Intelligent electric energy meter acquisition task scheduling management system and method

Abstract

The invention relates to the technical field of intelligent electric energy meters, and particularly discloses an intelligent electric energy meter collection task scheduling management system and method. According to the method, coding and depth extraction are carried out on a communication capability influence factor set of each electric energy meter by adopting a data processing and coding technology based on deep learning to obtain a set of communication capability multi-level depth features of the electric energy meters; and carrying out anchor point self-adaptive query response on each feature and the whole feature set to obtain the optimal probability of the communication capability of each electric energy meter, and carrying out sorting to generate a communication capability sorting table. According to the method, the communication capability influence factor set of the electric energy meter can be coded to extract multi-level and complex features, and meanwhile, the importance of each feature can be dynamically adjusted according to actual conditions, so that the flexibility and the adaptability of the model are further improved, and finer scheduling can be carried out subsequently according to different priorities and requirements.

Description

Technical Field

[0001] This application relates to the technical field of intelligent electricity meters, and more specifically, to an intelligent electricity meter acquisition task scheduling and management system and method. Background Art

[0002] With the rapid development of the smart grid, as an important terminal device in the power system, the electricity meter undertakes the key tasks of data acquisition and transmission. In order to ensure the efficient operation and precise management of the power system, it is necessary to conduct real-time monitoring and data acquisition of a large number of electricity meters.

[0003] Patent CN115237552A proposes a collection task scheduling and management method based on an object-oriented protocol. First, a set of influencing factors and a ranking table of the communication capabilities of electricity meters are established. Then, the status of the collection tasks is loaded and monitored, and the tasks are associated with the corresponding electricity meters according to the priorities. Tasks are assigned based on the communication capability ranking table, and electricity meters with strong communication capabilities are preferentially selected to execute high-priority tasks, and the communication scores are continuously updated during the execution process. The above steps are repeated until all tasks are completed to ensure efficient and accurate data acquisition and management.

[0004] In this patent, the ranking table multiplies the proportion of the influencing factors of each electricity meter by the corresponding coefficients and obtains a comprehensive score through a linear combination. Finally, the electricity meters are ranked from high to low according to the comprehensive score. However, this method only relies on simple statistical calculations of the proportion of each factor and cannot capture the deep patterns and complex relationships in the data. In addition, the method in the patent uses fixed weight coefficients, lacking flexibility and being difficult to dynamically adjust to adapt to changes in different application scenarios or data distributions. This limits its applicability and optimization potential in complex power grid environments, resulting in the system may not be able to fully exert its optimal performance.

[0005] Therefore, an optimized intelligent electricity meter acquisition task scheduling and management solution is desired. Summary of the Invention

[0006] This application provides an intelligent electricity meter acquisition task scheduling and management system and method, which can encode the set of influencing factors of the communication capabilities of electricity meters to extract multi-level and complex features, and at the same time can dynamically adjust the importance of each feature according to the actual situation, further enhancing the flexibility and adaptability of the model, so that more refined scheduling can be performed according to different priorities and requirements in the future.

[0007] According to one aspect of this application, an intelligent electricity meter acquisition task scheduling and management method is provided, including: S1: Obtain the set of influencing factors of the communication capabilities of each electricity meter, and rank the communication capabilities of each electricity meter based on the set of influencing factors of the communication capabilities of each electricity meter to obtain a communication capability ranking table.

[0008] S2: Load and activate the acquisition task, and associate the acquisition task with the relevant electricity meters.

[0009] S3: Based on the communication capability ranking table, perform task priority reading and acquisition for each electricity meter.

[0010] S4: Repeat S2 - S3 until all tasks are completed.

[0011] Among them, the S1 includes: S11: Respectively perform communication capability semantic anchor query adaptive decision response on the communication capability influencing factor sets of each electricity meter and the communication capability influencing factor set of all electricity meters to obtain a set of communication capability evaluation results as the communication capability evaluation results of each electricity meter; S12: Rank the communication capabilities of each electricity meter based on the set of communication capability evaluation results to obtain the communication capability ranking table.

[0012] According to another aspect of the present application, there is provided an intelligent electricity meter acquisition task scheduling and management system, including: an electricity meter communication efficiency analysis module, configured to obtain the communication capability influencing factor sets of each electricity meter, and rank the communication capabilities of each electricity meter based on the communication capability influencing factor sets of each electricity meter to obtain a communication capability ranking table.

[0013] An electricity meter task association management module, configured to load and activate the acquisition task, and associate the acquisition task with the relevant electricity meters.

[0014] An electricity meter communication priority scheduling module, configured to perform task priority reading and acquisition for each electricity meter based on the communication capability ranking table.

[0015] An electricity meter task loop execution module, configured to make the electricity meter task association management module and the electricity meter communication priority scheduling module sequentially repeat corresponding actions until all tasks are completed.

[0016] Among them, the electricity meter communication efficiency analysis module includes: a communication capability adaptive evaluation unit, configured to respectively perform communication capability semantic anchor query adaptive decision response on the communication capability influencing factor sets of each electricity meter and the communication capability influencing factor set of all electricity meters to obtain a set of communication capability evaluation results as the communication capability evaluation results of each electricity meter; a communication capability ranking unit, configured to rank the communication capabilities of each electricity meter based on the set of communication capability evaluation results to obtain the communication capability ranking table.

[0017] An intelligent electric energy meter acquisition task scheduling and management system and method provided by this application use data processing and encoding technologies based on deep learning to encode and deeply extract the set of factors affecting the communication capabilities of each electric energy meter to obtain a set of multi-level deep features of the communication capabilities of the electric energy meters. Then, each feature is subjected to anchor point adaptive query response with the entire set of features to obtain the probability of the best communication capabilities of each electric energy meter and sort them to generate the communication capability ranking table. This application can encode the set of factors affecting the communication capabilities of electric energy meters to extract multi-level and complex features, and at the same time can dynamically adjust the importance of each feature according to the actual situation, further improving the flexibility and adaptability of the model, so that more refined scheduling can be carried out according to different priorities and requirements in the future. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of this application, the drawings of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below only relate to some embodiments of this application and do not limit this application.

[0019] Figure 1 It is a schematic flowchart of the intelligent electric energy meter acquisition task scheduling and management method of the embodiment of this application.

[0020] Figure 2 It is a schematic flowchart of step S1 in the intelligent electric energy meter acquisition task scheduling and management method of the embodiment of this application.

