Intelligent identification management method and system for whole life cycle of power assets

Abstract

The application discloses a smart identification management method and system for the whole life cycle of power assets, relates to the technical field of power asset management, and comprises the following steps: generating a unique identification for the power asset and establishing an archive in a trusted data platform; embedding modeling is performed on multi-source data under the association of the unique identification, context representation is generated, and a causal correlation result is obtained through causal refinement; a privacy topology factor is introduced in the causal correlation to form a privacy-enhanced input; based on the privacy-enhanced input and anchor representation, trusted identification coding is obtained by executing conditional adversarial generation; and the coding is used to drive the dynamic linkage of runtime clocks, calibration clocks and warranty clocks, trigger compliance tasks and write back the results to the archive. The balance between privacy protection and topology maintenance is achieved, the robustness and credibility of the identification coding are improved, a closed loop of task linkage and value updating is formed in the whole life cycle management, and the integrity and compliance of the power asset management are improved.

Description

Technical Field

[0001] This invention relates to the field of power asset management technology, and in particular to an intelligent identification management method and system for the entire life cycle of power assets. Background Technology

[0002] As a core resource for power system operation, the full lifecycle management of power assets directly impacts power supply security, operational efficiency, and economic benefits. Traditional power asset management models typically rely on independent information systems for various business departments, such as procurement, operation monitoring, calibration, and finance systems. The lack of a unified data index and interconnectivity mechanism among these systems leads to issues like duplicate data entry, inconsistent definitions, and information silos at different stages. For example, status parameters collected during equipment operation are difficult to trace back to design files in a timely manner, and calibration results and maintenance records are often disconnected from operational status data, thus affecting the accuracy of asset lifespan assessments and maintenance decisions.

[0003] In recent years, with the development of new technologies such as IoT sensing, blockchain, and artificial intelligence, the full lifecycle management of power assets has gradually evolved towards intelligence and digitalization. IoT sensors can collect operational data such as voltage, current, and temperature in real time; blockchain technology can achieve tamper-proof data storage and evidence sharing among multiple parties; and artificial intelligence algorithms have been applied to health assessment and lifespan prediction. However, existing methods still have the following technical shortcomings: First, there is a lack of a unified asset identification management mechanism, making it difficult to match device codes in different systems, resulting in data not forming a complete archive; second, existing state modeling mostly relies on surface correlation analysis, making it difficult to identify the true causal relationship between variables, thus easily leading to misjudgments in prediction and diagnosis; third, there is a contradiction between privacy protection and data sharing, as key parameters (such as geographical location and inspection records) must both participate in analysis and be protected from leakage; fourth, existing smart identification generation methods are mostly single hash or encoding methods, lacking robustness against attacks and easily forged or tampered with; fifth, in terms of asset compliance management, operating cycles, inspection cycles, and warranty cycles are often managed separately, lacking a unified linkage mechanism, resulting in delayed or duplicated task scheduling and failing to form a closed loop for value updating. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an intelligent identification management method for the entire life cycle of power assets, which solves the problems of severe information silos, insufficient identification of causal relationships, and difficulty in forming a closed loop in the existing power asset life cycle management in terms of data integration, causal modeling, and compliance linkage.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an intelligent identification management method for the entire life cycle of power assets, which includes collecting power asset life cycle data, establishing an archive of power asset life cycle data and unique identifiers and storing it in a trusted data platform.

[0008] By linking power asset lifecycle data with unique identifiers, contextual information is formed based on embedded modeling, and causal relationship results are obtained through causal refinement.

[0009] Introducing privacy-preserving topological factors into causal association results to mask sensitive information while maintaining topological criticality creates privacy-enhanced inputs.

[0010] Adversarial generation is performed on privacy-enhanced inputs to obtain robust and trustworthy identifier encodings;

[0011] By leveraging trusted identifier encoding to drive the linkage of operating clocks, calibration clocks, and maintenance clocks, compliance tasks are triggered and the results are recorded back into the power asset archive, thereby achieving value updates.

[0012] As a preferred embodiment of the intelligent identification management method for the entire life cycle of power assets described in this invention, the power asset life cycle data includes the design parameters, operating status, and maintenance records of the power assets.

[0013] The process of establishing archives for power asset lifecycle data and unique identifiers includes generating a unique identifier for each power asset in a trusted data platform, collecting and archiving the power asset's design parameters, operating status, and maintenance records into corresponding archive entries, and binding and storing the unique identifier in the archive entries.

[0014] As a preferred embodiment of the intelligent identification management method for the entire life cycle of power assets described in this invention, the formation of context information includes: in a trusted data platform, using the unique identifier of the power asset as an index, uniformly aligning design parameters, operating status data, and maintenance event records; extracting statistical operating features according to multi-scale time windows of short-term, medium-term, and long-term; and combining the time decay effect of maintenance events and the diffusion relationship of the power network topology neighborhood to generate a context embedding vector that reflects the historical status of equipment, the impact of events, and network coupling.

[0015] The process of obtaining causal association results through causal refinement includes: based on the context embedding vector, using a causal discovery method constrained by a directed acyclic graph, calculating the causal association relationships between various operational features, and forming a causal association result matrix. The causal association results can distinguish between direct causal effects and superficial correlations, and output the feature representation after causal propagation.

[0016] As a preferred embodiment of the intelligent identification management method for the entire life cycle of power assets described in this invention, the step of introducing privacy topology factors into the causal association results includes setting masking coefficients for corresponding dimensions in the causal association result matrix according to preset sensitive field rules in a trusted data platform, generating retention coefficients by combining the feature importance after causal propagation, and merging the masking coefficients and retention coefficients to obtain privacy topology factors. The privacy topology factors are a set of adjustment parameters set dimension by dimension, used to simultaneously reduce the impact of sensitive features in subsequent processing and keep the main causal relationship intact.

[0017] The privacy-enhanced input includes multiplying the privacy topology factor with the feature representation after causal propagation in each dimension to obtain the privacy-preserving context input, which weakens the sensitive features in each dimension to generate privacy-enhanced input data.

