Method and apparatus for processing cloud device parameters, electronic device, and storage medium

Abstract

The present application relates to the technical field of electric power data processing, and provides a kind of regulation and control cloud equipment parameter processing method, device, electronic equipment and storage medium.The implementation scheme is: the nameplate image of the equipment to be collected is obtained;Text information in the nameplate image is recognized, and original parameter text is obtained;The original parameter text is structured and parsed, and target structured parameter is obtained;The target structured parameter is checked at multiple levels, and target check result is obtained;If the target check result is not checked, the abnormal parameter item in the target structured parameter is corrected, and the corrected target structured parameter is checked again;If the target check result is checked, the target structured parameter is determined as the target parameter of the equipment to be collected.The present application embodiment can improve the efficiency and accuracy of regulation and control cloud equipment data collection.

Description

Technical Field

[0001] This invention relates to the field of power data processing technology, and in particular to a method, apparatus, electronic device, and storage medium for processing parameters of cloud control equipment. Background Technology

[0002] In power system operation and maintenance, the dispatching cloud platform serves as the core carrier for power grid operation data management and analysis. The accuracy of its primary equipment parameters directly impacts the reliability of power grid dispatching decisions and the efficiency of new equipment commissioning. Currently, over 95% of the primary equipment parameters in the dispatching cloud platform originate from equipment nameplates, making the collection and entry of equipment nameplate information a crucial step before new equipment is commissioned.

[0003] Traditional methods for collecting parameters from cloud-based control equipment typically involve manually photographing the equipment nameplates, then having maintenance personnel manually identify and transcribe the parameter information from the nameplate images before inputting it into the control cloud system. This method not only relies on human experience and requires high accuracy in identifying equipment models, parameter units, and technical terminology, but also, in scenarios with a large number of devices and complex nameplate content, manual entry is prone to omissions, errors, and inconsistent formats, making it difficult to guarantee the accuracy of the parameter data.

[0004] Meanwhile, the manual data collection and entry process involves multiple steps such as shooting, sorting, identification, and verification, and the overall process is time-consuming, making it difficult to meet the demand for rapid data entry when new equipment is put into operation or when large-scale equipment parameters are updated.

[0005] Therefore, how to improve the efficiency and parameter accuracy of the data acquisition process of cloud devices is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] This invention provides a method, apparatus, electronic device, and storage medium for regulating cloud device parameters, which can solve at least one of the above-mentioned technical problems.

[0007] In a first aspect, embodiments of the present invention provide a method for controlling cloud device parameter processing, including: Acquire the nameplate image of the device to be acquired; The text information in the nameplate image is identified to obtain the original parameter text; The original parameter text is parsed in a structured manner to obtain the target structured parameters; The target structured parameters are validated at multiple levels to obtain the target validation results. If the target verification result is that the verification fails, the abnormal parameter items in the target structured parameters are corrected and the corrected target structured parameters are verified a second time. If the target verification result is successful, then the target structured parameters are determined as the target parameters of the device to be acquired.

[0008] Secondly, embodiments of the present invention provide a device for regulating cloud device parameter processing, comprising: The acquisition module is used to acquire the nameplate image of the device to be acquired. The recognition module is used to recognize the text information in the nameplate image to obtain the original parameter text; The structured parsing module is used to perform structured parsing on the original parameter text to obtain the target structured parameters; A multi-level verification module is used to perform multi-level verification on the target structured parameters to obtain the target verification result; The correction module is used to correct the abnormal parameter items in the target structured parameters and perform a second verification on the corrected target structured parameters if the target verification result is that the verification fails. The target parameter determination module is used to determine the target structured parameters as the target parameters of the device to be acquired if the target verification result is a successful verification.

[0009] Thirdly, embodiments of the present invention also provide an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in any one of the embodiments of the present invention.

[0010] Fourthly, embodiments of the present invention also provide a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method described in any one of the embodiments of the present invention.

[0011] By employing the technical solution of this invention, the traditional parameter acquisition process, which relies on manual transcription, is transformed into a machine recognition process by acquiring the nameplate image of the device to be collected and automatically recognizing the text information within the nameplate image. This reduces manual intervention, shortens the overall process time for parameter acquisition and input, and improves the efficiency of device data acquisition. Furthermore, by performing structured parsing on the recognized raw parameter text, target structured parameters are obtained. This allows parameters from different devices to be organized and output in a unified format, avoiding formatting issues and omissions that occur during manual organization, thus improving the consistency and accuracy of parameter processing at the structural level. Moreover, by performing multi-level verification on the target structured parameters, abnormal parameter items can be identified before the parameters are entered into the database, reducing the probability of erroneous parameters directly entering the control cloud system, thereby improving the reliability and accuracy of parameter data. When the verification result is negative, the abnormal parameter items in the target structured parameters are corrected, and the corrected target structured parameters are verified a second time. This achieves closed-loop processing and repeated confirmation of the problematic parameters, avoiding the risk of omissions caused by a single verification. When the verification result is positive, the target structured parameters are directly determined as the target parameters of the device to be collected, achieving automatic confirmation and storage of the parameters. In this way, the embodiments of the present invention can improve the efficiency and accuracy of data collection from cloud devices.

[0012] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0013] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of the invention. Wherein: Figure 1 This is a flowchart of a method for processing cloud device parameters according to an embodiment of the present invention; Figure 2 This is a structural block diagram of a cloud device parameter processing device according to an embodiment of the present invention; Figure 3 This is a schematic block diagram of an electronic device used to implement the methods of embodiments of the present invention. Detailed Implementation

[0014] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0015] Figure 1This is a flowchart of a method for processing parameters of a cloud device according to an embodiment of the present invention.

[0016] like Figure 1 As shown, the method for processing parameters of the cloud device can include: S110, acquire the nameplate image of the device to be acquired; S120, Recognize the text information in the nameplate image to obtain the original parameter text; S130, perform structured parsing on the original parameter text to obtain the target structured parameters; S140, Perform multi-level verification on the target structured parameters to obtain the target verification result; S150, If the target verification result is that the verification fails, the abnormal parameter items in the target structured parameters are corrected and the corrected target structured parameters are verified a second time. S160, if the target verification result is that the verification is passed, then the target structured parameters are determined as the target parameters of the device to be acquired.

[0017] For example, the device to be acquired may include, but is not limited to, the following primary equipment: transformers (such as 110 kV main transformers); circuit breakers (such as... Circuit breakers; disconnecting switches; current transformers, voltage transformers; AC contactors or high-voltage switchgear. For example, the equipment to be collected is a 110kV transformer, whose nameplate is marked with information such as "rated capacity, rated voltage, rated current, impedance voltage, manufacturer, and serial number".

[0018] In this example, on-site maintenance personnel use mobile smart terminal devices (such as industrial tablets or smartphones) to point at the physical nameplate of the equipment to be acquired. The dedicated data acquisition application on the terminal device calls upon the built-in camera module to perform one or more image captures, ensuring image clarity, suitable lighting, and complete nameplate content. After capturing the images, the application compresses and encodes the generated digital image files and temporarily stores them on local storage for later use.

