresponding to each time window, and is used to characterize the overall degradation degree of the equipment in the entire operating cycle. The weighting coefficients in the weighted cumulative calculation are determined by statistical analysis of the correlation between each parameter in the historical equipment failure sample and the equipment failure result.

[0122] S607, based on the cumulative index sequence of condition deterioration, performs continuous deterioration level mapping on the equipment condition and generates a corresponding continuous deterioration condition label sequence. The continuous degradation level mapping includes: If the cumulative state deterioration index is lower than the first-level threshold, it is mapped to a normal operating state. If the cumulative state degradation index is between the first-level threshold and the second-level threshold, it is mapped to a slightly degraded state. If the cumulative state degradation index is between the second-level threshold and the third-level threshold, it is mapped to a moderately degraded state. If the cumulative state degradation index is between the third-level threshold and the fourth-level threshold, it is mapped to a severely degraded state. If the cumulative state degradation index exceeds the fourth level threshold, it is mapped to a failure risk state; The thresholds for each level are determined by fitting the probability distribution of the degradation index distribution in the historical equipment life cycle status samples, and combining the mean, standard deviation and quantiles (such as the 95th or 99th percentile) to perform segmented statistics on the samples, thereby determining the threshold interval boundaries corresponding to each level.

[0123] S608, based on the continuous degradation state label sequence and the full-cycle continuous state connection chain, performs time-series correlation on the state evolution path within the entire operating cycle of the equipment, and generates a continuous degradation state sequence covering the entire operating cycle of the equipment. The continuous degradation state sequence records the dynamic evolution of the equipment state from normal operation to continuous degradation and even failure risk state in chronological order.

[0124] Reference Figure 2 As shown in the diagram, this invention provides a structured management and operational status analysis system for electrical testing data. The system includes a discrete state sequence generation module, an inertia-preserving evolution module, a defect disturbance correction module, and a dynamic calibration and full-cycle degradation module. The modules are interconnected. The discrete state sequence generation module is used to segment the equipment operation time axis based on the irreversible time anchor point, perform directional drift analysis on the state parameters in each segment time interval, establish state transition correlations, and generate discrete state sequences. The inertia preservation evolution module is used to establish state inertia preservation relationships based on discrete state sequences and generate continuous inertial state trajectories. The defect disturbance correction module is used to establish a defect state inversion relationship based on defect data to generate state disturbance characteristics and correct the continuous inertial state trajectory to obtain the defect-corrected state trajectory. The dynamic calibration and full-cycle degradation module is used to perform deviation calibration using new observation data and defect correction state trajectories, synchronously update the inertial hold and defect inversion relationship, and generate a continuous degradation state sequence covering the entire operating cycle.

[0125] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0126] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0127] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0128] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for structured management of electrical detection data and operational status analysis, characterized in that, include: The equipment operating time axis is segmented based on the irreversible time anchor point. Oriented drift analysis is performed on the state parameters within each segment time interval, and state transition correlations are established to generate discrete state sequences. Establish state inertia preservation relationships based on discrete state sequences and generate continuous inertial state trajectories; Based on defect data, a defect state inversion relationship is established to generate state disturbance characteristics, and the continuous inertial state trajectory is corrected to obtain the defect corrected state trajectory. By using new observation data and the defect correction state trajectory for deviation calibration, the relationship between inertial maintenance and defect inversion is updated synchronously to generate a continuous deterioration state sequence covering the entire operating cycle; The establishment of state inertia preservation relationships based on discrete state sequences includes: Obtain the discrete state sequence of the equipment, establish a state inertial analysis sequence in chronological order, and extract the state inertial characteristic parameters between adjacent discrete states; Based on state inertia characteristic parameters, identify state intervals that continuously maintain the same direction of change and generate a set of inertial state intervals; Trend continuity analysis is performed on each inertial state interval to generate state inertial continuity trajectory; For time intervals where no new observation data is available, non-observation state compensation processing is performed based on the state inertia continuation trajectory, and continuity constraints are applied to establish a state inertia maintenance relationship. The generation of continuous inertial state trajectories includes: A state inertia evolution baseline sequence is generated based on the state inertia preservation relationship; The runtime data and defect data are aligned and merged according to time windows to generate a set of reverse trend triggering candidates. The deviation indicators in the candidate set are weighted and fused to obtain the reverse trend trigger strength parameter, and it is determined whether the local state reversal mechanism is triggered. When triggered, a reverse trend state correction is performed and merged with the non-triggered part. After continuity verification and smoothing, a continuous inertial state trajectory is generated.