[0021] Figure 3 It is a schematic flowchart of step S11 in the intelligent electric energy meter acquisition task scheduling and management method of the embodiment of this application.

[0022] Figure 4 It is a schematic flowchart of step S114 in the intelligent electric energy meter acquisition task scheduling and management method of the embodiment of this application.

[0023] Figure 5 It is a schematic flowchart of step S1142 in the intelligent electric energy meter acquisition task scheduling and management method of the embodiment of this application.

[0024] Figure 6 It is a schematic block diagram of the intelligent electric energy meter acquisition task scheduling and management system of the embodiment of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0025] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the drawings. Obviously, the described embodiments are only some embodiments of this application, rather than all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall also fall within the protection scope of this application.

[0026] Based on this, as Figure 1 shown, this application proposes an intelligent electricity meter acquisition task scheduling and management method. As Figure 1 shown, the intelligent electricity meter acquisition task scheduling and management method includes: S1: Obtain the set of communication ability influencing factors of each electricity meter, and sort the communication abilities of each electricity meter based on the set of communication ability influencing factors of each electricity meter to obtain a communication ability ranking table; S2: Load and activate the acquisition task, and associate the acquisition task with the relevant electricity meters; S3: Based on the communication ability ranking table, perform task priority reading and acquisition for each electricity meter; S4: Repeat S2 - S3 until all tasks are completed.

[0027] Specifically, step S1: Obtain the set of communication ability influencing factors of each electricity meter, and sort the communication abilities of each electricity meter based on the set of communication ability influencing factors of each electricity meter to obtain a communication ability ranking table. In particular, the set of communication ability influencing factors includes: average single - frame communication transmission time, the number of consecutive recent reading failures on the same day, the number of acquisition tasks involved in the electricity meter, and the carrier network level of the electricity meter, etc. It should be understood that in the actual sub - station area environment, the communication status of electricity meters is interfered by various factors. The signal intensities at the locations of different electricity meters are different, and the quality of the connection lines is also uneven, resulting in differences in communication stability and speed. By collecting influencing factors such as average single - frame communication transmission time and the number of consecutive recent reading failures on the same day, the communication characteristics of each electricity meter can be comprehensively understood, and its communication performance in a complex environment can be grasped. And the data acquisition of electricity meters depends on stable and reliable communication. If the communication ability differences are not considered, a large amount of time and resources may be wasted on electricity meters with poor communication ability, resulting in problems such as low acquisition efficiency, data loss, or errors. Understanding the communication ability influencing factors and sorting them helps to identify electricity meters with poor communication ability in advance and take targeted measures.

[0028] Based on this, the technical concept of this application is to use data processing and coding techniques based on deep learning to encode and deeply extract the set of communication ability influencing factors of each electricity meter to obtain a set of multi - level deep features of the electricity meter communication ability. Then, each feature is subjected to anchor - point adaptive query response with the entire feature set to obtain the probability of the best communication ability of each electricity meter and sort them to generate the communication ability ranking table. This application can encode the set of communication ability influencing factors of electricity meters to extract multi - level and complex features, and at the same time can dynamically adjust the importance of each feature according to the actual situation, further improving the flexibility and adaptability of the model, so that more refined scheduling can be performed according to different priorities and requirements in the future.

[0029] In one embodiment, as Figure 2 shown, the S1 includes: S11: respectively performing communication ability semantic anchor query adaptive decision response on the communication ability influencing factor set of each electric energy meter and the communication ability influencing factor set of all electric energy meters to obtain a set of communication ability evaluation results as the communication ability evaluation results of each electric energy meter; S12: sorting the communication abilities of each electric energy meter based on the set of communication ability evaluation results to obtain the communication ability sorting table.

[0030] In one embodiment, as Figure 3 shown, the S11 includes: S111: performing low-dimensional embedding encoding on each communication ability influencing factor in the communication ability influencing factor set of each electric energy meter to obtain a set of communication ability influencing factor embedding encoding vectors of each electric energy meter; S112: performing multi-level encoding on the set of communication ability influencing factor embedding encoding vectors of each electric energy meter to obtain a set of multi-level deep implicit encoding vectors of the electric energy meter communication ability; S113: extracting the multi-level deep implicit encoding vector of the target electric energy meter from the set of multi-level deep implicit encoding vectors of the electric energy meter communication ability as the electric energy meter communication ability query vector; S114: performing query-communication ability set semantic anchor adaptive query response on the electric energy meter communication ability query vector and the set of multi-level deep implicit encoding vectors of the electric energy meter communication ability to obtain a target electric energy meter communication ability semantic query response encoding vector; S115: based on the target electric energy meter communication ability semantic query response encoding vector, obtaining a communication ability evaluation result, where the communication ability evaluation result is the probability that the target electric energy meter is the electric energy meter with the best communication ability; S116: circularly executing S113 to S115 to obtain the set of communication ability evaluation results.

[0031] Specifically, step S111: performing low-dimensional embedding encoding on each communication ability influencing factor in the communication ability influencing factor set of each electric energy meter to obtain a set of communication ability influencing factor embedding encoding vectors of each electric energy meter. It should be understood that considering the various factors in the communication ability influencing factor set of the electric energy meter, such as the average single-frame communication transmission time, the number of consecutive recent read failures on the same day, etc., the original data has a high dimension and may have a complex feature structure. In order to map the high-dimensional data to a low-dimensional space, simplify the data structure, retain important information while removing redundant information, and make subsequent calculations and analyses more efficient, the present application performs low-dimensional embedding encoding on each communication ability influencing factor in the communication ability influencing factor set of each electric energy meter to better capture the internal relationships and patterns between features, and obtain a set of communication ability influencing factor embedding encoding vectors of each electric energy meter.