[0019] For example, step S110, which involves obtaining the nameplate image of the device to be collected, includes one or more of the following methods: First, the system obtains a device list file containing information such as device name, device type, and voltage level from a preset management module. This method allows the system to generate information on the entire station or multiple intervals of devices to be collected at once, significantly improving task creation efficiency. Second, after receiving the interval number information uploaded by the task executor's terminal, the system automatically associates it with a preset interval device template and generates standard-configured device information based on the template, quickly completing the collection task construction. Third, the system automatically generates a corresponding non-standard device collection information task based on the custom device information fields uploaded by the task executor's terminal. Subsequently, based on the task executor's geographical location information and the geographical range corresponding to the task, the system automatically pushes the collection task to the task executor's terminal, enabling the task executor to capture the nameplate image of the device to be collected and upload it to the system. After receiving the device nameplate image uploaded by the task executor, the system automatically associates the image with the corresponding device information to obtain the nameplate image of the device to be collected.

[0020] For example, the original parameter text refers to the parameter content text obtained by recognizing the text information in the nameplate image and presented in natural language or a non-standard structure.

[0021] For example, the nameplate displays "Model: SZ11-40000 / 110; Rated capacity: 40000 kVA; High voltage side voltage: 110 kV; Low voltage side voltage: 35 kV"; the original parameter text obtained after text recognition can be "Model SZ11-40000 / 110 Rated capacity 40000 kVA High voltage side voltage 110 kV Low voltage side voltage 35 kV".

[0022] In this example, the system takes the acquired nameplate image as input and sends it to an integrated Optical Character Recognition (OCR) engine. This OCR engine first preprocesses the image, including grayscale conversion, binarization, noise filtering, and tilt correction, to enhance the recognizability of the text regions. Then, the engine uses its deep learning model to segment and recognize the characters in the image, and reassembles all the recognized characters into a plain text string according to their spatial layout and reading order in the original image. This string, which completely reproduces all the text content of the nameplate without any semantic processing, is the original parameter text.

[0023] For example, target structured parameters refer to a set of parameter data with fixed fields and hierarchical relationships formed after semantic parsing, parameter attribute extraction, and hierarchical encapsulation of the original parameter text.

[0024] In this example, the system applies a domain-knowledge-based Natural Language Processing (NLP) module to perform deep parsing of the original parameter text. This module first uses a pre-built semantic feature library of power equipment parameters to extract key information from the text, identifying parameter attributes such as "model" and "rated capacity" and their corresponding parameter states such as "SFP-180000 / 220" and "180000kVA," and establishing mapping relationships. Next, based on a pre-defined parameter classification framework and the logical relationships parsed from the text, these mapped parameters are hierarchically categorized and their subordinate relationships are defined. Finally, following a standard hierarchical format, all categorized parameter information is structured and arranged to generate a structured data object with clear hierarchy and explicit semantics; this is the target structured parameter.

[0025] For example, the final generated target structured parameters are in JavaScript Object Notation (JSON) format: {"Basic Equipment Information": {"Equipment Model": "SFP-180000 / 220"}, "Electrical Performance Parameters": {"Rated Capacity": "180000kVA", "Rated Voltage": "220kV / 10.5kV", "Insulation Class": "Class A"}}.

[0026] For example, the target verification result refers to the comprehensive judgment result obtained after performing compliance verification, correlation verification and scenario adaptation verification on the target structured parameters in sequence.

[0027] For example, target parameters refer to the data set selected and determined as the final valid equipment parameters from the target structured parameters when the target verification result is successful, and are used for storage or retrieval. For instance, when the target verification result is successful, the final determined target parameters may be: Equipment type: transformer; Model: SZ11-40000 / 110; Rated capacity: 40000kVA; High-voltage side rated voltage: 110kV; Low-voltage side rated voltage: 35kV; These target parameters can be directly written into the equipment ledger of the control cloud system for scheduling analysis or operation and maintenance management.

[0028] In this example, the target structured parameters undergo multi-level validation, and the results of each level are recorded, ultimately summarizing to form a target validation result containing the pass / fail status of all validation items. When the target validation result is a failure, the system first integrates all marked-failure abnormal parameter items into a validation issue list. Then, this list is fed back to the on-site maintenance personnel or relevant execution entities through the user interface, requesting manual verification and correction of the abnormal parameter items. Upon receiving the corrected parameters submitted by the execution entity, the system replaces the old abnormal parameter items with the new parameters and re-executes the entire multi-level validation process of step S140 for the updated complete target structured parameters; this is the secondary validation. This feedback correction and secondary validation cycle can continue until all parameters pass validation. When the target validation result, whether after the initial validation or after secondary or multiple validations, finally reaches the status of "validation passed," the system determines that the current version of the target structured parameters is accurate, complete, and logically consistent. The system then terminated the verification process and officially designated this high-quality structured data object, which had passed all verifications, as the target parameter for the device to be collected, which can be used by the cloud platform for data entry.

[0029] According to the above implementation method, by acquiring the nameplate image of the device to be collected and automatically recognizing the text information in the nameplate image, the original parameter text is obtained, realizing the automated conversion of device parameters from image to text, avoiding the inefficiency and error-prone problems caused by manual copying. Based on this, the original parameter text is structured and parsed to obtain target structured parameters with a unified hierarchy and standardized format, expressing the originally scattered and disordered nameplate parameters in a structured form, facilitating subsequent processing and storage. Subsequently, the target structured parameters undergo multi-level verification, checking the parameters from multiple dimensions such as basic compliance, logical relationships between parameters, and application scenario adaptability, obtaining target verification results, thereby identifying abnormal parameter items before the parameters are entered into the database. When the target verification result is unsuccessful, the abnormal parameter items in the target structured parameters are corrected, and the corrected target structured parameters are verified a second time, realizing closed-loop correction and review of problematic parameters; when the target verification result is successful, the target structured parameters are directly determined as the target parameters of the device to be collected. Thus, this invention constructs an automated processing flow from nameplate image acquisition to parameter confirmation output, which not only reduces manual intervention and shortens the equipment parameter acquisition and processing cycle, but also reduces the parameter entry error rate through multi-level verification and secondary verification mechanisms, thereby effectively improving the efficiency and accuracy of data acquisition for control cloud equipment.

[0030] In one implementation, the original parameter text is subjected to structured parsing to obtain target structured parameters, including: semantically decomposing the original parameter text to obtain multiple key semantic segments; for each key semantic segment, calculating the similarity between parameter attributes in the key semantic segment and parameter attributes in a preset feature library to obtain attribute similarity values; matching parameter states in the key semantic segments with parameter states in the feature library to obtain state similarity values; extracting parameter attributes and parameter states from each key semantic segment based on each attribute similarity value and each state similarity value to obtain each target parameter attribute and each target parameter state; mapping each target parameter attribute and each target parameter state based on a preset mapping relationship between parameter attributes and parameter states to obtain each target mapping relationship; classifying each target parameter attribute into subordinate relationships to obtain each main parameter category and at least one secondary parameter item corresponding to the main parameter category; and structurally encapsulating each target mapping relationship, each main parameter category, and at least one secondary parameter item corresponding to the main parameter category to obtain target structured parameters.

[0031] For example, a key semantic fragment refers to the smallest unit of information obtained by decomposing the original parameter text according to semantic logic. Each fragment contains explicit parameter attributes (e.g., rated voltage) and state information (e.g., 220kV / 10.5kV) for subsequent attribute and state similarity calculations.

