Distribution area meter reading link AI score and time slot level parameter issuing system and method

Abstract

The invention discloses a low-voltage distribution area meter reading link AI score and time slot level parameter issuing system and method. The system comprises an edge node, a link feature acquisition module, an AI score module, a time slot level parameter issuing module and an electric energy meter terminal. The link feature acquisition module is used for acquiring link state information in real time at an edge node side; the AI scoring module is used for processing the multi-dimensional features through a lightweight artificial intelligence model according to the link state information, and outputting the probability score and confidence of the link transmission success rate; and the time slot level parameter issuing module is used for generating a dynamic parameter combination according to the probability score and the confidence coefficient of the link transmission success rate, distributing the parameter combination to a specific time slot of the meter reading frame, and issuing parameters to the electric energy meter terminal at a time slot boundary through a clock synchronization mechanism. The technical problems that an existing system is low in one-time copying rate, high in retransmission frequency, long in end-to-end time delay and difficult to meet the requirements for high reliability, low time delay and balanced coverage of an intelligent power grid are solved.

Description

Technical Field

[0001] This invention relates to the field of low-voltage distribution radio area technology, and in particular to a system and method for AI scoring and time slot-level parameter distribution of distribution radio area meter reading links. Background Technology

[0002] With the rapid development of smart grids and distribution automation, smart meters have been widely deployed in low-voltage distribution substations, and automatic meter reading is completed through concentrators or acquisition terminals. Among these, power line carrier communication, which utilizes existing power lines for data transmission without the need for additional wiring, is flexible in deployment and low in cost, making it the main communication method for automatic meter reading in low-voltage distribution substations. However, in actual operation, the power line channels in low-voltage substations are complex and variable. Link quality is affected by various factors such as electrical noise, harmonic interference, impedance changes, and user electrical start-up and shutdown, exhibiting significant time-varying and variability. The communication quality varies significantly between different users, and some users' links are in weak coverage areas for extended periods, leading to problems such as insufficient first-pass yield, excessive retransmissions, and poor end-to-end delay convergence.

[0003] Most existing technologies employ fixed parameter configuration, uniformly setting parameters such as modulation scheme, coding level, and transmit power within the entire substation area. While simple to implement, this approach cannot differentiate for different link environments, limiting its ability to protect weak links. Furthermore, traditional systems often use single thresholds such as signal-to-noise ratio or bit error rate as the basis for judgment. Additionally, parameter adjustments are typically made at the hourly or substation-wide level, failing to respond quickly to transient interference or tributary degradation, resulting in weak links not receiving timely relief and limiting overall meter reading performance. Moreover, current policy governance and resource allocation mechanisms also have shortcomings. On one hand, parameter distribution lacks versioning and auditing mechanisms; improper parameter configuration can easily trigger widespread communication jitter, lacking controllable and traceable guarantees. On the other hand, existing solutions often neglect fairness in resource allocation, with strong link nodes consistently dominating due to better conditions, while weak link nodes may remain in a state of failure for extended periods, resulting in a structural imbalance of "the strong getting stronger and the weak getting weaker." These problems result in insufficient intelligence in the optimization of existing low-voltage distribution radio area meter reading links, a lack of refined control methods, and effective protection for weak links.

[0004] Patent application number 202510617428.9 discloses a method and system for constructing intelligent agents based on an agency workflow. It defines core functional modules of the intelligent agent and assigns unique identifiers, establishes communication protocol standards between modules, clarifies message formats and priority rules, allocates computing resources to each module and sets elastic scaling strategies, generates module dependency graphs and architecture metadata files, detects the loss and delay of multimodal input data, generates compensation features based on historical context, calculates the quality confidence of each modality, and performs dynamic weighted fusion. Existing technologies cannot solve the above-mentioned technical problems; therefore, there is an urgent need to propose a system and method for AI scoring and time-slot-level parameter distribution in distribution radio area meter reading links. Summary of the Invention

[0005] The main objective of this invention is to propose an AI scoring and time slot-level parameter distribution system and method for distribution area meter reading links, aiming to solve the technical problems of existing systems, such as low first-time reading rate, high retransmission frequency, and long end-to-end time delay, which make it difficult to meet the requirements of smart grids for high reliability, low latency, and balanced coverage.

[0006] To achieve the above objectives, the present invention provides a low-voltage distribution transformer area meter reading link AI scoring and time slot-level parameter distribution system, wherein the low-voltage distribution transformer area meter reading link AI scoring and time slot-level parameter distribution system includes an edge node, a link feature acquisition module, an AI scoring module, a time slot-level parameter distribution module, and an energy meter terminal connected in sequence.

[0007] The link feature acquisition module is used to acquire link status information in real time at the edge node side;

[0008] The AI ​​scoring module is used to process multi-dimensional features based on link status information using a lightweight artificial intelligence model, and output a probability score and confidence level for link transmission success rate.