[0018] As a preferred embodiment of the intelligent identification management method for the entire lifecycle of power assets described in this invention, the generation of the trusted identification code includes: in a trusted data platform, using the privacy-enhanced input as the generation condition, and combining it with the anchoring representation bound to the unique identifier, configuring a generation network and a discrimination network, and performing conditional adversarial generation under the combined action of binding constraints, privacy suppression constraints, structural compliance constraints, and stability constraints to obtain a trusted identification code that satisfies the constraints; binding and storing the trusted identification code with the asset file in the trusted data platform to drive subsequent lifecycle management tasks.

[0019] As a preferred embodiment of the intelligent identification management method for the entire lifecycle of power assets described in this invention, the linkage between the operating clock, calibration clock, and warranty clock driven by the trusted identification code includes: extracting operating stress factor and metering deviation information based on the trusted identification code and the corresponding context input in a trusted data platform; dynamically accumulating the operating clock according to the operating stress factor; gradually accumulating the calibration clock according to the metering deviation information; and decreasing the warranty clock in a countdown manner according to the warranty period, so as to realize the dynamic update of the status of the operating clock, calibration clock, and warranty clock, and maintain continuous binding with the trusted identification code.

[0020] As a preferred embodiment of the intelligent identification management method for the entire lifecycle of power assets described in this invention, the step of triggering compliance tasks and recording the results back to the power asset archive includes: when the state of any clock reaches a preset threshold condition, generating a trigger indication; executing task scheduling in the trusted data platform according to compliance priority based on the trigger indication, forming corresponding operation and maintenance work orders, inspection tasks, or warranty processing orders; and binding and storing the operation records with the trusted identification code after the task is completed, while updating the status information and value information in the power asset archive, thereby realizing closed-loop management of the entire lifecycle.

[0021] Secondly, this invention provides an intelligent identification management system for the entire lifecycle of power assets, including:

[0022] The data acquisition module is used to collect design parameters, operating status and maintenance records of power assets, and to establish files with unique identifiers and bind and store them in a trusted data platform;

[0023] The context modeling module is used to uniformly align the design parameters, running status data, and maintenance event records in the trusted data platform using a unique identifier as an index, and generate a context embedding vector.

[0024] The causal refinement module is used to obtain the causal association result matrix and output the feature representation after causal propagation based on the context embedding vector and the causal discovery method constrained by the directed acyclic graph.

[0025] The privacy topology factor construction module is used to generate privacy topology factors in a trusted data platform according to preset sensitive field rules, and combine them with the feature representation after causal propagation to form privacy-enhanced input;

[0026] The adversarial generation module is used in the trusted data platform to perform conditional adversarial generation using the privacy-enhanced input as the generation condition and in combination with the anchor representation bound to the unique identifier, to obtain the trusted identifier code and bind and store it with the asset file.

[0027] The clock linkage module is used to drive the dynamic update of the status of the running clock, verification clock and maintenance clock based on the trusted identifier encoding, and to trigger a compliance task when the threshold condition is reached, and record the result back to the power asset file to realize value update.

[0028] Thirdly, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the intelligent identification management method for the entire life cycle of power assets as described in the first aspect of the present invention.

[0029] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the intelligent identification management method for the entire life cycle of power assets as described in the first aspect of the present invention.

[0030] The beneficial effects of this invention are as follows: By generating unique identifiers and establishing digital archives for power assets in a trusted data platform, unified archiving and secure storage of design parameters, operating status, and maintenance records are achieved, avoiding information silos caused by the fragmentation of multiple systems. Based on this, embedded modeling combined with multi-scale time windows and topology diffusion mechanisms is used to generate contextual information, and direct causal relationships between operating features are identified through causal refinement methods, significantly improving the accuracy of state modeling. Furthermore, privacy topology factors are introduced to mask and retain the weights of sensitive fields in causal results dimension by dimension, protecting privacy while preserving key topological relationships, forming privacy-enhanced input data. Subsequently, a robust trusted identifier code is generated under multiple constraints based on a conditional adversarial generative structure, preventing forgery and tampering at the algorithmic level. Finally, the trusted identifier code drives the dynamic linkage of the operating clock, verification clock, and maintenance clock, automatically triggering compliance tasks and writing back the archives when threshold conditions are reached, achieving closed-loop management throughout the entire lifecycle. This method not only improves the intelligence and security of power asset management but also ensures the consistency and value updating of the operation, verification, and maintenance processes. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart of a smart identification management method for the entire lifecycle of power assets. Detailed Implementation

[0033] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0034] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0035] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0036] Reference Figure 1 This is one embodiment of the present invention, which provides a smart identification management method for the entire lifecycle of power assets, including the following steps:

[0037] S1: Collect power asset lifecycle data, establish archives with unique identifiers for power asset lifecycle data, and store them on a trusted data platform.

[0038] Before power assets are put into management, the trusted data platform generates a unique identifier for each power asset and establishes a corresponding file entry in the platform. The generation of the unique identifier is achieved through encrypted computation:

[0039]

[0040] in, This serves as a unique identifier for the generation of power assets. This means taking the first 96 bits of the hash result as the output to ensure that the unique identifier meets the fixed length requirement; This represents the message authentication code calculation based on the SHA-256 hash function; The key is assigned by the key management system of the trusted data platform and is used to control the security of identifier generation; This indicates the manufacturer code, used to identify the manufacturer of the power asset; This indicates the model code, used to identify the model category of power assets; It indicates the factory serial number, used to distinguish different power assets of the same model; Represents a timestamp, used to record the time information when a unique identifier for a power asset is generated; Represents a random number, generated by a hardware random number generator, used to enhance the unpredictability of unique identifiers; This represents the concatenation operator, used to concatenate the manufacturer code, model code, serial number, timestamp, and random number in sequence to form an input sequence.