[0032] In this example, the system first applies a text segmentation algorithm based on natural language processing to the original parameter text. This algorithm divides the continuous long text into multiple independent, shorter text segments based on punctuation marks, line breaks, and predefined parameter delimiters (such as semicolons or spaces). Next, the system performs a preliminary semantic content assessment on these segments, filtering out those that explicitly contain parameter description information and discarding irrelevant decorative text or blank lines. These smallest units of information capable of carrying independent parameter information are the key semantic segments.

[0033] For example, the original parameter text is "Model: SFP-180000 / 220 Rated Capacity: 180000kVA Rated Voltage: 220kV / 10.5kV". By recognizing colons and spaces, it can be broken down into several key semantic segments, such as: Key Semantic Segment 1: "Model: SFP-180000 / 220"; Key Semantic Segment 2: "Rated Capacity: 180000kVA"; Key Semantic Segment 3: "Rated Voltage: 220kV / 10.5kV".

[0034] For example, for each key semantic segment, the system retrieves the corresponding parameter attributes from a preset parameter feature library, calculates the similarity between the parameter attributes in the key semantic segment and the parameter attributes in the feature library, and obtains the attribute similarity value for each key semantic segment. The similarity calculation method can employ word vector cosine similarity.

[0035] For example, for each key semantic segment, the system obtains the parameter state within it and matches or calculates numerical similarity with the corresponding parameter state in the feature library to obtain a state similarity value. The numerical similarity calculation can be based on text description, range, or numerical deviation assessment. For instance, if the state similarity value for the temperature (parameter state) in the feature library's state description "high-temperature operating range 60~85 degrees Celsius" is 0.95, then after comparison, the key semantic segment "temperature 60~80 degrees Celsius" falls within this range, and the state similarity value for the key semantic segment "temperature 60~80 degrees Celsius" is 0.95.

[0036] It should be noted that a key semantic segment may include one or more parameter attributes. Since parameter attributes correspond one-to-one with parameter states, a key semantic segment also includes one or more parameter states.

[0037] For example, for a key semantic segment containing multiple parameter attributes and corresponding parameter states, after obtaining the similarity values ​​of each attribute and each state corresponding to the key semantic segment using the method described in the previous example, the parameter corresponding to the maximum value among various similarity values ​​is extracted as the target object. For example, for the attribute similarity values ​​corresponding to the key semantic segment, the parameter attribute corresponding to the largest attribute similarity value is extracted as the target parameter attribute. As another example, a semantic hierarchical decomposition of a certain parameter text yields the key semantic segment "rated capacity," while the parameter attributes in the feature library include ["rated capacity," "rated voltage," "rated current," "capacity"]. Assume that through calculation, the attribute similarity values ​​between "rated capacity" in the key semantic segment and each parameter attribute in the feature library are 1.00, 0.70, 0.70, and 0.65, respectively. Thus, "rated capacity" is extracted to obtain the target parameter attribute "rated capacity."

[0038] Similarly, the parameter state corresponding to the largest state similarity value is extracted from the key semantic segment and used as the target parameter state to obtain the target parameter attribute and target parameter state corresponding to the key semantic segment.

[0039] Understandably, calculating similarity values ​​and extracting parameters (parameter states and attributes) based on these values ​​is equivalent to confirming the parameters in key semantic segments. This avoids the subsequent encapsulation and verification operations targeting non-parametric information, thereby improving the efficiency of subsequent data processing. Furthermore, even if the original parameter text identified in S110 has flaws, the correct target structured parameters can still be obtained through this step.

[0040] For example, the target mapping relationship refers to the mapping relationship obtained by mapping the extracted target parameter attributes and target parameter states based on the preset mapping relationship between parameter attributes and parameter states.

[0041] In this example, based on the inherent relationship between corresponding parameter attributes and parameter states in the feature library, all target parameter attributes and target parameter states are paired and mapped. This pairing process establishes a direct mapping from parameter attributes to their corresponding parameter states. All these successfully paired combinations are stored as a set containing multiple key-value pairs (key is the parameter attribute, value is the parameter state), which is the target mapping relationship obtained in this parsing. For example, "Rated Capacity" → "180000kVA"; "Rated Voltage" → "220kV / 10.5kV"; "Equipment Model" → "SFP-180000 / 220".

[0042] For example, a primary parameter category refers to a high-level classification of target parameters according to their attributes or functions, used for overall structured organization. Each primary parameter category may contain at least one secondary parameter item. A secondary parameter item refers to a specific parameter instance further subdivided under a primary parameter category, including its attributes and status information, reflecting the refined characteristics of the parameter.

[0043] In this example, the attributes of each target parameter are categorized into subordinate relationships, resulting in each main parameter category and at least one corresponding secondary parameter item. The subordinate relationships of target parameter attributes such as {"rated capacity", "rated voltage", and "equipment model"} are categorized. "Equipment basic information" and "electrical performance parameters" are identified as the main parameter categories, with "equipment model" belonging to the former, and "rated capacity" and "rated voltage" belonging to the latter.

[0044] Using the primary parameter category as the top-level node, its corresponding secondary parameter items as the next-level nodes, and the target parameter state corresponding to the secondary parameter items as the leaf nodes, a hierarchical tree-like data structure is constructed. This tree-like data structure is the target structured parameter obtained after structured encapsulation. For example, by encapsulating all information in a structured manner according to the hierarchy of "primary category, secondary item, state," a target structured parameter in JSON format can be generated.

[0045] For example, {"Basic Equipment Information": {"Equipment Model": "SFP-180000 / 220"}, "Electrical Performance Parameters": {"Rated Capacity": "180000kVA", "Rated Voltage": "220kV / 10.5kV"}}.

[0046] According to the above implementation method, the original parameter text is first semantically decomposed to obtain multiple key semantic segments, transforming the parameter information from continuous natural language text into independently processable semantic units. Then, for each key semantic segment, the similarity between its parameter attributes and those in a preset feature library, as well as the similarity between its parameter state and those in the feature library, are calculated. Based on the obtained attribute similarity and state similarity values, parameter attributes and states are automatically extracted from the key semantic segments to obtain each target parameter attribute and each target parameter state, thereby achieving accurate identification of the semantic elements of the nameplate parameters. On this basis, using the preset mapping relationship between parameter attributes and parameter states, each target parameter attribute and each target parameter state is matched and mapped to obtain each target mapping relationship, establishing a clear correspondence between the parameter meaning and its specific value. Further, the subordinate relationships of each target parameter attribute are divided to obtain the main parameter category and its corresponding at least one secondary parameter item, thereby constructing a parameter hierarchy structure that conforms to business logic. Finally, each target mapping relationship, each main parameter category, and its corresponding at least one secondary parameter item are structurally encapsulated to generate target structured parameters.

[0047] Thus, this invention can transform the original parameter text from an unstructured description into structured parameter data with clear hierarchy and semantics. This not only avoids the process of manually understanding and organizing parameters item by item, but also improves the accuracy of parameter attribute and parameter state recognition through similarity matching and mapping relationship constraints.