[0032] In a specific embodiment, there is a group of electricity meters for which data collection is required, and each electricity meter has multiple communication ability influencing factors, such as the average single-frame communication transmission time, the number of consecutive recent reading failures on the same day, the number of collection tasks involved in the electricity meter, and the electricity meter carrier network level, etc. To simplify these high-dimensional features and capture their internal relationships, a low-dimensional embedding coding method can be used. First, collect the data of the communication ability influencing factors of each electricity meter. For example, for a specific electricity meter A, its data is as follows: the average single-frame communication transmission time is 50 ms, the number of consecutive recent reading failures on the same day is 3 times, the number of collection tasks involved in the electricity meter is 15, and the electricity meter carrier network level is 4. Next, convert these original high-dimensional features into low-dimensional vectors through low-dimensional embedding coding. Here, an Embedding Layer is used to implement this process. Define an embedding layer with an embedding dimension of 8, which means each influencing factor will be mapped into an 8-dimensional vector space. Create an embedding matrix, where each row corresponds to the embedding vector of an influencing factor. Initially, this matrix can be randomly initialized or initialized according to some prior knowledge. In a specific embodiment, the embedding matrix after training is as follows: the embedding vector corresponding to the average single-frame communication transmission time is [0.1, -0.2, 0.3, 0.4, -0.5, 0.6, -0.7, 0.8]; the embedding vector corresponding to the number of consecutive recent reading failures on the same day is [-0.1, 0.2, -0.3, 0.4, -0.5, 0.6, -0.7, 0.8]; the embedding vector corresponding to the number of collection tasks involved in the electricity meter is [0.2, -0.3, 0.4, -0.5, 0.6, -0.7, 0.8, -0.9]; the embedding vector corresponding to the electricity meter carrier network level is [-0.2, 0.3, -0.4, 0.5, -0.6, 0.7, -0.8, 0.9]. Here, this is only an example and does not constitute a limitation. For electricity meter A, the embedding vectors of its respective influencing factors are: the average single-frame communication transmission time is [0.1, -0.2, 0.3, 0.4, -0.5, 0.6, -0.7, 0.8]; the number of consecutive recent reading failures on the same day is [-0.1, 0.2, -0.3, 0.4, -0.5, 0.6, -0.7, 0.8]; the number of collection tasks involved in the electricity meter is [0.2, -0.3, 0.4, -0.5, 0.6, -0.7, 0.8, -0.9]; the electricity meter carrier network level is [-0.2, 0.3, -0.4, 0.5, -0.6, 0.7, -0.8, 0.9]. Finally, these communication ability influencing factor embedding coding vectors are aggregated into a complete set of communication ability influencing factor embedding coding vectors.For example, for electricity meter A, the set of embedded coding vectors of factors affecting its communication capability can be represented as a matrix, where each row is an embedded vector of an influencing factor: [[0.1, -0.2, 0.3,0.4, -0.5, 0.6, -0.7, 0.8], [-0.1, 0.2, -0.3, 0.4, -0.5, 0.6, -0.7, 0.8],[0.2, -0.3, 0.4, -0.5, 0.6, -0.7, 0.8, -0.9], [-0.2, 0.3, -0.4, 0.5, -0.6,0.7, -0.8, 0.9]]. In this way, the originally complex high-dimensional features are converted into a set of low-dimensional embedded coding vectors, which facilitates subsequent multi-level encoding and deeper feature extraction.

[0033] Specifically, step S112: multi-level encoding is performed on the set of embedded coding vectors of the communication capability influencing factors of each electric energy meter to obtain a set of multi-level deep implicit coding vectors of the communication capability of the electric energy meter. It should be understood that, considering that the communication capability of the electric energy meter is affected by many factors, there may be complex nonlinear relationships between these factors. Although low-dimensional embedded coding can simplify feature representation, it may not be able to fully capture these complex nonlinear relationships. Based on this, in the technical solution of the present application, the S112 includes: multi-level fully connected coding is performed on the set of embedded coding vectors of the communication capability influencing factors of each electric energy meter based on a multi-layer perceptron model to obtain a set of multi-level deep implicit coding vectors of the communication capability of the electric energy meter. It can be understood that the multi-layer perceptron model (MLP) has a strong nonlinear mapping capability, and through the fully connected structure of multi-layer neurons, a deeper feature extraction of the embedded coding vector can be performed. The relationship between factors such as the average single-frame communication transmission time and the number of recent consecutive reading failures on the same day is not a simple linear one. The multi-layer perceptron model can dig out the complex associations hidden in the deep layers of the data, capture more detailed feature information, and provide richer data support for accurately evaluating the communication capabilities of the electric energy meter. In this way, the obtained set of multi-level deep implicit coding vectors of the communication capability of the electric energy meter contains richer and more accurate information related to the communication capability, which can significantly improve the accuracy of subsequent communication capability evaluation, make the sorting results more in line with the actual situation, and provide a more reliable basis for the scheduling of collection tasks.

[0034] Specifically, step S113: Extract the multi-level depth implicit coding vector of the communication ability of the target electricity meter from the set of multi-level depth implicit coding vectors of the communication ability of the electricity meter as the query vector for the communication ability of the electricity meter. It should be understood that considering that the set of multi-level depth implicit coding vectors of the communication ability of the electricity meter contains the comprehensive communication ability information of many electricity meters. However, in the actual acquisition task scheduling process, it is often necessary to perform separate analysis and decision-making for a specific target electricity meter. Therefore, in the technical solution of this application, the multi-level depth implicit coding vector of the communication ability of the target electricity meter is extracted from the set of multi-level depth implicit coding vectors of the communication ability of the electricity meter as the query vector for the communication ability of the electricity meter. In this way, personalized feature representation can be provided for this electricity meter to ensure that the evaluation result is more accurate.

[0035] Specifically, step S114: Perform query-communication ability set semantic anchor point adaptive query response on the query vector of the communication ability of the electricity meter and the set of multi-level depth implicit coding vectors of the communication ability of the electricity meter to obtain the semantic query response coding vector of the communication ability of the target electricity meter. It should be understood that considering that a single query vector of the communication ability of the target electricity meter only reflects some characteristics of the meter itself. In reality, the communication ability of the electricity meter is affected by various factors, and there are complex associations between different electricity meters. In order to capture the overall distribution and pattern of the communication ability of the electricity meter in the entire power grid environment and dynamically model the communication state of the target electricity meter, this application performs query-communication ability set semantic anchor point adaptive query response on the query vector of the communication ability of the electricity meter and the set of multi-level depth implicit coding vectors of the communication ability of the electricity meter to obtain the semantic query response coding vector of the communication ability of the target electricity meter. In this way, more comprehensive information can be obtained from an overall perspective. In a complex substation area environment, different electricity meters are affected by signal interference, line loss, etc. to different degrees. By comparing the communication ability characterization information of all electricity meters, the relative advantages and disadvantages of the target electricity meter in the whole, as well as its similarities and differences with other electricity meters, can be found, and thus its communication ability can be comprehensively understood.