[0041] The unique identifier is calculated using the manufacturer code, model code, serial number, timestamp, and random number, and is controlled by a key. After being processed by a hash function, it is truncated to a fixed length to ensure uniqueness and anti-counterfeiting.

[0042] After generating a unique identifier, power asset lifecycle data is collected, including design parameters, operating status, and maintenance records. Design parameters are imported at the equipment manufacturing stage, operating status is acquired in real time through online sensors and monitoring systems, and maintenance records are uploaded by the operation and maintenance terminal during inspections and maintenance. All power asset lifecycle data is archived one by one into corresponding file entries within a trusted data platform and bound to a unique identifier, stored in an append-only manner to ensure data integrity and immutability.

[0043] By using a uniquely generated encrypted identifier as an index for lifecycle data, design parameters, operating status, and maintenance records are uniformly bound together, ensuring data continuity and security, and providing trusted input for subsequent causal refinement and privacy topology processing.

[0044] In traditional technologies, the identification of power assets often relies on manual numbering or decentralized management systems, which are prone to duplication, omissions, or tampering, leading to inconsistent data traceability. The system generates a unique identifier for each power asset through a trusted data platform and binds it to design parameters, operating status, and maintenance records, forming an immutable archival entry. This ensures that assets have a unified digital identity index throughout their entire lifecycle, avoiding duplicate records and information silos, and enabling the archives to serve as reliable input for subsequent causal modeling and clock management.

[0045] S2: Based on the association between power asset lifecycle data and unique identifiers, contextual information is formed through embedded modeling, and causal association results are obtained through causal refinement.

[0046] After completing the unique identifier generation and lifecycle data archiving described in S1, the archived data is used as an index to perform context modeling and causal refinement.

[0047] First, within the trusted data platform, the design parameters, operating status, and maintenance records of the power asset are uniformly aligned, using the archive entries corresponding to unique identifiers as units. For operating status data, time windows are divided into short-term, medium-term, and long-term time scales, and statistical features are extracted within each time scale to ensure that both short-term operational fluctuations and medium- to long-term stable trends are reflected. For maintenance records and alarm events, time decay weights are calculated based on the chronological order of events and the time interval from the current date, thereby highlighting the impact of recent events on equipment status while reducing the interference of older events. Based on this, and combined with the power network topology information, the operating characteristics of adjacent nodes of the equipment corresponding to the unique identifier are weighted according to topological diffusion relationships, incorporating the operational influence from neighboring equipment. Finally, a context embedding vector integrating design parameters, operating status, maintenance events, and topological neighborhood information is obtained.

[0048] After generating contextual information, causal refinement is further performed on this information. The core of causal refinement lies in using a causal discovery method constrained by a directed acyclic graph to identify genuine causal relationships from the context embedding vectors. Its optimization process is defined as follows:

[0049]

[0050] in, This represents the feature matrix composed of context embedding vectors; This represents a causal adjacency matrix, where the matrix elements reflect the causal strength between different features. This represents the number of samples, used to normalize the objective function. The Frobenius norm is used to measure the size of the residuals. This represents the sparsity parameter, which controls the sparsity of causal edges; This represents the L1 norm, used to encourage sparse solutions; This represents Hadamard element-wise multiplication; Indicates matrix exponentiation; Represents the trace operation of a matrix; Represents the feature dimension, used for acyclic constraints; This represents the optimal causal relationship matrix obtained through optimization.

[0051] In obtaining the causal adjacency matrix Subsequently, the characteristic representation of causal consistency is obtained through causal propagation calculation:

[0052]

[0053] in, Represents a d-dimensional identity matrix; It represents the contextual features after causal propagation, integrating direct and indirect causal effects.

[0054] To measure the criticality of different features within a causal framework, a causal criticality index is defined:

[0055]

[0056] in, This represents the causal importance of feature r; the larger the value, the more significant its influence on other features. This represents the index of the target feature pointed to by feature r, used for the summation range.

[0057] Ultimately, the result of causal refinement includes a causal adjacency matrix. Characteristic representation after causal propagation and a set of causal keyness indicators These three factors serve as inputs for the subsequent construction of privacy-preserving topology factors. Through this step, the proposed solution elevates the approach from traditional correlation analysis to causal relationship modeling, significantly improving the explanatory power of power asset operation mechanisms and enhancing the robustness of subsequent privacy protection and identifier generation.

[0058] Traditional methods often rely on monitoring data from a single time window, failing to comprehensively reflect the historical status and topological environment of the equipment, thus easily leading to biases when inferring operational relationships. This paper addresses this issue by using a unique identifier as an index to uniformly align design parameters, operational status data, and maintenance event records. Features are extracted according to multi-scale time windows, and causal refinement methods are combined to calculate the causal relationships between these operational features. This design captures both the short-term dynamics of equipment operation and preserves long-term trends, while avoiding interference from superficial correlations through causal constraints, thereby outputting an interpretable causal structure.

[0059] S3: Introduce privacy-preserving topological factors into the causal association results to mask sensitive information and maintain topological criticality, thus forming privacy-enhanced inputs.

[0060] The platform assigns sensitivity scores to each feature dimension corresponding to the causal relationship result matrix based on preset sensitive field rules (field classification, compliance tags, legal level, etc.), thus obtaining a sensitivity score. (Range [0, 1]). Subsequently, the sensitivity scores are converted into soft mask weights according to rules. (Range [0, 1]): The higher the score, the lower the corresponding soft mask weight, indicating that this dimension will be weakened more strongly in subsequent processing.

[0061] In the power grid adjacency matrix The principal eigenvectors are calculated and normalized to obtain device-level centrality, which is then mapped to the feature dimension:

[0062]

[0063]

[0064]

[0065] in, The adjacency matrix of a power network (constructed from a single connection or equivalent distance, with nodes representing power asset equipment); express The principal eigenvector (the right eigenvector corresponding to the largest eigenvalue). Representation matrix The largest eigenvalue; Represents device node Normalization centrality; Representing vectors At the node The component at the location; Represents the summation index over all nodes (used for summing). (component summation and normalization); Representing feature dimension Topological criticality weights (mapping device centrality to feature dimensions). Represents the elements of the feature-device mapping matrix (if the feature Originating from equipment but (otherwise it is 0); Indicates the device node index; Indicates the feature dimension index.