[0048] In one implementation, the hierarchical classification of each target parameter attribute is performed to obtain each main parameter category and at least one secondary parameter item corresponding to each main parameter category. This includes: hierarchically classifying each target parameter attribute based on its label to obtain first-level classification results, wherein each first-level classification result includes at least one secondary candidate parameter item corresponding to each candidate main parameter category; and extracting each first-level classification result based on at least one secondary candidate parameter item corresponding to each candidate main parameter category and a preset judgment condition to obtain second-level classification results, wherein the judgment condition is that each secondary candidate parameter item satisfies the functional requirements of the candidate main parameter category. Based on the preset parameter storage architecture and data retrieval rules of the control cloud system, the adaptability of each second-level partitioning result is verified; the second-level partitioning results that are not conducive to parameter management or fast access are adjusted to obtain the third-level partitioning results; the second-level partitioning results that are conducive to parameter management or fast access are determined as the fourth-level partitioning results; the fourth-level partitioning results and the third-level partitioning results are verified, and the candidate main parameter categories and at least one secondary candidate parameter item corresponding to each candidate main parameter category are determined as the main parameter category and at least one secondary parameter item corresponding to the main parameter category, respectively.

[0049] For example, firstly, for each target parameter attribute, a label identifying its category level (e.g., a 'category' label or a 'sub-item' label) is extracted from the database. Then, all target parameter attributes with the 'category' label are identified as candidate primary parameter categories. Next, all target parameter attributes with the 'sub-item' label are iterated over, and the similarity value between the 'sub-item' label target parameter attribute and each 'category' label target parameter attribute is calculated. Target parameter attributes with 'sub-item' labels whose similarity values ​​exceed a preset similarity threshold are classified under the corresponding 'category' label. The resulting complete hierarchical structure, containing all candidate primary parameter categories and their initially assigned secondary candidate parameter sub-items, is the first-level partitioning result.

[0050] For example, we obtain a set of target parameter attributes with labels: {"Electrical Parameters": category, "Equipment Basic Information": category, "Rated Capacity": sub-item, "Rated Voltage": sub-item, "Equipment Model": sub-item}. First, "Electrical Parameters" and "Equipment Basic Information" are identified as candidate primary parameter categories. Then, through similarity calculation and analysis, "Rated Capacity" and "Rated Voltage" are initially assigned to "Electrical Parameters," and "Equipment Model" is assigned to "Equipment Basic Information," forming the first-level classification result.

[0051] For example, a functional requirement compliance check is performed on each candidate primary parameter category and its subordinate secondary candidate parameter items in the first-level partitioning result. This check is based on a preset judgment condition, namely, whether each secondary candidate parameter item is a necessary indicator to meet the core functional requirements of its respective candidate primary parameter category. All hierarchical structures that pass this check and have clear functional dependencies are retained. These hierarchical structures retained after functional requirement verification constitute the second-level partitioning result.

[0052] For example, the first-level classification results are validated. The judgment condition is that "the sub-item must be a necessary indicator to meet the functional requirements of the main category". "Equipment Model" meets the "Identifying Equipment" functional requirement of "Equipment Basic Information", and this relationship passes the validation. Both "Rated Capacity" and "Rated Voltage" meet the "Describing Electrical Behavior" functional requirement of "Electrical Parameters", and this relationship also passes the validation. Suppose there is an incorrect classification, such as "Insulation Class" being classified under "Equipment Basic Information", this relationship will be removed because it does not meet the functional requirements. Finally, all the validated hierarchical relationships are extracted to form the second-level classification results.

[0053] For example, the second-level partitioning result is compared with the preset cloud parameter storage architecture and data retrieval rules to perform an adaptation check. This check mainly examines whether the partitioning result meets the requirements for parameter management efficiency and call performance. For example, whether there are too many sub-items under a certain main parameter category, or whether a frequently used parameter is placed in too deep a level. For hierarchical structures determined during the check to be detrimental to parameter management or fast call performance, an adjustment procedure will be initiated. For example, by merging similar sub-items or promoting the hierarchical position of frequently used items, optimization will be performed. The new hierarchical structure obtained after this adjustment is the third-level partitioning result.

[0054] For example, an adaptation check is performed on the second-level classification results. The check rules include "no more than 10 sub-items under the main category" and "the number of levels for commonly used parameters should not exceed 2." Suppose that "Electrical Parameters" is initially divided into 12 sub-items, and "Rated Voltage" is incorrectly placed in the third level. This is determined to be detrimental to parameter management. Therefore, an adjustment is performed, merging two similar sub-items and promoting "Rated Voltage" to the second level. The adjusted new structure is the third-level classification result.

[0055] For example, firstly, all portions of the second-level partitioning results that pass the adaptability verification are identified as fourth-level partitioning results. Then, these fourth-level partitioning results are merged with the adjusted third-level partitioning results from the previous step to form a candidate hierarchical relationship scheme for final verification. This candidate scheme is then compared and verified against a reference benchmark that includes industry technical standards and historical practice cases. Only hierarchical relationships that meet both the requirements and general industry logic are ultimately confirmed. This finally confirmed hierarchical structure represents the final determined categories of primary parameters and their corresponding secondary parameter items. For example, the structure "Equipment Basic Information: Equipment Model" in the second-level partitioning results is deemed beneficial for parameter management during adaptability verification and is therefore identified as a fourth-level partitioning result.

[0056] The fourth-level classification results are merged with the third-level classification results (the adjusted electrical parameters section) to form a complete candidate scheme, which is then subjected to final verification. For example, a comparison with power industry standards confirms that classifying "rated voltage" under "electrical parameters" conforms to industry specifications. Finally, after all verifications pass, the complete hierarchical relationship is determined: equipment basic information includes equipment model, and electrical parameters include rated voltage, etc.

[0057] According to the above implementation method, firstly, based on the labels of each target parameter attribute, the target parameter attributes are hierarchically divided to form a first-level division result containing candidate primary parameter categories and their corresponding at least one secondary candidate parameter item, thereby achieving preliminary clustering of parameter attributes. Subsequently, based on the judgment condition that "all secondary candidate parameter items meet the functional requirements of the candidate primary parameter category", the first-level division result is filtered and extracted to obtain the second-level division result, so as to ensure that parameter items under the same candidate primary parameter category have consistency and intrinsic correlation in functional semantics. On this basis, combined with the preset parameter storage architecture and data retrieval rules of the control cloud system, the second-level division result is subjected to adaptability verification, and the division results that are not conducive to parameter management or fast access are adjusted to obtain the third-level division result. At the same time, the second-level division result whose adaptability verification result is conducive to parameter management or fast access is directly determined as the fourth-level division result. Finally, the results of the third and fourth level partitioning are uniformly verified. The verified candidate primary parameter categories and their corresponding at least one secondary candidate parameter item are determined as the final primary parameter category and at least one secondary parameter item corresponding to the primary parameter category, respectively. This ensures that the parameter classification results simultaneously meet the storage structure and retrieval rule requirements of the cloud system while maintaining semantic consistency of parameter functions, avoiding the hierarchical confusion or decreased retrieval efficiency problems caused by relying solely on semantic clustering.

[0058] In one implementation, the fourth-level and third-level partitioning results are verified, and each candidate primary parameter category and at least one secondary candidate parameter item corresponding to each candidate primary parameter category that passes verification are respectively determined as each primary parameter category and at least one secondary parameter item corresponding to each primary parameter category. This includes: determining the fifth-level partitioning result based on each fourth-level and third-level partitioning result; comparing and verifying the fifth-level partitioning result based on a preset hierarchical partitioning reference benchmark; if the verification fails, adjusting at least one secondary candidate parameter item corresponding to each candidate primary parameter category in the fifth-level partitioning result that does not match the hierarchical partitioning reference benchmark; if the verification passes, determining each candidate primary parameter category and at least one secondary candidate parameter item corresponding to each candidate primary parameter category in the fifth-level partitioning result as each primary parameter category and at least one secondary parameter item corresponding to each primary parameter category.