[0036] In one embodiment, as Figure 4As shown, the S114 includes: S1141: Semantically response-anchoring each of the multi-level deep implicit coding vectors of the electricity meter communication capabilities in the set of the electricity meter communication capabilities query vector and the multi-level deep implicit coding vectors of the electricity meter communication capabilities to obtain a set of query-electricity meter communication capabilities feature response-anchoring coding matrices; S1142: Calculating the decision-anchor adaptive splicing factors of each query-electricity meter communication capabilities feature response-anchoring coding matrix in the set of query-electricity meter communication capabilities feature response-anchoring coding matrices to obtain a set of query-electricity meter communication capabilities decision-anchor adaptive splicing weight factors; S1143: Based on the set of query-electricity meter communication capabilities decision-anchor adaptive splicing weight factors, fusing the set of query-electricity meter communication capabilities feature response-anchoring coding matrices to obtain the target electricity meter communication capabilities semantic query response coding vector.

[0037] In one embodiment, the S1141 includes: Extracting the deep implicit feature fully connected coding of the electricity meter communication capabilities query vector to obtain the electricity meter communication capabilities query deep implicit coding vector. Specifically, this process is represented by the formula: ; where is the electricity meter communication capabilities query vector, is matrix multiplication, and are the electricity meter communication capabilities query weight matrix and the electricity meter communication capabilities query bias vector respectively, is the activation function, is the electricity meter communication capabilities query deep implicit coding vector.

[0038] Performing deep implicit feature fully connected coding enhancement extraction on each of the multi-level deep implicit coding vectors of the electricity meter communication capabilities in the set of the multi-level deep implicit coding vectors of the electricity meter communication capabilities to obtain a set of electricity meter communication capabilities multi-level deep implicit coding enhancement vectors. Specifically, this process is represented by the formula:

[0039]

[0040] where is the set of the multi-level deep implicit coding vectors of the electricity meter communication capabilities, and are the first, second, th, and th multi-level deep implicit coding vectors of the electricity meter communication capabilities in the set of the multi-level deep implicit coding vectors of the electricity meter communication capabilities respectively, and are the weight matrix of the communication ability of the electricity meter and the bias vector of the communication ability of the electricity meter, respectively, is the th multi-level deep implicit coding enhancement vector of the communication ability of the electricity meter in the set of multi-level deep implicit coding enhancement vectors of the communication ability of the electricity meter.

[0041] The query-depth implicit coding vector of the communication ability of the electricity meter and each multi-level deep implicit coding enhancement vector of the communication ability of the electricity meter in the set of multi-level deep implicit coding enhancement vectors of the communication ability of the electricity meter are respectively subjected to semantic response decision anchoring of the communication ability of the electricity meter to obtain the set of query-electricity meter communication ability feature response anchoring coding matrices. Specifically, this process is represented by the formula:

[0042] where, is transpose vector of, is length of, is the th query-electricity meter communication ability feature response anchoring coding matrix in the set of query-electricity meter communication ability feature response anchoring coding matrices.

[0043] It should be understood that by performing a full - connection encoding extraction on the deep implicit features of the power meter communication ability query vector to obtain the deep implicit encoding vector of the power meter communication ability query, this process aims to transform the original communication ability query vector into a higher - dimensional and more semantically - informative representation. Through the mapping of a fully - connected neural network, the original features are projected into a latent space where the features not only have a richer semantic representation ability but also can better capture global and local non - linear relationships. This transformation enables the system to more accurately understand the communication ability characteristics of each power meter, providing a solid foundation for further task scheduling. Next, a deep implicit feature full - connection encoding enhancement extraction is performed on each power meter multi - level deep implicit encoding vector in the set of power meter multi - level deep implicit encoding vectors of the power meter communication ability to obtain a set of power meter multi - level deep implicit encoding enhancement vectors. This step further enhances the expression ability of each power meter communication ability feature. Then, the power meter communication ability query deep implicit encoding vector and each power meter multi - level deep implicit encoding enhancement vector in the set of power meter multi - level deep implicit encoding enhancement vectors are respectively subjected to power meter communication ability semantic response decision anchoring to obtain a set of query - power meter communication ability feature response anchoring encoding matrices. In this process, the system models the semantic relationship between the query vector and the enhancement vector to generate a series of local response matrices. These matrices not only reflect the communication ability interaction information between power meters but also serve as a bridge for capturing global - local semantic alignment relationships. Specifically, these matrices help the system identify which power meters perform better in a specific task or context and provide a basis for subsequent task allocation. For example, in actual operation, when the system needs to select a power meter from many power meters to perform a high - priority data acquisition task, it will first calculate the deep implicit encoding vectors of the communication abilities of each power meter. After the full - connection encoding extraction, these vectors have a stronger semantic representation ability. Then, the system enhances all the multi - level deep implicit encoding vectors of the power meters to ensure that the feature representation of each power meter is as comprehensive and accurate as possible. Then, the system uses these enhanced feature vectors for semantic response decision anchoring to generate a set of local response matrices. These matrices not only reveal the communication ability details of each power meter but also show their mutual relationships, enabling the system to make more reasonable scheduling decisions.

[0044] In one embodiment, as Figure 5As shown, the S1142 includes: S11421: determining the decision anchor adaptive splicing factors of each query-meter communication capability response anchor encoding matrix based on the feature distribution of each query-meter communication capability response anchor encoding matrix in the set of query-meter communication capability response anchor encoding matrices to obtain a set of query-meter communication capability decision anchor adaptive splicing factors; S11422: weighting the set of query-meter communication capability decision anchor adaptive splicing factors based on the Softmax function to obtain a set of query-meter communication capability decision anchor adaptive splicing weight factors.

[0045] In one embodiment, the S11421 includes: calculating the corresponding query-meter communication capability decision anchor adaptive splicing factor of the query-meter communication capability response anchor encoding matrix based on the maximum value, mean value, variance, and difference amplification coefficient of the query-meter communication capability response anchor encoding matrix and the meter communication capability drift coefficient; wherein, the meter communication capability drift coefficient is obtained by performing local-global importance score smoothing transition processing on the query-meter communication capability response anchor encoding matrix. Specifically, this process is represented by the formula:

[0046]

[0047]

[0048]

[0049] Wherein, represents the variance of represents the mean value of is for calculating the number of eigenvalues of represents the number of eigenvalues of as the difference amplification coefficient, represents ， the maximum value of represents the meter communication capability drift coefficient, is the intermediate state transition, is the th eigenvalue of is the

[0050] Specifically, the calculation process of step S11422 is expressed by the formula: ; where is a normalization function, is the -th query - electricity meter communication ability decision anchor adaptive splicing weight factor in the set of query - electricity meter communication ability decision anchor adaptive splicing weight factors.