[0066] Platform for causal criticality With topological criticality Perform summation and normalization separately to obtain two sets of dimensionless weight vectors, ensuring consistent dimensions and comparability. Then, use the average sensitivity of the current sample (or batch) (i.e., all...) Using the mean of the input, an adaptive balance coefficient is calculated through a smoothed Sigmoid function. When the overall sensitivity is high, Taking a larger value, the system places more emphasis on causal criticality to reduce the risk of topological information leakage; when the overall sensitivity is low, By selecting the smaller value, the system places greater emphasis on topological criticality to preserve structural information relevant to operational safety. Finally, the platform obtains the joint criticality weight for each dimension through a weighted fusion of "causal criticality weight and topological criticality weight". .

[0067] Soft mask With joint criticality Multiply the components and arrange them dimension by dimension into a diagonal matrix, applying a lower bound to the diagonal elements if necessary. To ensure numerical stability:

[0068]

[0069] in, Representing the privacy topology factor, it is a set of adjustment parameters in the form of a one-dimensional diagonal matrix (size...). ); This indicates the operation of taking the larger of the two values ​​one dimension at a time; The lower limit constant for the values ​​of the diagonal elements ( To ensure numerical stability; Indicates the soft mask weight; Indicates the joint keyness weight; Indicates the number of feature dimensions; Indicates by index Dimension by dimension are placed into the diagonal to form a diagonal matrix.

[0070] Using PTF to represent causal propagation Dimensionally weighted input yields privacy-enhanced input:

[0071]

[0072] in, Contextual input indicating enhanced privacy.

[0073] Traditional privacy protection methods often employ static masking or global noise addition. While these methods can mask sensitive information, they frequently disrupt crucial topological relationships, leading to distorted subsequent analysis. This system sets masking and retention coefficients for different dimensions based on causal criticality, generating dimensionally adjustable privacy topological factors. These factors are then multiplied with the feature representation after causal propagation to form privacy-enhanced input. This approach masks sensitive fields while maintaining the integrity of the causal framework, ensuring that the input data satisfies privacy requirements while preserving the expressive power of core features, providing reliable input for subsequent adversarial generation.

[0074] S4: Perform adversarial generation on the privacy-enhanced input to obtain a robust and trusted identifier encoding.

[0075] Read the unique identifier ID of the asset (generated and stored by S1) from the trusted data platform; map the ID to a fixed-dimensional anchored representation. (Implemented through a learnable or fixed mapping layer) to serve as an "identity anchor" in the generator. (Implementation details: If irreversibility is required, a one-way embedding can be performed on the platform using parameters managed by the key; if training is required, the mapping network can be trained synchronously, but the mapping process must run under the platform's controlled key / access.) Define a regularized real sample generator (e.g., an encoding framework and check bit structure based on the ID) in the trusted data platform to generate a "real codeword" reference distribution for the discriminator to learn the constraints of real semantics and structure; use this regularized real sample as a positive sample for the discriminator in adversarial training to ensure that the final generated result is not only "realistic" but also "compliant and verifiable".

[0076] In obtaining privacy-enhanced input and the anchoring representation of the unique identifier Next, the system constructs a conditional adversarial generative structure within the trusted data platform. Specifically, the generator network is first configured. ,by As input, output candidate identifier code Simultaneously configure a discriminator network. Its input is a combination of "encoding and conditions", used to distinguish candidate identifier encodings from regularized true samples. In this structure, the discriminator improves its discriminative ability by minimizing the following loss function:

[0077]

[0078] in, Indicates the discriminator network (parameters) The input is (encoding, condition), and the output is the probability value of belonging to "true sample"; This represents a "true codeword" sample generated by rules (derived from the verifiable code obtained from the ID according to the rules). This represents the candidate code currently output by the generator (the generator's output). Input indicating enhanced privacy; This represents the expected value for the training batch of samples; This indicates the loss of the discriminator.

[0079] In this optimization process, the generator must not only "fool" the discriminator, but also simultaneously satisfy binding constraints, privacy suppression constraints, structural compliance constraints, and stability constraints. The final output... The results of the trusted identifier coding candidates that meet multiple conditions will proceed to the next step of screening and archiving.

[0080] By minimizing The discriminator gains stronger distinguishing ability. Meanwhile, the generator minimizes the following expression based on the discriminator's feedback:

[0081]

[0082] in, The overall loss of the generator; Represents a generator network; A linear mapping matrix (or part of a mapping network) used to map the generated code to the anchored representation. Comparable representation space; Anchored representation for ID; It is the Euclidean norm. For binding constraint terms, the weight coefficient is... ; This represents a privacy suppression regularization term with a weight coefficient of . ; This represents the sensitive subspace that has been masked out. Element-wise multiplication; The weights represent structural compliance constraints (such as differentiable check bit difference measures). >0; For stability constraints (e.g., consistency loss due to noise sampling), weights ; All are positive hyperparameters, used for tuning in engineering using a validation set.

[0083] The training process includes:

[0084] Constructing a training batch: sampling a number of assets Obtained through a rule-based process And calculate the corresponding Discriminator step: Minimize renew .

[0085] Generator step: Minimize renew With mapping Set parameters (and update the privacy regularization estimator, such as MINE / InfoNCE). Alternately train the discriminator and generator until they stabilize (using engineering techniques such as early stopping, learning rate decay, and gradient penalty).

[0086] Regular evaluation: verify compliance pass rate, privacy breach estimate, robustness metrics (adversarial / noise testing), etc.; save the best model and version it.

[0087] Reasoning (online): Input With corresponding ID; calculate anchor representation Generate fixed noise or deterministic seeds, and execute... .

[0088] Quantize / binarize and perform structural verification; if verification fails, trigger the nearest correctable solution (such as minimum distance error correction).