[0059] For example, the two hierarchical partitioning results generated in the preceding steps are aggregated and merged. The fourth-level partitioning result represents the set of hierarchical relationships that do not require adjustment during adaptability verification; the third-level partitioning result represents the set of hierarchical relationships after adaptability adjustment. These two sets are then combined to form a complete candidate hierarchical relationship scheme that includes all target parameter attributes and whose hierarchical structure has initially met internal requirements. This scheme is the fifth-level partitioning result.

[0060] For example, a hierarchical classification reference benchmark is retrieved, which includes power industry technical standards and specifications as well as historical application cases of similar equipment. Then, each master-slave hierarchical relationship in the fifth-level classification result is traversed and compared with the general logic defined in the reference benchmark. This comparison process aims to verify whether the hierarchical classification in the candidate scheme conforms to common industry practices and technical specifications, and to generate a "verification passed" or "verification failed" judgment result for each hierarchical relationship.

[0061] For example, the hierarchical classification reference standard includes a rule: "The 'insulation class' of the transformer should belong to the 'electrical parameters' category." During the comparison and verification of the fifth-level classification results, it was found that the scheme incorrectly classified "insulation class" under the "mechanical parameters" category. Therefore, the verification result for this hierarchical relationship of "insulation class" is "verification failed."

[0062] For example, in the previous comparison and verification step, all hierarchical relationships that were judged to "fail verification" are removed to adjust them. Specifically, a sub-candidate parameter item that was misclassified is removed from its current candidate primary parameter category.

[0063] For example, after all adjustment operations are completed, or if all hierarchical relationships are determined to be "verified" in the initial comparison and verification, the current hierarchical division result is considered accurate. Then, all candidate primary parameter categories and secondary candidate parameter items in the result are determined as the final primary parameter categories and their corresponding secondary parameter items, thereby completing the entire subordinate relationship division process.

[0064] It should be noted that for each candidate primary parameter category, and at least one secondary candidate parameter item corresponding to that primary parameter category, the steps given in the previous example are processed to obtain each primary parameter category and at least one secondary parameter item corresponding to that primary parameter category.

[0065] According to the above implementation method, by uniformly summarizing the results of each fourth-level division and the results of the third-level division, a fifth-level division result is formed. This allows for a comprehensive consideration of the results after adaptation adjustment and the results that directly pass the adaptation verification within the same hierarchical framework. Subsequently, the fifth-level division result is compared and verified based on a preset hierarchical division reference benchmark to determine whether the master-slave relationship between the candidate primary parameter categories and their corresponding secondary candidate parameter items conforms to industry-standard logic and practical feasibility. When the verification result is unsuccessful, at least one secondary candidate parameter item corresponding to the candidate primary parameter category in the fifth-level division result that does not match the hierarchical division reference benchmark is specifically adjusted so that its master-slave relationship gradually approaches the reasonable hierarchical structure defined by the reference benchmark. When the verification result is successful, each candidate primary parameter category and its corresponding at least one secondary candidate parameter item in the fifth-level division result are determined as the final primary parameter category and at least one secondary parameter item corresponding to the primary parameter category, respectively. In this way, structural biases introduced by relying solely on the results of previous automatic partitioning can be avoided, and the parameter hierarchy can meet the system architecture requirements while further conforming to industry standards and historical application experience, thereby improving the accuracy and consistency of the partitioning results of primary parameter categories and secondary parameter sub-items.

[0066] In one implementation, multi-level verification is performed on the target structured parameters to obtain a target verification result, including: performing compliance verification on the target structured parameters to obtain a compliance verification result; performing correlation verification on the target structured parameters based on preset logical constraints to obtain a correlation verification result; performing adaptability verification on the target structured parameters based on preset application scenario requirements to obtain a scenario verification result; if the compliance verification result, correlation verification result, and scenario verification result are all passed, then the target verification result is determined to be verified as passed; if any one of the compliance verification result, correlation verification result, and scenario verification result is failed, then the target verification result is determined to be verified as failed.

[0067] For example, the target structured parameter to be verified is compared with a preset parameter baseline specification. This verification process mainly covers three dimensions: First, checking the parameter's expression format to determine whether its name, unit, etc., conform to the standard terminology of the power industry; second, checking the data integrity to confirm whether any required parameter items are missing; and third, checking the compliance of the basic format, such as whether the decimal places of numerical parameters and the format of date parameters meet the system's storage requirements. All parameter items that do not conform to the baseline specification are marked, and finally, a compliance verification result containing the verification status of all parameter items is generated.

[0068] For example, during a compliance check of the target structured parameters of a transformer, it was found that the unit of "rated voltage" was incorrectly recorded as "V" instead of the standard "kV," and the required field "rated capacity" was missing. Therefore, these two parameter items were marked as failing, and a compliance check result specifying the exact problem was generated.

[0069] For example, the correlation verification process is initiated only for parameter items that pass the compliance check. This process invokes a set of preset logical constraints. Based on these constraints, the system cross-checks multiple technically related parameter items in the target structured parameters to determine whether there are any logical contradictions or numerical conflicts among them.

[0070] In this example, logical constraints, such as those for rated capacity, rated voltage, and rated current, are set as follows: In the formula, This is the theoretical rated capacity; Rated voltage; The rated current is I; at the same time, the rated current I should not exceed the maximum permissible current value derived from the rated capacity and rated voltage.

[0071] When the target structured parameters include a rated capacity of 1000kVA, a rated voltage of 10kV, and a rated current of 50A, the system calculates a theoretical rated capacity of 866kVA based on logical constraints. This is inconsistent with the input rated capacity of 1000kVA and exceeds the preset error threshold. Therefore, it is determined that there is a numerical conflict between the parameters and the combination of "rated capacity, rated voltage, and rated current" is marked as failing the verification.

[0072] For example, for parameters that have passed the first two levels of verification, the highest level of compatibility verification is performed. This process will consider the specific device operating scenario corresponding to this data collection task, the application requirements of the control cloud system, and historical parameter application cases of similar devices to determine the rationality of the parameter values.

[0073] For example, the system will compare the value of a parameter with the standard value range of its voltage level or equipment type. For parameters whose compatibility is questionable, the system will trigger a review process and ultimately mark all parameter items that are determined to be inconsistent with the actual application scenario, forming a scenario verification result.

[0074] For example, the system performs a compatibility check on the target structured parameters of a circuit breaker in a 110kV substation. The system finds that its "rated breaking current" is recorded as 2kA, but according to application scenario requirements and historical cases, this parameter for 110kV circuit breakers is usually no less than 20kA. Because this value deviates significantly from the reasonable range, the system marks it as "verification failed" and generates the scenario verification result.

[0075] For example, a final logical AND operation is performed on all the verification results in the aforementioned example. Specifically, the compliance verification result, the correlation verification result, and the scenario verification result are checked to see if any parameter item is marked as failing. If the check finds that all parameter items are in the "verification passed" state in all three levels of verification, the target verification result is determined to be "verification passed". Conversely, if the check finds that any parameter item is marked as "verification failed" in any of the three levels of verification, the target verification result is determined to be "verification failed".