[0051] Specifically, after generating the query - electricity meter communication ability feature response anchor coding matrix, the system will conduct an in - depth analysis of these matrices to determine the feature distribution in each matrix. Through this analysis, the system can identify which features are more important in a specific task or context. For example, when performing a high - priority data collection task, some electricity meters may have higher weights due to factors such as strong communication ability and stable historical performance. These weights not only reflect the actual communication ability of the electricity meters but also consider the specific requirements of the task and environmental factors. To ensure that these weights can accurately reflect the contribution degree of each electricity meter, the system further calculates the query - electricity meter communication ability decision anchor adaptive splicing factors. These factors measure the dominance of local features in a specific task or context and serve as the weight basis for subsequent feature fusion. In this way, the system can dynamically adjust the priority and weight of each electricity meter in task scheduling to ensure the reasonable allocation and efficient utilization of resources. Next, a weight processing based on the Softmax function is performed on the set of query - electricity meter communication ability decision anchor adaptive splicing factors to obtain the set of query - electricity meter communication ability decision anchor adaptive splicing weight factors. The core of this step is to convert the adaptive splicing factors into a weight sequence with the property of probability distribution. The Softmax function can not only normalize these factors so that their sum is 1 but also enhance the significance difference of local features through the characteristics of the exponential function, strengthening the distribution discrimination ability of features.

[0052] Preferably, for the electricity meter communication ability drift coefficient in the query - electricity meter communication ability decision anchor adaptive splicing factors, for the state transition of the input feature set distribution of the query - electricity meter communication ability feature response anchor coding matrix from weak overall interpretability of the mean to strong local interpretability of the maximum value, the global dominance basis of local features is enhanced through the weak - to - strong interpretable generalization of the electricity meter communication ability drift coefficient .

[0053] Specifically, taking as the intermediate state transition representation from weak interpretability to strong interpretability, for each eigenvalue , using it as the importance score of the input query - electricity meter communication ability feature response anchoring coding matrix for the global smooth state transition, to perform importance score weight global control for the intermediate state transition to achieve the interpretable generalization inference based on the weight of the query - electricity meter communication ability decision anchor adaptive splicing factor for the global state transition.

[0054] Specifically, in step S1143, after completing the analysis of the communication ability feature response anchoring coding matrix for each electricity meter and determining the corresponding decision anchor adaptive splicing weight factors, the system starts feature fusion. These weight factors not only measure the dominance of each electricity meter in a specific task or context but also serve as the weight basis for subsequent feature fusion. In this way, the system can dynamically adjust the priority and weight of each electricity meter in task scheduling to ensure reasonable resource allocation and efficient utilization. Next, the system will fuse the query - electricity meter communication ability feature response anchoring coding matrix according to these weight factors. The core of this process is to perform weighted summation of multiple local feature response matrices according to their weights to generate a comprehensive semantic query response coding vector. This vector not only contains the detailed communication ability information of each electricity meter but also reflects the mutual relationship and synergy among them. Specifically, this process is expressed by the formula:

[0055] where, is the semantic query response coding matrix of the target electricity meter communication ability, is the shape reshaping operation, is the semantic query response coding vector of the target electricity meter communication ability.

[0056] Specifically, step S115: Based on the semantic query response encoding vector of the target electricity meter communication ability, obtain a communication ability evaluation result, where the communication ability evaluation result is the probability that the target electricity meter is the electricity meter with the best communication ability. In one embodiment, S115 includes: inputting the semantic query response encoding vector of the target electricity meter communication ability into a communication ability evaluation module based on a classifier to obtain the communication ability evaluation result, where the communication ability evaluation result is the probability that the target electricity meter is the electricity meter with the best communication ability. That is, the semantic query response encoding vector of the target electricity meter obtained by performing semantic query using the electricity meter communication ability query vector and the set of multi-level deep implicit encoding vectors of the electricity meter communication ability is classified, so as to use the classification and prediction capabilities of the classifier to convert this complex information into a specific probability value. This probability value intuitively quantifies the possibility that the target electricity meter becomes the electricity meter with the best communication ability, making the evaluation of the communication ability clearer and more definite, and facilitating subsequent analysis and decision-making. In a specific embodiment, the communication ability evaluation module based on a classifier uses a Softmax classifier. The Softmax classifier is a commonly used multi-classification algorithm, especially suitable for tasks that require outputting the probability distribution of multiple classes. It converts the input vector into a probability distribution, such that the probability value of each class is between 0 and 1, and the sum of the probability values of all classes is 1.

[0057] Specifically, step S116: Loop through S113 to S115 to obtain a set of the communication ability evaluation results. Subsequently, loop through S113 to S115 to obtain a set of the communication ability evaluation results. That is, in an actual electricity meter acquisition system, the communication ability of each electricity meter is crucial and interrelated. Evaluating only some of the electricity meters cannot meet the system's need for a comprehensive understanding of the overall communication situation. By looping through these steps, all electricity meters can be traversed, ensuring that each electricity meter can obtain a corresponding evaluation, forming a complete communication ability evaluation system.

[0058] Specifically, in step S12: Sort the communication capabilities of each electricity meter based on the set of communication capability evaluation results to obtain the communication capability ranking table. It should be understood that the data collection of electricity meters depends on stable and reliable communication. Due to factors such as their locations and the quality of connection lines, there may be significant differences in the communication conditions of different electricity meters. If these differences are not considered and data collection is directly carried out in a fixed order or priority, a large amount of time and resources may be wasted on electricity meters with poor communication capabilities, thus affecting the overall collection efficiency and even potentially resulting in data loss or errors. By evaluating the communication capabilities of each electricity meter and sorting them, electricity meters with strong communication capabilities and high success rates can be processed first to ensure the efficient and accurate completion of the data collection task. Secondly, the power grid environment is complex and changeable, and the communication status of electricity meters may change at any time. For example, some electricity meters may experience unstable communication due to signal interference, equipment failures, etc. If the communication capability ranking of electricity meters can be dynamically adjusted, these problems can be detected and addressed in a timely manner, and the collection strategy can be flexibly adjusted to avoid unnecessary resource consumption.