[0089] If the final code passes verification, it is written back to the trusted data platform as a trusted identifier and bound to the asset file; at the same time, training / inference metadata is recorded for auditing.

[0090] Traditional identifier generation methods often rely on fixed rule encoding, lacking defense mechanisms against anomalous attacks or forgery, making asset identifiers susceptible to tampering or counterfeiting. This system, based on privacy-enhanced input and anchored representations, constructs a conditional adversarial structure between the generator and discriminator networks, generating trusted identifier codes under multiple constraints. This generation method ensures structural compliance while resisting forgery interference, producing outputs that match asset profiles and possess robustness, thus guaranteeing the stability and trustworthiness of identifiers throughout their lifecycle.

[0091] S5: Utilize trusted identifier encoding to drive the linkage of operating clock, calibration clock, and maintenance clock to trigger compliance tasks and record the results back to the power asset archive, thereby achieving value updates.

[0092] In the trusted data platform, trusted identifier encoding is used to complete identity verification and reverse lookup, uniquely locating the asset file of the device; the current status of three types of clocks (last update time, current cumulative amount / remaining amount), threshold configuration, rule version, as well as the processing timestamp and data window of the current period are read; privacy-enhanced input is used as context input and cached together with asset file parameters before entering the clock update stage.

[0093] To ensure that the clock advance reflects real-world operating conditions, this step extracts the composite stress factor from the privacy-enhanced input and estimates the rate of measurement deviation by combining it with historical calibration records.

[0094] Comprehensive stress factor calculation:

[0095]

[0096] in, Indicates time The comprehensive stress factor represents the overall working condition intensity that the equipment experiences at that moment; This represents a smooth nonlinear function, used to ensure that the output is non-negative and has saturation characteristics; This represents the stress offset parameter (real number), obtained from platform calibration, and reflects the foundation loss level of the equipment under zero load. This represents the stress weight vector (with the same dimensions as the input), calibrated using historical data, reflecting the sensitivity of different working conditions to stress. Indicates time The privacy-enhancing input vector is obtained by the dimension-wise absolute value.

[0097] Point estimate of measurement deviation rate The platform completes the process by using the deviation residuals from previous calibration records and combining them with current operating parameters (such as temperature and load) to update the point estimate at the current moment online through a regression model.

[0098] At a unified discrete time step The following dynamic accumulation is performed on the three types of clocks using different mechanisms to ensure that it reflects real operating conditions, calibration drift trends, and warranty wear; the formula for the three-clock accumulation calculation is expressed as follows:

[0099]

[0100] in, This represents the cumulative amount of the running clock, and the equivalent running time weighted by the operating conditions. This indicates the cumulative amount of the calibration clock, and the cumulative amount of the measurement deviation. The remaining amount on the warranty clock indicates the remaining warranty period; This represents the platform-defined uniform time step, used for discrete clock updates. The update factor for various clock types is determined by the equipment category / operating specifications; This represents the overall stress factor.

[0101] After updating the three clock states, the trusted data platform compares them with preset thresholds. When the accumulated value of the running clock exceeds the specified operating threshold, the system determines that the equipment has reached the condition requiring operation and maintenance; when the accumulated value of the verification clock reaches or exceeds the set verification threshold, the system determines that the equipment has entered a critical state of measurement deviation and requires verification; when the remaining value of the warranty clock is less than or equal to the warranty threshold, the system determines that the equipment is about to be out of warranty and requires warranty processing. The comparison process involves judging each condition one by one, and a corresponding trigger indication is generated as soon as any condition is met. This trigger indication is sent to the scheduling module to enter the compliance task generation stage.

[0102] When one or more trigger indications occur simultaneously, the trusted data platform schedules tasks according to compliance priority rules. First, verification and warranty triggers involving legal regulations or security requirements are prioritized; second, the timeliness of operation and maintenance is considered; and finally, operation and maintenance costs and resource consumption are comprehensively assessed. Based on the scheduling results, the platform generates corresponding work orders: for clock-based triggers, an operation and maintenance work order is generated; for clock-based triggers, a verification work order is generated; and for clock-based triggers, a warranty processing order is generated. All work orders are equipped with a trusted identification code to ensure that the task is strongly bound to a specific device and is traceable.

[0103] Maintenance or inspection personnel scan trusted identification codes on-site using mobile terminals or edge devices, automatically retrieving equipment files and work order information. After on-site operations are completed, maintenance records, inspection results, warranty processing details, and other data are uploaded to the trusted data platform. The platform re-binds these records with the corresponding trusted identification codes and writes them into the power asset file, updating the equipment's operating status and value information. Simultaneously, the platform automatically updates three types of clock states based on the nature of the task; for example, after inspection, the inspection clock is reset to zero or reset; after warranty processing, the warranty clock is reset, ensuring that the clock state is consistent with the actual operation. Thus, the compliance task is processed in a closed loop.

[0104] To ensure traceability throughout the entire process, the trusted data platform generates complete audit logs for every trigger, scheduling, task execution, and file write-back. These audit logs include operation time, threshold configuration, task type, execution result, operator information, and system signature information. Through this audit information, regulatory agencies or internal audit departments can verify the compliance of equipment lifecycle management at any time. Simultaneously, the platform sends stress factors, deviation estimates, and task execution efficiency to the continuous optimization module for adjusting thresholds, optimizing clock parameters, and improving resource scheduling strategies, thus forming a data-driven, self-evolving closed loop.

[0105] In the refined causal feature representation, to avoid directly exposing sensitive fields while maintaining key dependencies in the power network topology, a privacy topology factor is introduced as a dimension-wise adjustment parameter within the trusted data platform. This factor differentiates and adjusts different dimensions based on causal criticality, enabling both masking and retention to be achieved simultaneously. This ensures that the input data meets privacy protection requirements while maintaining the integrity of the causal propagation structure.