[0076] According to the above implementation method, the target structured parameters are comprehensively verified from three dimensions: parameter standardization, logical consistency between parameters, and matching with the actual application scenario by sequentially performing compliance verification, correlation verification, and adaptability verification on the target structured parameters. When the compliance verification result, correlation verification result, and scenario verification result are all passed, the target verification result is determined to be passed, thereby ensuring that the target structured parameters meet the system's preset requirements in terms of format, content, and application. When any of the above verification results is failed, the target verification result is determined to be failed, to avoid parameters with missing, conflicting, or unreasonable values ​​being directly used. By constructing a multi-level verification system composed of compliance verification, correlation verification, and adaptability verification, parameter data that does not conform to the specifications or has logical vulnerabilities can be identified and intercepted before the parameters are entered into the database or called. This not only improves the accuracy and reliability of the target structured parameters, but also reduces the risk caused by parameter errors in subsequent control and analysis stages.

[0077] In one implementation, if the target verification result is that the verification fails, the abnormal parameter items in the target structured parameters are corrected and the corrected target structured parameters are verified a second time, including: marking the parameters that fail the verification in the target structured parameters as abnormal parameter items; correcting the abnormal parameter items to obtain the corrected target structured parameters; and performing multi-level verification on the corrected target structured parameters to perform a second verification on the corrected target structured parameters.

[0078] For example, after completing the multi-level verification process, the system iterates through all verification results. For any parameter item (e.g., rated voltage, rated capacity) or combination of parameter items that is determined to fail at any verification level (including compliance, correlation, or adaptability verification), the system adds an "abnormal" status marker to its corresponding data structure. Simultaneously, the system also records the specific reasons for this marker as a list of verification issues, such as "format error," "logical contradiction," or "out of reasonable range."

[0079] For example, a checklist containing all abnormal parameter items and their problem descriptions is presented to the relevant implementing entity (such as on-site maintenance personnel) through a front-end user interface. Based on this checklist, and referring to physical device nameplates or other authoritative data sources, the implementing entity manually checks and modifies the erroneous parameter items and submits the corrected data through the interface. Upon receiving this submitted data, the system updates the parameter values ​​of the corresponding abnormal parameter items in the target structured parameters with the corrected values, thereby obtaining the corrected target structured parameters.

[0080] For example, in addition to the methods shown in the previous examples, abnormal parameter items can also be found using a difference highlighting algorithm. Specifically, the difference highlighting algorithm can be expressed as: In the formula, This serves as a difference identifier for parameter i; For parameter i in the target structured parameters; The parameter i is the parameter item in the original parameter text; This is the preset deviation threshold.

[0081] For example, performing secondary verification on the corrected target structured parameters according to the multi-level verification method described in the aforementioned example helps reduce the risk of erroneous parameters re-entering the system.

[0082] According to the above implementation method, by marking parameters that fail the validation in the target structured parameters as abnormal parameter items, the problematic parameters are accurately located. Then, targeted corrections are performed on the abnormal parameter items to obtain the corrected target structured parameters. Based on this, a preset difference highlighting algorithm and difference threshold are introduced to highlight the differences between the parameters before and after correction. A secondary validation is then performed on the corrected target structured parameters to focus on verifying whether the corrected content meets the preset specifications and constraints. By constructing a targeted correction and secondary validation mechanism centered on abnormal parameter items, this invention avoids redundant calculations and manual intervention caused by repeatedly validating all parameters, concentrating validation resources on problematic parameter items.

[0083] In one implementation, recognizing text information in a nameplate image to obtain original parameter text includes: preprocessing the nameplate image to obtain a preprocessed nameplate image; performing quality detection on the preprocessed nameplate image; if the quality detection result is a failure, generating a nameplate image acquisition command and sending the nameplate image acquisition command to the user corresponding to the device to be acquired; if the quality detection result is a pass, calling a character recognition algorithm to detect the preprocessed nameplate image to obtain the target text region; extracting characters from the target text region to obtain target character information; and integrating the target character information to obtain the original parameter text.

[0084] For example, the preprocessing includes normalizing the image size to unify the resolution specifications, then performing denoising to eliminate salt-and-pepper noise or Gaussian noise caused by the shooting environment, and then performing grayscale conversion and contrast enhancement to highlight the brightness difference between the text area and the background area, and finally outputting the preprocessed nameplate image.

[0085] Example Xingdi performs sharpness assessment, occlusion detection, and integrity detection on the preprocessed nameplate image. Specifically, it determines whether there is severe blur by calculating the edge gradient value of the image, determines whether the nameplate is complete by detecting whether key areas are occluded, and statistically analyzes the proportion of effective text areas in the image to form a comprehensive quality detection result.

[0086] For example, if the text outlines in a nameplate image are blurred due to jitter, and the system detects that the edge gradient value is lower than a preset threshold, then the system determines that the quality of the preprocessed nameplate image does not meet the recognition requirements, resulting in a detection failure.

[0087] For example, if the quality inspection result is that the inspection fails, a collection instruction containing a re-shooting prompt is generated based on the specific reason for the failure, and the collection instruction is sent to the corresponding user terminal through a task push mechanism, thereby guiding the user to re-acquire a qualified nameplate image.

[0088] For example, when the system detects that the nameplate image is obstructed, it generates a collection instruction containing "Please adjust the shooting angle to ensure that the nameplate is fully visible" and sends the instruction to the user terminal responsible for the collection task of the device.

[0089] For example, if the quality inspection result is "passed," a text detection model (e.g., Efficient and Accurate Scene Text Detector (EAST) or Differentiable Binarization (DB)) is used to identify regions in the image that may contain characters, and the target text regions are marked with rectangular or polygonal boxes, thereby distinguishing text regions from non-text regions. For instance, in a qualified switchgear nameplate image, the system detects regions containing fields such as "rated current," "model," and "manufacturer," and identifies these regions as target text regions.

[0090] For example, each target text region is input into a character recognition model (e.g., a Convolutional Recurrent Neural Network (CRNN)), which recognizes the characters line by line and outputs the corresponding character sequences, thereby forming a discrete set of character information. For instance, for a target text region containing "rated voltage: 10kV", the system recognizes and outputs the character sequences "rated voltage" "10kV" as part of the target character information.

[0091] For example, the extracted target character information is spliced ​​and recombined according to the spatial order or field logic relationship of the target text region in the nameplate image, so as to transform the discrete character information into a parameter description text with semantic coherence, thereby forming the original parameter text.

[0092] For example, the character information such as "Equipment Model: XYZ-100", "Rated Voltage: 10kV", and "Rated Capacity: 1000kVA" that are identified separately are integrated into the original parameter text "Equipment Model is XYZ-100, Rated Voltage is 10kV, Rated Capacity is 1000kVA", which is used for subsequent parameter parsing and structured processing.

[0093] According to the above implementation method, by preprocessing the nameplate image and performing quality detection, the image clarity and recognizability are controlled at the source. When the quality detection result is unsatisfactory, a nameplate image acquisition command is automatically generated and issued to guide the user to re-acquire a nameplate image that meets the requirements. When the quality detection result is satisfactory, a character recognition algorithm is invoked to detect the preprocessed nameplate image to obtain the target text region, and character extraction and integration processing is performed on the target text region to generate the original parameter text. In this way, the problem of a large number of misidentifications caused by low-quality images directly entering the recognition process can be avoided, and character extraction is based on high-quality images, thereby significantly improving the accuracy of nameplate text recognition. At the same time, by locating the target text region and performing character integration processing, the interference of irrelevant background on the recognition results is reduced, improving the stability and consistency of parameter text acquisition, and thus providing reliable input for subsequent structured parsing and verification processing.