[0059] Specifically, in step S2: Load and activate the collection task, and associate the collection task with the relevant electricity meters. It should be understood that the collection task is the specific instruction for obtaining the data of electricity meters. Only by loading and activating it can the subsequent data collection operations be started. The collection task contains key information such as the collection cycle, collection content, and task priority, which are the basis for data collection. If the task is not loaded and activated, the collection system will not know which collection operations need to be performed and will not be able to complete the data collection work. Different electricity meters may be responsible for collecting different types of data, or according to the electricity consumption nature and management requirements, there may be special settings for the collection task priorities of certain electricity meters. Only by clarifying the corresponding relationship between the task and the electricity meter can the collection task be accurately executed.

[0060] Specifically, loading and activating the acquisition tasks and associating the acquisition tasks with relevant electricity meters includes the following steps: First, it is necessary to obtain and load all the acquisition tasks to be executed from the master station. These tasks usually include daily freeze reading tasks, monthly freeze reading tasks, curve reading tasks, and event acquisition tasks, etc. Each task has its specific acquisition cycle and priority. For example, the daily freeze task has an acquisition cycle of 1 day and a priority of 0, while the monthly freeze task has an acquisition cycle of 1 month and a priority of 1. Once the tasks are loaded, the next step is to activate these tasks. Activation means that these tasks are ready to be executed and can start to be assigned to the corresponding electricity meters for data acquisition. To achieve this, the status of the tasks will be monitored regularly to check which tasks are in the activated state. If a certain task is activated, it will be associated with the corresponding electricity meter according to its priority. Specifically, when a task is activated, it will be assigned to the associated electricity meter according to the priority of the task. This step involves two main aspects: one is to determine which electricity meters need to execute this task, and the other is how to establish an association relationship between the task and these electricity meters. Suppose there is a task list containing various types of tasks, such as: daily freeze task (priority 0); monthly freeze task (priority 1); curve reading task 1 (priority 1); curve reading task 2 (priority 1); event acquisition task (priority 2). For each type of task, it is necessary to clarify its specific acquisition content and the electricity meters involved. For example, the daily freeze task needs to acquire the forward active energy indication value and reverse active energy indication value of all electricity meters; while the curve reading task 1 needs to acquire the forward active total energy indication value, reverse active total energy indication value, voltage, current, active power, reactive power, power factor, etc. of all electricity meters. After clarifying the specific requirements of the tasks, appropriate electricity meters will be selected to execute these tasks according to the communication ability ranking table. For example, assume that a certain electricity meter A has a high communication ability score, then it may be preferentially assigned high-priority tasks. Specifically, if there are multiple tasks to be executed currently, the task with the highest priority will be assigned to electricity meter A first, and then other tasks will be assigned in turn. During this process, the communication ability ranking table of the electricity meters will also be updated dynamically. For example, when executing a certain task, if it is found that the communication of a certain electricity meter is unstable or the number of failure times increases, the communication ability score of this electricity meter will be adjusted in time, thereby affecting the subsequent task assignment strategy. In addition, in order to ensure that the tasks can be executed accurately and without error, detailed association records will be established between the tasks and the electricity meters. These records include not only the basic information of the tasks (such as task name, priority, acquisition content, etc.), but also the information of the associated electricity meters (such as electricity meter number, belonging channel, etc.). In this way, the execution situation of each task can be clearly traced, and adjustments or reassignments can be made when necessary.

[0061] Specifically, in step S3: Based on the communication ability ranking table, perform task-priority reading and acquisition for each electricity meter. Correspondingly, considering that there are differences in the communication abilities of different electricity meters, which are affected by various factors, such as the average single-frame communication transmission time, the number of consecutive reading failures recently on the same day, the number of acquisition tasks involved in the electricity meter, and the carrier network level of the electricity meter, etc. The communication ability ranking table comprehensively considers these factors and can truly reflect the communication status of each electricity meter. By performing reading and acquisition according to this table, tasks can be arranged based on the actual communication ability of the electricity meter, ensuring a more reasonable acquisition process.

[0062] Specifically, based on the communication ability ranking table, perform task-priority reading and acquisition for each electricity meter, including the following steps: First, check the status of the current task. If a certain task is activated, associate it with the corresponding electricity meter according to the priority. For electricity meter A, according to the associated task list, perform readings in sequence according to the task priority. For example, assume that electricity meter A needs to complete the following three tasks: daily freeze task (priority 0), curve meter reading task 1 (priority 1), and event acquisition task (priority 2). First, execute the daily freeze task with the highest priority to obtain data such as the forward active energy indication value and the reverse active energy indication value; then execute curve meter reading task 1 and the event acquisition task in sequence. During the execution of the tasks of electricity meter A, monitor its communication status in real time. If it is found that the communication of electricity meter A is unstable or there are multiple reading failures, continue to select the next electricity meter with strong communication ability according to the communication ability ranking table to execute the task. For example, assume that electricity meter A has a communication failure during the execution of curve meter reading task 1, and it will automatically switch to electricity meter B with the second-highest communication ability to continue executing the remaining tasks. Throughout the process, the communication ability ranking table of the electricity meter will also be continuously updated. For example, after each reading task is completed, collect the data on the factors affecting the communication ability of the electricity meter, such as the average single-frame communication transmission time and the number of consecutive reading failures in this reading. These data will be uploaded to the electricity meter ranking module, and the proportion of the communication ability factor data of each electricity meter in all factors will be calculated by the interval method, and combined with the importance coefficient of each factor, recalculate the communication ability coefficient of the electricity meter. According to the new communication ability coefficient, update the communication ability ranking table to ensure more reasonable and efficient subsequent task allocation. In addition, the task allocation of the electricity meter will also be processed separately according to different communication channels (such as the carrier channel and the 485 channel). For example, for the electricity meters under the carrier channel, give priority to selecting the electricity meters with strong communication ability to execute tasks; for the RS485 electricity meters, the default level is 0, and the tasks will also be allocated accordingly according to their communication abilities.

[0063] Specifically, step S4: Repeat steps S2 - S3 until all tasks are completed. In this way, by continuously repeating steps S2 - S3, the system can continuously check and process the unfinished tasks to ensure that no task is overlooked. Whether it is a high - priority or low - priority task, each electricity meter involved can be accessed and the data collection can be completed.