[0106] After receiving privacy-enhanced input, a conditional adversarial generative network (GDN) is used to generate identifier codes. The generator produces candidate codes under the constraints of the privacy-enhanced input and anchoring information, and the discriminator judges the consistency of the generated results with reference conditions. This process is repeatedly trained under multiple constraints, ensuring that the generated identifier codes maintain stability and correspondence with actual assets, and preventing abnormal interference from compromising identifier consistency.

[0107] After the identification code is generated, the system binds it to the operating clock, calibration clock, and warranty clock. The operating clock dynamically accumulates based on the operating stress factor, the calibration clock gradually accumulates based on the measurement deviation, and the warranty clock decreases based on the warranty period. The status of all three changes in conjunction with the update of the identification code. When any clock reaches a preset condition, the system automatically triggers a task and generates a corresponding work order, calibration record, or warranty processing order. After the task is completed, the results are written back to the asset file.

[0108] Through the above steps, the identification information of power assets forms a unified closed-loop management with the operation, inspection, and maintenance processes, ensuring consistency in privacy protection, identification credibility, and task triggering, and enabling continuous updates of asset status and value information at the archive level.

[0109] This embodiment also provides an intelligent identification management system for the entire lifecycle of power assets, including:

[0110] The data acquisition module is used to collect the design parameters, operating status and maintenance records of power assets, and to establish files with unique identifiers and bind and store them in a trusted data platform.

[0111] The context modeling module is used to uniformly align the design parameters, running status data, and maintenance event records in the trusted data platform using a unique identifier as an index, and generate a context embedding vector.

[0112] The causal refinement module is used to obtain the causal association result matrix and output the feature representation after causal propagation based on the context embedding vector and the causal discovery method with directed acyclic graph constraints.

[0113] The privacy topology factor construction module is used to generate privacy topology factors in a trusted data platform according to preset sensitive field rules, and combine them with the feature representation after causal propagation to form privacy-enhanced input.

[0114] The adversarial generation module is used in a trusted data platform to perform conditional adversarial generation using the privacy-enhanced input as the generation condition and in combination with the anchor representation bound to the unique identifier, so as to obtain a trusted identifier code and bind and store it with the asset file.

[0115] The clock linkage module is used to drive the dynamic update of the status of the running clock, verification clock and maintenance clock based on the trusted identifier encoding, and to trigger a compliance task when the threshold condition is reached, and record the result back to the power asset file to realize value update.

[0116] One embodiment of the present invention provides an intelligent identification management method for the entire life cycle of power assets. To verify the beneficial effects of the present invention, a simulation experiment is conducted for scientific demonstration.

[0117] The test subjects were 50 pole-mounted transformers (rated capacity 63~400kVA) on three 10kV feeders of a municipal power distribution network, which were continuously monitored for 8 weeks (56 days). A unique identifier was generated for each device and a file was created in the trusted data platform. File items included: design parameters (rated voltage, rated capacity, impedance, manufacturer, manufacturing date, etc.), operating status (phase current, phase voltage, oil temperature, ambient temperature, load factor, power factor, etc. sampled for 1 minute), and maintenance records (inspection, repair, fault and handling). To ensure data traceability, each on-site operation was completed by scanning the trusted identifier code with a mobile terminal to bind and upload the data. The platform integrated four types of systems: materials and equipment ledger system, operation and maintenance system, metering and verification system, and financial asset system, making the unique identifier the data primary key across systems.

[0118] Data modeling is performed according to the following steps: First, using a unique identifier as an index, the three types of data are aligned uniformly along the timeline, and multi-scale time windows (short-term 24h, medium-term 30d, long-term 180d) are set. Mean, variance, quantiles, limit violation rate, and load harmonic indicators are extracted, and topological neighborhood diffusion characteristics are superimposed (operational statistics of adjacent equipment on the same feeder are weighted by diffusion coefficients). Then, causal refinement is performed: a causal discovery process with directed acyclic graph constraints is used to perform structural learning and parameter estimation of the operational features. "Event-based intervention" (setting maintenance event counts or limit violation events to zero in the simulation) is used to verify directly impacting edges. The causal association result matrix and the feature representation after causal propagation are output, while the causal criticality of each dimension is recorded.

[0119] During the privacy protection phase, based on the platform's sensitive field rules (including geographic coordinates, fine-grained user-side load, personnel information, etc.), privacy topology factors are generated for the corresponding dimensions of the causal matrix. Masking coefficients are applied to sensitive dimensions, while retention coefficients are applied to key structures based on causal criticality, forming a weight set adjusted dimension by dimension. This factor is multiplied dimension by dimension with the features after causal propagation to obtain privacy-enhanced inputs, suppressing the influence of sensitive fields while maintaining topological criticality.

[0120] During the trusted identifier encoding generation phase, a conditional adversarial structure is configured: the generator network outputs candidate codes based on "privacy-enhanced input + anchored representation bound to a unique identifier"; the discriminator network performs consistency discrimination based on "encoding + condition" input. During training, binding constraints, privacy suppression constraints, structural compliance constraints, and stability constraints are superimposed. 80% of devices are used as the training set, and 20% as the validation set. An early stopping strategy avoids overfitting. The output trusted identifier code has a fixed length (96-bit equivalent, presented as a readable string on the platform) and is bound to the device profile.

[0121] During operation, three types of clocks are linked: the operating clock dynamically accumulates based on the stress factor of the operating condition (a comprehensive mapping of load rate and oil temperature); the calibration clock gradually accumulates based on the rate of measurement deviation (obtained by correcting historical calibration residuals with current temperature); and the warranty clock decreases in a countdown manner according to the warranty period. After setting thresholds, when any clock meets the conditions, the platform generates maintenance work orders, calibration tasks, or warranty processing orders according to compliance priority. After the work is completed, the mobile terminal scans the code to bind the result back to the archive, and the value is updated on the financial side. For comparison, a traditional control group is set up: using fixed QR code numbers (no adversarial generation), correlation-based statistical modeling (no causal refinement and privacy topology factors), and performing maintenance / calibration at fixed intervals (no linkage of the three clocks). The control group selects 50 devices of similar scale and load structure on the same network side, with the observation period consistent with the sampling.