[0094] In one implementation, after obtaining the target parameters of the device to be collected, the target parameters can be synchronized to the control cloud system connected to the control cloud device parameter processing device. Specifically, the control cloud device parameter processing device uniformly encapsulates the target parameters, the nameplate image of the device to be collected corresponding to the target parameters, and the verification record formed for the target parameters to generate a data packet for transmission. The data packet is encrypted using an encryption format (e.g., a symmetric encryption format based on Advanced Encryption Standard 256-bit Galois / Counter Mode (AES-256-GCM) or a hybrid encryption format based on a combination of RSA public-key encryption algorithm (Rivest-Shamir-Adleman, RSA) and Advanced Encryption Standard (AES)) to ensure the integrity and security of the data during transmission.

[0095] After the data packet is generated, it is synchronized to the control cloud system using at least one preset synchronization method. For example, an intranet synchronization method can be used, in which the data packet is imported into an intranet workstation, and then uploaded to the control cloud system through a data synchronization service deployed on the intranet workstation, thereby achieving secure data transmission based on a dedicated network environment.

[0096] For example, an application synchronization method can also be used, that is, through a preset application, data packets are directly transmitted to the target interface of the control cloud system according to the specified interface protocol provided by the control cloud system, so as to realize the remote synchronization of target parameters, nameplate images and verification records.

[0097] Through the above methods, this embodiment can reliably synchronize target parameters and their associated data to the control cloud system while ensuring data security, providing a data foundation for subsequent parameter management, control analysis, and rapid access.

[0098] Figure 2 This is a structural block diagram of a cloud device parameter processing device according to an embodiment of the present invention.

[0099] like Figure 2 As shown, the parameter processing device for regulating cloud equipment may include: The acquisition module 510 is used to acquire the nameplate image of the device to be acquired; The recognition module 520 is used to recognize the text information in the nameplate image to obtain the original parameter text; The structured parsing module 530 is used to perform structured parsing on the original parameter text to obtain the target structured parameters; The multi-level verification module 540 is used to perform multi-level verification on the target structured parameters to obtain the target verification result. The correction module 550 is used to correct the abnormal parameter items in the target structured parameters and perform a second verification on the corrected target structured parameters if the target verification result is that the verification fails. The target parameter determination module 560 is used to determine the target structured parameters as the target parameters of the device to be acquired if the target verification result is a successful verification.

[0100] In one implementation, the structured parsing module includes: The semantic hierarchy decomposition unit is used to decompose the original parameter text into multiple key semantic segments. The similarity calculation unit is used to calculate the similarity between the parameter attributes in the key semantic segments and the parameter attributes in the preset feature library for each key semantic segment, and to obtain an attribute similarity value; and to match the parameter states in the key semantic segments with the parameter states in the feature library, and to obtain a state similarity value. The extraction unit is used to extract parameter attributes and parameter states from each key semantic segment based on each attribute similarity value and each state similarity value, so as to obtain each target parameter attribute and each target parameter state; The mapping unit is used to map each of the target parameter attributes and each of the target parameter states based on a preset mapping relationship between parameter attributes and parameter states, so as to obtain each target mapping relationship; The subordinate relationship division unit is used to divide the subordinate relationships of each of the target parameter attributes to obtain each main parameter category and at least one secondary parameter item corresponding to the main parameter category; The structured encapsulation unit is used to structure and encapsulate each of the target mapping relationships, each of the main parameter categories, and at least one secondary parameter item corresponding to the main parameter category to obtain the target structured parameters.

[0101] In one embodiment, the subordinate relationship partitioning unit includes: The hierarchical partitioning subunit is used to perform hierarchical partitioning of each target parameter attribute based on the label of each target parameter attribute to obtain each first hierarchical partitioning result, wherein the first hierarchical partitioning result includes at least one secondary candidate parameter sub-item corresponding to each candidate main parameter category. The adaptability verification subunit is used to perform adaptability verification on each of the second-level division results based on the preset control cloud system parameter storage architecture and data retrieval rules. The adjustment subunit is used to adjust the second-level partitioning results that are not conducive to parameter management or fast calling, so as to obtain the third-level partitioning results; The fourth-level partitioning result determination subunit is used to determine each second-level partitioning result that is conducive to parameter management or fast calling as each fourth-level partitioning result; The verification subunit is used to verify each of the fourth-level partitioning results and the third-level partitioning results, and to determine each candidate main parameter category and at least one secondary candidate parameter item corresponding to each candidate main parameter category as at least one secondary parameter item corresponding to each main parameter category.

[0102] In one implementation, the verification subunit is specifically used for: Based on the fourth-level partitioning results and the third-level partitioning results, the fifth-level partitioning results are determined; Based on the preset hierarchical division reference benchmark, the fifth level division result is compared and verified; If the verification fails, at least one secondary candidate parameter item corresponding to each candidate primary parameter category that does not match the hierarchical division reference benchmark in the fifth-level division result shall be adjusted. If the verification is successful, then each candidate primary parameter category and at least one secondary candidate parameter item corresponding to each candidate primary parameter category in the fifth-level division result will be determined as each primary parameter category and at least one secondary parameter item corresponding to each primary parameter category.

[0103] In one embodiment, the multi-level verification module includes: A compliance verification unit is used to perform compliance verification on the target structured parameters and obtain a compliance verification result. The correlation verification unit is used to perform correlation verification on the target structured parameters based on preset logical constraints, and obtain the correlation verification result. The adaptability verification unit is used to perform adaptability verification on the target structured parameters based on preset application scenario requirements, and obtain scenario verification results. The first verification result unit is used to determine the target verification result as verified if the compliance verification result, the correlation verification result, and the scenario verification result are all verified. The second verification result unit is used to determine the target verification result as failing if any one of the compliance verification result, the correlation verification result, and the scenario verification result fails the verification.

[0104] In one embodiment, the correction module includes: A marking unit is used to mark parameters that fail the validation in the target structured parameters as abnormal parameter items; An abnormal parameter item correction unit is used to correct the abnormal parameter item to obtain the corrected target structured parameter; The secondary verification unit is used to perform the multi-level verification on the corrected target structured parameters to perform secondary verification on the corrected target structured parameters.

[0105] In one embodiment, the identification module includes: The preprocessing unit is used to preprocess the nameplate image to obtain a preprocessed nameplate image; A quality inspection unit is used to perform quality inspection on the preprocessed nameplate image; The first quality inspection result unit is used to generate a nameplate image acquisition command if the quality inspection result is that the inspection fails, and to send the nameplate image acquisition command to the user corresponding to the device to be acquired. The second quality inspection result unit is used to call a character recognition algorithm to detect the preprocessed nameplate image and obtain the target text region if the quality inspection result is that the inspection is passed. A character extraction unit is used to extract characters from the target text region to obtain target character information; An integration unit is used to integrate the target character information to obtain the original parameter text.