[0064] In summary, the intelligent electricity meter acquisition task scheduling and management system and method according to the embodiments of the present application are elucidated. It uses data processing and encoding techniques based on deep learning to encode and deeply extract the set of communication ability influencing factors of each electricity meter to obtain a set of multi - level deep features of the electricity meter communication ability. Then, each feature is subjected to anchor - point adaptive query response with the set of the entire features to obtain the probability of the best communication ability of each electricity meter and sort them to generate the communication ability ranking table. The present application can encode the set of communication ability influencing factors of the electricity meter to extract multi - level and complex features, and at the same time can dynamically adjust the importance of each feature according to the actual situation, further improving the flexibility and adaptability of the model, so that more refined scheduling can be performed according to different priorities and requirements in the future.

[0065] Figure 6 It is a schematic block diagram of the intelligent electricity meter acquisition task scheduling and management system according to the embodiments of the present application. As Figure 6 shown, the intelligent electricity meter acquisition task scheduling and management system 100 includes: an electricity meter communication efficiency analysis module 110, configured to obtain the set of communication ability influencing factors of each electricity meter, and sort the communication abilities of each electricity meter based on the set of communication ability influencing factors of each electricity meter to obtain a communication ability ranking table; an electricity meter task association management module 120, configured to load and activate the acquisition task, and associate the acquisition task with the relevant electricity meters; an electricity meter communication priority scheduling module 130, configured to perform task priority reading and acquisition on each electricity meter based on the communication ability ranking table; and an electricity meter task loop execution module 140, configured to sequentially repeat the corresponding actions of the electricity meter task association management module and the electricity meter communication priority scheduling module until all tasks are completed.

[0066] In one embodiment, the electricity meter communication efficiency analysis module includes: a communication ability adaptive evaluation unit, configured to respectively perform communication ability semantic anchor point query adaptive decision response on the set of communication ability influencing factors of each electricity meter and the set of communication ability influencing factors of all electricity meters to obtain a set of communication ability evaluation results as the communication ability evaluation results of each electricity meter; and a communication ability ranking unit, configured to sort the communication abilities of each electricity meter based on the set of communication ability evaluation results to obtain the communication ability ranking table.

[0067] In one embodiment, the communication capability adaptive evaluation unit includes: an influencing factor low-dimensional embedding encoding subunit, configured to perform low-dimensional embedding encoding on each communication capability influencing factor in the set of communication capability influencing factors of each electric energy meter to obtain a set of embedded encoding vectors of the communication capability influencing factors of each electric energy meter; a communication capability multi-level encoding subunit, configured to perform multi-level encoding on the set of embedded encoding vectors of the communication capability influencing factors of each electric energy meter to obtain a set of multi-level deep implicit encoding vectors of the communication capability of the electric energy meter; a target electric energy meter vector extraction subunit, configured to extract the multi-level deep implicit encoding vector of the communication capability of the target electric energy meter from the set of multi-level deep implicit encoding vectors of the communication capability of the electric energy meter as the query vector of the communication capability of the electric energy meter; a communication capability adaptive query response processing subunit, configured to perform query - communication capability set semantic anchor adaptive query response on the query vector of the communication capability of the electric energy meter and the set of multi-level deep implicit encoding vectors of the communication capability of the electric energy meter to obtain a semantic query response encoding vector of the communication capability of the target electric energy meter; a communication capability evaluation result determination subunit, configured to obtain a communication capability evaluation result based on the semantic query response encoding vector of the communication capability of the target electric energy meter, where the communication capability evaluation result is the probability that the target electric energy meter is the electric energy meter with the best communication capability; an evaluation result set acquisition subunit, configured to enable the target electric energy meter vector extraction subunit, the communication capability adaptive query response processing subunit, and the communication capability evaluation result determination subunit to sequentially repeat corresponding actions to obtain a set of communication capability evaluation results.

[0068] Here, those skilled in the art can understand that the specific operations of the above-mentioned various modules in the intelligent electric energy meter acquisition task scheduling and management system have been introduced in detail in the description of the intelligent electric energy meter acquisition task scheduling and management method referred to above Figures 1 to 5 and thus, the repeated description thereof will be omitted.

[0069] As described above, the above is only the specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all of them should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for scheduling and managing the acquisition tasks of an intelligent electricity meter, characterized in that, Including: S1: Obtain the set of communication ability influencing factors of each electric energy meter, and sort the communication abilities of each electric energy meter based on the set of communication ability influencing factors of each electric energy meter to obtain a communication ability ranking table; S2: Load and activate the acquisition task, and associate the acquisition task with the relevant electric energy meters; S3: Based on the communication ability ranking table, perform task priority reading and acquisition for each electric energy meter; S4: Repeat steps S2 - S3 until all tasks are completed; where, S1 includes: S11: Respectively perform communication ability semantic anchor point query adaptive decision response on the set of communication ability influencing factors of each electric energy meter and the set of communication ability influencing factors of all electric energy meters to obtain a set of communication ability evaluation results as the communication ability evaluation results of each electric energy meter; S12: Sort the communication abilities of each electric energy meter based on the set of communication ability evaluation results to obtain the communication ability ranking table.

2. The intelligent electric energy meter acquisition task scheduling and management method according to claim 1, wherein S11 includes: S111: Perform low-dimensional embedding coding on each communication ability influencing factor in the set of communication ability influencing factors of each electric energy meter to obtain a set of embedded coding vectors of communication ability influencing factors of each electric energy meter; S112: Perform multi-level coding on the set of embedded coding vectors of communication ability influencing factors of each electric energy meter to obtain a set of multi-level deep implicit coding vectors of electric energy meter communication ability; S113: Extract the multi-level deep implicit coding vector of the target electric energy meter's communication ability from the set of multi-level deep implicit coding vectors of electric energy meter communication ability as the communication ability query vector of the electric energy meter; S114: Perform query - communication ability set semantic anchor point adaptive query response on the communication ability query vector and the set of multi-level deep implicit coding vectors of electric energy meter communication ability to obtain the semantic query response coding vector of the target electric energy meter's communication ability; S115: Based on the semantic query response coding vector of the target electric energy meter's communication ability, obtain the communication ability evaluation result, and the communication ability evaluation result is the probability that the target electric energy meter is the electric energy meter with the best communication ability; S116: Loop through steps S113 to S115 to obtain the set of communication ability evaluation results.

3. The intelligent electric energy meter acquisition task scheduling and management method according to claim 2, wherein S112 includes: Perform multi-level full connection coding based on a multi-layer perceptron model on the set of embedded coding vectors of communication ability influencing factors of each electric energy meter to obtain the set of multi-level deep implicit coding vectors of electric energy meter communication ability.