[0122] Experimental data (key indicators, all statistically analyzed over a 56-day observation period) include: record consistency rate: 99.96% for this method and 97.84% for the control group.

[0123] Counterfeit / forgery interception rate: 99.60% for this method, 87.40% for the control group.

[0124] Number of misbinding events (times / 50 units): 0.00 for this method, 6.00 for the control group.

[0125] Re-identification (privacy attack) success rate (simulated link attack): 2.30% for this method, 21.50% for the control group.

[0126] Topological consistency score (0~1): 0.93 for this method, 0.72 for the control group.

[0127] Causal structure accuracy (based on the hit rate of the leave-out intervention set): 0.84 for this method and 0.61 for the control group.

[0128] On-time completion rate of compliance tasks: 98.60% for this method and 88.70% for the control group.

[0129] Average lead time for compliance triggering (days): 5.40 for this method, 2.10 for the control group.

[0130] Average overdue time (hours / task): 4.20 for this method, 36.80 for the control group.

[0131] Work order repetition rate: 1.10% in this method, 7.50% in the control group.

[0132] Average downtime (hours / unit / period): 9.20 for this method, 18.40 for the control group.

[0133] Monthly maintenance cost (RMB / unit / month): Method 1,164.50, control group 1,250.00.

[0134] Net effect of asset value update (RMB / unit / period, including risk reduction measurement): This method +1,380.00, control group +210.00.

[0135] Alarm-to-noise ratio (false alarms / total alarms): 6.40% for this method, 15.30% for the control group.

[0136] Write-back latency after task closure (hours, P50): 1.80 for this method, 12.60 for the control group.

[0137] Data shows that the methodology chain of "embedding and causal refinement → privacy topology factor → adversarial generative coding → three-clock linkage" produces synergistic effects across multiple dimensions. Firstly, the file consistency rate is improved to 99.96%, and the number of misbinding events is reduced to 0.00. This corresponds to the effects of steps S1-S2: using unique identifiers as primary keys to uniformly align multi-source data, and retaining the direct relationship between operational status and maintenance events after causal refinement, reducing erroneous merging caused by superficial correlations. The control group, using correlation statistics, cannot distinguish between the direct and indirect relationships of "temperature-load-maintenance," making mismatches more likely during data backtracking.

[0138] Secondly, the privacy attack success rate decreased to 2.30% and the topology consistency score was 0.93, demonstrating the effectiveness of step S3: the privacy topology factor is not a fixed mask, but rather dynamically adjusted based on causal propagation results, suppressing sensitive dimensions while preserving key topological structures. This solves the problem of traditional static masks destroying structural information, ensuring that subsequent generation and clock calculations still have structural interpretability. The control group used global noise addition or masking, which reduced the risk of leakage, but the damage to topological information led to increased noise in downstream tasks, resulting in an alarm-to-noise ratio of 15.30%, significantly higher than the 6.40% of this method.

[0139] Furthermore, the forgery / spoofing interception rate of 99.60% and the ticket duplication rate of 1.10% correspond to the contribution of step S4: the conditional adversarial generation binds the privacy-enhancing input with the anchored representation, continuously verifies the consistency of the discrimination network, and ensures that the generated trusted identifier encoding remains stable under attacks and environmental disturbances, thereby reducing false triggering and duplicate ticket creation during on-site identification. In contrast, the fixed QR codes in the control group are easily misread or forged under conditions of dirt, reflection, and duplication, leading to an increase in duplicate or erroneous tickets.

[0140] At the coordination level, the on-time completion rate of compliance tasks was 98.60%, the average lead time was 5.40 days, and the overdue time was 4.20 hours. This indicates that the three-clock mechanism in step S5 effectively maps operational stress, metering deviation, and warranty period into differentiated advancement logic. The stress-weighted operation clock allows equipment under high load and high temperature scenarios to receive maintenance opportunities earlier; the verification clock advances based on the deviation rate rather than a fixed cycle, allowing metering devices with faster drift to enter the verification queue first; and the warranty clock directly counts down the remaining period, reducing the situation where it is triggered after the expiration date. The control group used a fixed cycle, resulting in high-stress equipment "not having enough time" and low-stress equipment "over-maintaining," thus leading to the difference in results: an overdue time of 36.80 hours and an downtime of 18.40 hours.

[0141] The cost and value dimensions further validated the benefits of closed-loop management: monthly maintenance costs decreased from RMB 1,250.00 to RMB 1,164.50, mainly due to the reduction in duplicate work orders and false alarms, as well as the optimization of maintenance rhythm and resource scheduling; the net effect of asset value update was RMB 1,380.00 / unit / period, reflecting the positive impact of reduced risk and improved measurement reliability after verification on value assessment. The control group only saw an increase of RMB 210.00 / unit / period, indicating that the traditional process lacked sufficient linkage between "risk-compliance-value".

[0142] To further illustrate the role of key components, ablation analysis was conducted: After removing privacy topology factors, the re-identification success rate increased to 9.70%, and topology consistency decreased to 0.79; after removing adversarial generation, the forgery / spoofing interception rate decreased to 93.20%, and the ticket duplication rate increased to 4.80%; after removing the three-clock linkage, the compliance lead time decreased to 2.30 days, and the overdue time increased to 29.40 hours. The results show that all three are key components in forming the overall effect, and the absence of any one of them will result in significant degradation in privacy, trust, or compliance dimensions.

[0143] In summary, this embodiment, based on a unified identifier and a trusted platform, improves modeling accuracy through causal refinement in a real power distribution network scenario. It achieves a balance between protection and structural preservation through privacy topology factors, enhances identifier trustworthiness and stability through adversarial generation, and forms a closed loop of compliance and value through three-clock linkage. Data comparison shows that this method achieves substantial improvements in consistency, privacy protection, robust identification, compliance execution, and cost-effectiveness, demonstrating the engineering feasibility and comprehensive benefits for the full lifecycle management of power assets.