2. The method according to claim 1, characterized in that, The generation of the discrete state sequence includes: Based on the standard state parameter set, abnormal states that meet the conditions for irreversible states are identified, and a set of time-irreversible anchor points is generated. The time axis of equipment operation is segmented using irreversible time anchor points, and unidirectional state drift analysis is performed to generate a continuous state trajectory within the interval. Extract the state features of the intervals and establish the state transition mapping relationship between them and the anchor points. Then, perform forced transition connections on adjacent intervals to generate the equipment segmented state evolution chain. The continuous state trajectories of each interval are sequentially associated with the state evolution chain to generate a discrete state sequence of the device.

3. The method according to claim 1, characterized in that, The step of establishing a defect state inversion relationship based on defect data to generate state perturbation features includes: Perform time-series processing on the defect data to generate a defect event sequence; An initial feasible state domain is constructed before the defect occurs, and a defect constraint function is introduced to perform constraint truncation processing on the initial feasible state domain to generate the first stage feasible state domain. Following the reverse time sequence of the defect events, the defect constraints are backpropagated to update the historical state space of the equipment and generate a multi-stage feasible state domain sequence. The state domain shrinkage rate is calculated based on the multi-stage feasible state domain sequence, and the feasible state domain at each time node is centered to generate a state perturbation reference vector. Based on the state disturbance reference vector, component decomposition is performed on different defect types to generate state disturbance feature vectors.

4. The method according to claim 1, characterized in that, The modified continuous inertial state trajectory, to obtain the defect-corrected state trajectory, includes: Based on the state perturbation feature vector, the perturbation intensity of the continuous inertial state trajectory is analyzed, and the state perturbation driving function field is constructed. The continuous inertial trajectory is mapped to the observation state space, and observation consistency constraints are introduced to calculate the deviation sequence between the inertial state and the observation state. Based on the deviation sequence and disturbance intensity, a dual-constraint state projection optimization model is constructed. Projection correction processing is performed on the inertial state trajectory, and enhanced correction or smooth preservation is performed according to the disturbance driving intensity to obtain the defect-corrected state trajectory.

5. The method according to claim 1, characterized in that, The process of using new observation data and the defect correction state trajectory for deviation calibration, and synchronously updating the relationship between inertial preservation and defect inversion, includes: Acquire new observation data and generate an observation state sequence, align it with the defect correction state trajectory in time, and calculate the state deviation calibration sequence. Adaptive feedback correction is performed on the defect correction state trajectory based on the state deviation calibration sequence to generate an updated state correction trajectory; Based on the updated state correction trajectory, the state inertia preservation relationship and the defect state inversion relationship are updated respectively, and a joint reconstruction is performed to generate the updated continuous state trajectory of the equipment.

6. The method according to claim 5, characterized in that, The adaptive feedback correction of the defect correction state trajectory based on the state deviation calibration sequence, generating an updated state correction trajectory, includes: The original error propagation sequence is constructed based on the state deviation calibration sequence, the error propagation distance is calculated and the error propagation damping kernel function is established to generate the damped error propagation sequence; The influence intensity of error propagation in the damped error propagation sequence is normalized to generate an error propagation influence weight sequence. Based on the error propagation influence weight sequence, damped feedback correction processing is performed on the defect correction state trajectory to generate an initial feedback correction trajectory. The continuity of the initial feedback correction trajectory is verified, and local smoothing transition processing is performed on the state windows that do not meet the continuity constraints to generate the calibrated state correction trajectory.

7. The method according to claim 1, characterized in that, The generation of a continuous degradation state sequence covering the entire operating cycle includes: Extract continuous state evolution feature sequences based on the calibrated continuous state trajectory of the equipment; Based on the continuous state evolution feature sequence, the continuous deterioration stage is divided according to the state change gradient and the cumulative state offset. Trend fitting is performed on the state evolution characteristics within each continuous deterioration stage to generate stage state evolution functions, and the state connection boundaries between adjacent stages are smoothed to generate a full-cycle continuous state connection chain. Based on the continuous state connection chain throughout the entire cycle, calculate the cumulative index of state degradation, perform level mapping, and generate a continuous degradation state label sequence. The tag sequence is temporally associated with the connection chain to generate a continuous degradation state sequence covering the entire operating cycle of the device.

8. A system for structured management and operational status analysis of electrical testing data as described in any one of claims 1-7, characterized in that, include: The discrete state sequence generation module is used to segment the equipment operation time axis based on the irreversible time anchor point, perform directional drift analysis on the state parameters in each segment time interval, establish state transition correlations, and generate discrete state sequences. The inertia preservation evolution module is used to establish state inertia preservation relationships based on discrete state sequences and generate continuous inertial state trajectories. The defect disturbance correction module is used to establish a defect state inversion relationship based on defect data to generate state disturbance characteristics and correct the continuous inertial state trajectory to obtain the defect-corrected state trajectory. The dynamic calibration and full-cycle degradation module is used to perform deviation calibration using new observation data and defect correction state trajectories, synchronously update the inertial hold and defect inversion relationship, and generate a continuous degradation state sequence covering the entire operating cycle.

Images