4. The intelligent electric energy meter acquisition task scheduling and management method according to claim 3, characterized in that, The S114 includes: S1141: Semantically response-anchoring each multi-level deep implicit coding vector of the electricity meter communication ability in the set of the electricity meter communication ability query vector and the multi-level deep implicit coding vectors of the electricity meter communication ability to obtain a set of query-electricity meter communication ability feature response-anchoring coding matrices; S1142: Calculating the decision-anchor adaptive splicing factors of each query-electricity meter communication ability feature response-anchoring coding matrix in the set of the query-electricity meter communication ability feature response-anchoring coding matrices to obtain a set of query-electricity meter communication ability decision-anchor adaptive splicing weight factors; S1143: Based on the set of the query-electricity meter communication ability decision-anchor adaptive splicing weight factors, fusing the set of the query-electricity meter communication ability feature response-anchoring coding matrices to obtain the target electricity meter communication ability semantic query response coding vector.

5. The intelligent electric energy meter acquisition task scheduling and management method according to claim 4, wherein The S1141 includes: Extracting the deep implicit feature fully-connected coding of the electricity meter communication ability query vector to obtain the electricity meter communication ability query deep implicit coding vector; Extracting the enhanced deep implicit feature fully-connected coding of each multi-level deep implicit coding vector of the electricity meter communication ability in the set of the multi-level deep implicit coding vectors of the electricity meter communication ability to obtain a set of electricity meter communication ability multi-level deep implicit coding enhanced vectors; Respectively performing the electricity meter communication ability semantic response decision-anchoring on the electricity meter communication ability query deep implicit coding vector and each electricity meter communication ability multi-level deep implicit coding enhanced vector in the set of the electricity meter communication ability multi-level deep implicit coding enhanced vectors to obtain the set of the query-electricity meter communication ability feature response-anchoring coding matrices.

6. The intelligent electric energy meter acquisition task scheduling and management method according to claim 5, wherein The S1142 includes: S11421: Based on the feature distribution of each query-electricity meter communication ability feature response-anchoring coding matrix in the set of the query-electricity meter communication ability feature response-anchoring coding matrices, determining the decision-anchor adaptive splicing factors of each query-electricity meter communication ability feature response-anchoring coding matrix to obtain a set of query-electricity meter communication ability decision-anchor adaptive splicing factors; S11422: Weighting the set of the query-electricity meter communication ability decision-anchor adaptive splicing factors based on the Softmax function to obtain the set of the query-electricity meter communication ability decision-anchor adaptive splicing weight factors.

7. The intelligent electricity meter acquisition task scheduling and management method according to claim 6, wherein The S11421 includes: Calculating the corresponding query-electricity meter communication ability decision-anchor adaptive splicing factor of the query-electricity meter communication ability feature response-anchoring coding matrix based on the maximum value, mean value, variance, difference amplification coefficient of the query-electricity meter communication ability feature response-anchoring coding matrix and the electricity meter communication ability drift coefficient; wherein, the electricity meter communication ability drift coefficient is obtained by performing local-global importance score smoothing transition processing on the query-electricity meter communication ability feature response-anchoring coding matrix.

8. The intelligent electric energy meter acquisition task scheduling and management method according to claim 7, characterized in that The S115 includes: inputting the semantic query response encoding vector of the communication ability of the target electricity meter into a communication ability evaluation module based on a classifier to obtain the communication ability evaluation result, where the communication ability evaluation result is the probability that the target electricity meter is the electricity meter with the best communication ability.

9. An intelligent electric energy meter acquisition task scheduling and management system, characterized in that, It includes: An electricity meter communication efficiency analysis module, configured to obtain a set of communication ability influencing factors for each electricity meter, and rank the communication abilities of the electricity meters based on the set of communication ability influencing factors for each electricity meter to obtain a communication ability ranking table; An electricity meter task association management module, configured to load and activate a collection task, and associate the collection task with relevant electricity meters; An electricity meter communication priority scheduling module, configured to perform task priority reading and collection on each electricity meter based on the communication ability ranking table; an electricity meter task loop execution module, configured to repeatedly execute corresponding actions by the electricity meter task association management module and the electricity meter communication priority scheduling module in sequence until all tasks are completed; wherein, the electricity meter communication efficiency analysis module includes: a communication ability adaptive evaluation unit, configured to perform communication ability semantic anchor point query adaptive decision response on the set of communication ability influencing factors for each electricity meter and the set of communication ability influencing factors for all electricity meters respectively to obtain a set of communication ability evaluation results as the communication ability evaluation results for each electricity meter; a communication ability ranking unit, configured to rank the communication abilities of the electricity meters based on the set of communication ability evaluation results to obtain the communication ability ranking table.

10. The intelligent electric energy meter acquisition task scheduling and management system according to claim 9, characterized in that, The communication capability adaptive evaluation unit includes: an influencing factor low-dimensional embedding encoding subunit, configured to perform low-dimensional embedding encoding on each communication capability influencing factor in the set of communication capability influencing factors of each electric energy meter to obtain a set of communication capability influencing factor embedding encoding vectors of each electric energy meter; a communication capability multi-level encoding subunit, configured to perform multi-level encoding on the set of communication capability influencing factor embedding encoding vectors of each electric energy meter to obtain a set of multi-level deep implicit encoding vectors of the electric energy meter communication capability; a target electric energy meter vector extraction subunit, configured to extract the multi-level deep implicit encoding vector of the target electric energy meter from the set of multi-level deep implicit encoding vectors of the electric energy meter communication capability as the electric energy meter communication capability query vector; a communication capability adaptive query response processing subunit, configured to perform query-communication capability set semantic anchor point adaptive query response on the electric energy meter communication capability query vector and the set of multi-level deep implicit encoding vectors of the electric energy meter communication capability to obtain a target electric energy meter communication capability semantic query response encoding vector; a communication capability evaluation result determination subunit, configured to obtain a communication capability evaluation result based on the target electric energy meter communication capability semantic query response encoding vector, where the communication capability evaluation result is the probability that the target electric energy meter is the electric energy meter with the best communication capability; an evaluation result set acquisition subunit, configured to have the target electric energy meter vector extraction subunit, the communication capability adaptive query response processing subunit, and the communication capability evaluation result determination subunit sequentially repeat corresponding actions to obtain a set of the communication capability evaluation results.

Images