[0009] The time slot-level parameter distribution module is used to generate dynamic parameter combinations based on the probability score and confidence level of the link transmission success rate, and allocate the parameter combinations to specific time slots of the meter reading frame. The parameters are then distributed to the energy meter terminal at the time slot boundary through a clock synchronization mechanism.

[0010] One preferred embodiment is that the link status information includes signal-to-noise ratio, bit error rate, received signal strength, retransmission count, spectral energy density distribution, harmonic interference components, and line impedance disturbance.

[0011] In one preferred embodiment, the link state information represents link characteristics in vector form, specifically as follows:

[0012]

[0013] in, For the first Link status information for each user. For the first Signal-to-noise ratio of each user link For the first Bit error rate of individual user links, For the first The received signal strength of the individual user link, For the first Retransmission count for each user link For the first Spectral energy density distribution of individual user links, For the first Harmonic interference components of the user link For the first Line impedance disturbance of individual user links.

[0014] One preferred embodiment is that the lightweight artificial intelligence model includes one or more of logistic regression, decision tree, random forest, or lightweight Transformer network.

[0015] One preferred embodiment is that the AI ​​scoring module uses the Platt calibration method to calibrate the output link transmission success rate, specifically as follows:

[0016]

[0017] in, To improve the success rate of link transmission after calibration, for, For the first Link status information for each user. For calibration parameters, This represents the link transmission success rate before calibration.

[0018] In one preferred embodiment, the AI ​​scoring module uses a temperature scaling method to calibrate the output link transmission success rate.

[0019] One preferred embodiment is that the parameter combination includes modulation scheme, coding redundancy, repetition count, and transmit power.

[0020] One preferred embodiment is that the parameter combination is allocated to a specific time slot of the meter reading frame, wherein the specific time slot does not exceed the frame capacity of the meter reading frame, specifically as follows:

[0021]

[0022] in, To be assigned to the The number of time slots for each user's link. This represents the total time slot capacity per frame. This represents the total number of links for each user.

[0023] One preferred embodiment is that the low-voltage distribution radio area meter reading link AI scoring and time slot-level parameter distribution system also includes a parameter library and policy governance module. The parameter library and policy governance module is used to store versioned parameter templates and provide signature verification, gray release, atomic rollback and audit logs for parameter distribution.

[0024] A method for distributing AI scoring and time slot-level parameter distribution for distribution area meter reading links, including the aforementioned distribution area meter reading link AI scoring and time slot-level parameter distribution system, includes the following steps:

[0025] S1. Collect the link status information of each user meter on the edge node side;

[0026] S2. The link status information is processed through a lightweight artificial intelligence model to obtain the probability score and confidence level of the link transmission success rate.

[0027] S3. Based on the probability score and confidence level of the link transmission success rate, generate dynamic parameter combinations and allocate the parameter combinations to specific time slots of the meter reading frame. Send the parameters to the energy meter terminal at the time slot boundary through the clock synchronization mechanism.

[0028] S4. Implement version control, canary release, and atomic rollback for parameter templates, and record audit logs.

[0029] In the above technical solution of the present invention, the low-voltage distribution substation meter reading link AI scoring and time slot-level parameter distribution system includes an edge node, a link feature acquisition module, an AI scoring module, a time slot-level parameter distribution module, and an energy meter terminal connected in sequence. The link feature acquisition module is used to acquire link status information in real time at the edge node. The AI ​​scoring module is used to process multi-dimensional features based on the link status information using a lightweight artificial intelligence model, and output the probability score and confidence level of the link transmission success rate. The time slot-level parameter distribution module is used to generate dynamic parameter combinations based on the probability score and confidence level of the link transmission success rate, and allocate the parameter combinations to specific time slots of the meter reading frame, and distribute the parameters to the energy meter terminal at the time slot boundary through a clock synchronization mechanism. The present invention solves the technical problems of existing systems, such as low first-time reading rate, high retransmission frequency, and long end-to-end time delay, which make it difficult to meet the requirements of smart grids for high reliability, low latency, and balanced coverage.

[0030] In this invention, by collecting multi-dimensional features and introducing a lightweight artificial intelligence model, real-time scoring and confidence estimation of link quality are achieved, thereby significantly improving the accuracy, stability and interpretability of link status judgment.

[0031] In this invention, the time slot-level parameter distribution mechanism based on the meter reading frame structure dynamically combines the modulation method, coding redundancy, repetition count, and transmission power within a time slot at the second or minute level, thereby achieving rapid relief for weak links, reducing the number of retransmissions, and converging end-to-end latency. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0033] Figure 1 This is a schematic diagram of a low-voltage distribution radio area meter reading link AI scoring and time slot-level parameter distribution system according to an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram of an AI scoring and time slot-level parameter distribution method for a low-voltage distribution radio area meter reading link according to an embodiment of the present invention.