[0106] The specific functions and examples of each module and submodule of the system in this embodiment of the invention can be found in the relevant descriptions of the corresponding steps in the above method embodiments, and will not be repeated here.

[0107] The acquisition, storage, and application of user personal information involved in the technical solution of this invention all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0108] This invention also provides an electronic device, comprising: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method described in any one of the embodiments of the present invention.

[0109] The beneficial effects of the electronic device in this embodiment of the invention are equivalent to the beneficial effects of the above-described method for controlling cloud device parameters, and will not be repeated here.

[0110] This invention also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method described in any one of the embodiments of this invention.

[0111] The beneficial effects of the storage medium of the present invention are equivalent to the beneficial effects of the above-described method for controlling cloud device parameters, and will not be repeated here.

[0112] Figure 3 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present invention is shown. Electronic device 800 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 800 may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0113] like Figure 3As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. The RAM 803 may also store various programs and data required for the operation of the electronic device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0114] Multiple components in electronic device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of displays, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows electronic device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0115] The computing unit 801 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the cloud device parameter processing method. For example, in some embodiments, the cloud device parameter processing method can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the cloud device parameter processing method described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the cloud device parameter processing method by any other suitable means (e.g., by means of firmware).

[0116] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0117] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0118] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0119] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0120] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0121] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0122] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0123] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for controlling cloud device parameter processing, characterized in that, include: Acquire the nameplate image of the device to be acquired; The text information in the nameplate image is identified to obtain the original parameter text; The original parameter text is parsed in a structured manner to obtain the target structured parameters; The target structured parameters are validated at multiple levels to obtain the target validation results. If the target verification result is that the verification fails, the abnormal parameter items in the target structured parameters are corrected and the corrected target structured parameters are verified a second time. If the target verification result is successful, then the target structured parameters are determined as the target parameters of the device to be acquired.

2. The method according to claim 1, characterized in that, The step of performing structured parsing on the original parameter text to obtain the target structured parameters includes: The original parameter text is decomposed into semantic layers to obtain multiple key semantic segments; For each of the key semantic segments, the similarity between the parameter attributes in the key semantic segment and the parameter attributes in the preset feature library is calculated to obtain an attribute similarity value; and the parameter states in the key semantic segments are matched with the parameter states in the feature library to obtain a state similarity value. Based on the similarity values ​​of each attribute and the similarity values ​​of each state, parameter attributes and parameter states are extracted from each key semantic segment to obtain each target parameter attribute and each target parameter state; Based on the preset mapping relationship between parameter attributes and parameter states, each of the target parameter attributes and each of the target parameter states are mapped to obtain each target mapping relationship; The attribute of each target parameter is divided into subordinate relationships to obtain each main parameter category and at least one secondary parameter item corresponding to the main parameter category; The target structured parameters are obtained by structurally encapsulating each of the target mapping relationships, each of the main parameter categories, and at least one secondary parameter item corresponding to the main parameter category.

3. The method according to claim 2, characterized in that, The step of dividing the attribute of each target parameter into subordinate relationships to obtain each main parameter category and at least one secondary parameter item corresponding to the main parameter category includes: Based on the labels of each of the target parameter attributes, the target parameter attributes are hierarchically divided to obtain each first-level division result, wherein the first-level division result includes at least one secondary candidate parameter sub-item corresponding to each candidate main parameter category; Based on at least one secondary candidate parameter item corresponding to each of the candidate main parameter categories and a preset judgment condition, each of the first-level division results is extracted to obtain each of the second-level division results, wherein the judgment condition is that each of the secondary candidate parameter items satisfies the functional requirements of the candidate main parameter category. Based on the preset control cloud system parameter storage architecture and data retrieval rules, the adaptability of each second-level division result is verified. Adjust the second-level partitioning results that are not conducive to parameter management or fast access based on the adaptability verification results to obtain the third-level partitioning results; The second-level partitioning results that are conducive to parameter management or fast calling will be determined as the fourth-level partitioning results. The fourth-level partitioning results and the third-level partitioning results are verified. Each candidate main parameter category and each candidate main parameter category corresponding to at least one secondary candidate parameter item are determined as each main parameter category and each main parameter category corresponding to at least one secondary parameter item.

4. The method according to claim 3, characterized in that, The step of verifying each of the fourth-level partitioning results and the third-level partitioning results, and determining each verified candidate primary parameter category and at least one secondary candidate parameter item corresponding to each candidate primary parameter category as at least one secondary parameter item corresponding to each primary parameter category, includes: Based on the results of the fourth-level partitioning and the results of the third-level partitioning, the results of the fifth-level partitioning are determined; Based on the preset hierarchical division reference benchmark, the fifth level division result is compared and verified; If the verification fails, at least one secondary candidate parameter item corresponding to each candidate primary parameter category that does not match the hierarchical division reference benchmark in the fifth-level division result shall be adjusted. If the verification is successful, then each candidate primary parameter category and at least one secondary candidate parameter item corresponding to each candidate primary parameter category in the fifth-level division result will be determined as each primary parameter category and at least one secondary parameter item corresponding to each primary parameter category.

5. The method according to claim 1, characterized in that, The multi-level verification of the target structured parameters to obtain the target verification result includes: The target structured parameters are subjected to compliance verification to obtain the compliance verification results; Based on preset logical constraints, the correlation of the target structured parameters is verified to obtain the correlation verification result. The target structured parameters are subjected to adaptability verification based on the preset application scenario requirements to obtain the scenario verification result. If the compliance verification result, the correlation verification result, and the scenario verification result are all passed, then the target verification result is determined to be passed. If any one of the compliance verification result, the correlation verification result, and the scenario verification result fails the verification, then the target verification result is determined to be a verification failure.

6. The method according to claim 1, characterized in that, If the target verification result is a verification failure, then the abnormal parameter items in the target structured parameters are corrected and a second verification is performed on the corrected target structured parameters, including: Mark the parameters that fail the validation in the target structured parameters as abnormal parameter items; The abnormal parameter items are corrected to obtain the corrected target structured parameters; The multi-level verification is performed on the corrected target structured parameters to perform a secondary verification on the corrected target structured parameters.

7. The method according to claim 1, characterized in that, The process of recognizing the text information in the nameplate image to obtain the original parameter text includes: The nameplate image is preprocessed to obtain a preprocessed nameplate image; The preprocessed nameplate image is subjected to quality inspection; If the quality inspection result is that the inspection fails, a nameplate image acquisition command is generated and sent to the user corresponding to the device to be acquired. If the quality inspection result is a pass, then the character recognition algorithm is called to detect the preprocessed nameplate image to obtain the target text region; The target text region is subjected to character extraction to obtain target character information; The target character information is integrated to obtain the original parameter text.

8. A device for processing parameters of cloud equipment, characterized in that, include: The acquisition module is used to acquire the nameplate image of the device to be acquired. The recognition module is used to recognize the text information in the nameplate image to obtain the original parameter text; The structured parsing module is used to perform structured parsing on the original parameter text to obtain the target structured parameters; A multi-level verification module is used to perform multi-level verification on the target structured parameters to obtain the target verification result; The correction module is used to correct the abnormal parameter items in the target structured parameters and perform a second verification on the corrected target structured parameters if the target verification result is that the verification fails. The target parameter determination module is used to determine the target structured parameters as the target parameters of the device to be acquired if the target verification result is a successful verification.

9. An electronic device, characterized in that, include: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

Images