[0144] This embodiment also provides a computer device applicable to the intelligent identification management method for the entire life cycle of power assets, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent identification management method for the entire life cycle of power assets as proposed in the above embodiment.

[0145] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0146] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the intelligent identification management method for the entire life cycle of power assets as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0147] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A smart identification management method for the entire lifecycle of power assets, characterized in that: include: Collect lifecycle data of power assets, establish archives of power asset lifecycle data with unique identifiers, and store them in a trusted data platform; By linking power asset lifecycle data with unique identifiers, contextual information is formed based on embedded modeling, and causal relationship results are obtained through causal refinement. The formation of context information includes, in the trusted data platform, using the unique identifier of the power asset as an index, uniformly aligning the design parameters, operating status data and maintenance event records, extracting statistical operating features according to short-term, medium-term and long-term multi-scale time windows, and combining the time decay effect of maintenance events and the diffusion relationship of the power network topology neighborhood to generate a context embedding vector that reflects the historical status of equipment, event impact and network coupling. The process of obtaining causal association results through causal refinement includes: based on the context embedding vector, using a causal discovery method constrained by a directed acyclic graph, calculating the causal association relationships between various operational features, and forming a causal association result matrix. The causal association results can distinguish between direct causal effects and superficial correlations, and output the feature representation after causal propagation. Introducing privacy-preserving topological factors into causal association results to mask sensitive information while maintaining topological criticality creates privacy-enhanced inputs. The introduction of privacy topology factors in causal association results includes setting masking coefficients for corresponding dimensions in the causal association result matrix according to preset sensitive field rules in a trusted data platform, generating retention coefficients by combining the feature importance after causal propagation, and merging the masking coefficients and retention coefficients to obtain privacy topology factors. The privacy topology factors are a set of adjustment parameters set dimension by dimension, used to simultaneously reduce the influence of sensitive features in subsequent processing and keep the main causal relationship intact. The privacy-enhanced input includes multiplying the privacy topology factor with the feature representation after causal propagation in each dimension to obtain the privacy-preserving context input, which weakens the sensitive features in each dimension to generate privacy-enhanced input data. Adversarial generation is performed on privacy-enhanced inputs to obtain robust and trustworthy identifier encodings; Generating the trusted identifier code includes, in a trusted data platform, using the privacy-enhanced input as the generation condition, and combining it with the anchored representation bound to the unique identifier, configuring a generation network and a discriminator network, and performing conditional adversarial generation under the combined effect of binding constraints, privacy suppression constraints, structural compliance constraints and stability constraints to obtain a trusted identifier code that satisfies the constraints; The trusted identifier is encoded and bound to the asset file in the trusted data platform to drive subsequent full lifecycle management tasks; By leveraging trusted identifier encoding to drive the linkage of operating clocks, calibration clocks, and maintenance clocks, compliance tasks are triggered and the results are recorded back into the power asset archive, thereby achieving value updates.

2. The intelligent identification management method for the entire lifecycle of power assets as described in claim 1, characterized in that: The power asset lifecycle data includes the power asset's design parameters, operating status, and maintenance records; The process of establishing archives for power asset lifecycle data and unique identifiers includes generating a unique identifier for each power asset in a trusted data platform, collecting and archiving the power asset's design parameters, operating status, and maintenance records into corresponding archive entries, and binding and storing the unique identifier in the archive entries.

3. The intelligent identification management method for the entire lifecycle of power assets as described in claim 2, characterized in that: The linkage between the operating clock, calibration clock, and warranty clock driven by the trusted identifier code includes: extracting the operating stress factor and measurement deviation information based on the trusted identifier code and the corresponding context input in the trusted data platform; dynamically accumulating the operating clock according to the operating stress factor; gradually accumulating the calibration clock according to the measurement deviation information; and decreasing the warranty clock in a countdown manner according to the warranty period, so as to realize the dynamic update of the status of the operating clock, calibration clock, and warranty clock and maintain their continuous binding with the trusted identifier code.

4. The intelligent identification management method for the entire lifecycle of power assets as described in claim 3, characterized in that: The process of triggering compliance tasks and recording the results back to the power asset archive includes generating a trigger indication when the state of any clock reaches a preset threshold condition, executing task scheduling in the trusted data platform according to compliance priority based on the trigger indication, forming corresponding operation and maintenance work orders, inspection tasks or warranty processing orders, and binding and storing the operation record with the trusted identification code after the task is completed, while updating the status information and value information in the power asset archive to achieve closed-loop management of the entire life cycle.

5. A smart identification management system for the entire lifecycle of power assets, based on the smart identification management method for the entire lifecycle of power assets as described in any one of claims 1 to 4, characterized in that: The data acquisition module is used to collect design parameters, operating status and maintenance records of power assets, and to establish files with unique identifiers and bind and store them in a trusted data platform; The context modeling module is used to uniformly align the design parameters, running status data, and maintenance event records in the trusted data platform using a unique identifier as an index, and generate a context embedding vector. The causal refinement module is used to obtain the causal association result matrix and output the feature representation after causal propagation based on the context embedding vector and the causal discovery method constrained by the directed acyclic graph. The privacy topology factor construction module is used to generate privacy topology factors in a trusted data platform according to preset sensitive field rules, and combine them with the feature representation after causal propagation to form privacy-enhanced input; The adversarial generation module is used in the trusted data platform to perform conditional adversarial generation using the privacy-enhanced input as the generation condition and in combination with the anchor representation bound to the unique identifier, to obtain the trusted identifier code and bind and store it with the asset file. The clock linkage module is used to drive the dynamic update of the status of the running clock, verification clock and maintenance clock based on the trusted identifier encoding, and to trigger a compliance task when the threshold condition is reached, and record the result back to the power asset file to realize value update.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the intelligent identification management method for the entire life cycle of power assets as described in any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the intelligent identification management method for the entire life cycle of power assets as described in any one of claims 1 to 4.

Images