[0035] The realization of the objective, functional characteristics and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0038] See Figure 1 According to one aspect of the present invention, the present invention provides a low-voltage distribution transformer area meter reading link AI scoring and time slot level parameter distribution system, wherein the low-voltage distribution transformer area meter reading link AI scoring and time slot level parameter distribution system includes an edge node, a link feature acquisition module, an AI scoring module, a time slot level parameter distribution module and an energy meter terminal connected in sequence.

[0039] The link feature acquisition module is used to acquire link status information in real time at the edge node side;

[0040] The AI ​​scoring module is used to process multi-dimensional features based on link status information using a lightweight artificial intelligence model, and output a probability score and confidence level for link transmission success rate.

[0041] The time slot-level parameter distribution module is used to generate dynamic parameter combinations based on the probability score and confidence level of the link transmission success rate, and allocate the parameter combinations to specific time slots of the meter reading frame. The parameters are then distributed to the energy meter terminal at the time slot boundary through a clock synchronization mechanism.

[0042] Specifically, in this embodiment, the present invention, through the low-voltage distribution transformer area meter reading link AI scoring and time slot-level parameter distribution system, comprises an edge node, a link feature acquisition module, an AI scoring module, a time slot-level parameter distribution module, a parameter library and strategy governance module, and an energy meter terminal connected in sequence, forming a closed-loop structure of perception-decision-execution-feedback. The edge node, as the core hub of the system, is responsible for coordinating the acquisition of link data, the inference and calculation of the artificial intelligence model, and the dynamic distribution of parameters, while maintaining data interaction with the energy meter to ensure the orderly execution of the entire transformer area meter reading process. The edge node is responsible for receiving data reports from the energy meter terminal on the one hand, and coordinating key links such as link feature acquisition, AI scoring, and parameter distribution on the other hand.

[0043] Specifically, in this embodiment, the link feature acquisition module acquires link status information in real time at the edge node side. The link status information includes signal-to-noise ratio, bit error rate, received signal strength, retransmission count, spectral energy density distribution, harmonic interference components, and line impedance disturbance. The link status information reflects the dynamic characteristics of the link in the time domain and also characterizes its complex disturbance in the frequency domain, thereby providing sufficient data support for subsequent intelligent judgment and comprehensively depicting the channel status.

[0044] Specifically, in this embodiment, the link state information represents link features in vector form, specifically as follows:

[0045]

[0046] in, For the first Link status information for each user. For the first Signal-to-noise ratio of each user link For the first Bit error rate of individual user links, For the first The received signal strength of the individual user link, For the first Retransmission count for each user link For the first Spectral energy density distribution of individual user links, For the first Harmonic interference components of the user link For the first Line impedance disturbance of individual user links.

[0047] Specifically, in this embodiment, the collected link status information is input into a lightweight artificial intelligence model. This lightweight AI model includes one or more of logistic regression, decision trees, random forests, or lightweight Transformer networks, outputting a probability score for the link success rate. The system further estimates the confidence level of the results through temperature scaling or Platt calibration to ensure accurate scoring, reliability, and interpretability. This module is the core of weak link identification and dynamic scheduling. The output link transmission success rate is:

[0048]

[0049] in, This represents the output link transmission success rate.

[0050] Specifically, in this embodiment, to increase the reliability of the results, the AI ​​scoring module uses the Platt calibration method to calibrate the output link transmission success rate, specifically as follows:

[0051]

[0052] in, To improve the success rate of link transmission after calibration, for, For the first Link status information for each user. For calibration parameters, This represents the link transmission success rate before calibration.

[0053] Specifically, in this embodiment, the AI ​​scoring module uses a temperature scaling method to calibrate the output link transmission success rate.

[0054] Specifically, in this embodiment, the time slot-level parameter distribution module allocates the parameter combination (modulation mode, coding redundancy, repetition count, and transmit power) to a specific time slot of the meter reading frame based on the probability score and confidence level of the link transmission success rate. The system ensures that the parameters take effect immediately at the time slot boundary through a timestamp interface and clock synchronization mechanism. For weak links with poor scores, the module prioritizes distributing enhanced configurations (parameter combinations of low-order modulation, high redundancy, and high transmit power). For links with scores reaching a set threshold, the default settings are maintained, achieving differentiated protection and efficient resource utilization. The allocation of parameter combinations to specific time slots of the meter reading frame, where the specific time slot does not exceed the frame capacity of the meter reading frame, specifically involves:

[0055]

[0056] in, To be assigned to the The number of time slots for each user's link. This represents the total time slot capacity per frame. This represents the total number of links for each user.

[0057] Specifically, in this embodiment, the low-voltage distribution transformer area meter reading link AI scoring and time slot-level parameter distribution system also includes a parameter library and policy governance module. This module stores versioned parameter templates and provides signature verification, canary release, atomic rollback, and audit logs for parameter distribution. The parameter library and policy governance module has a built-in parameter library covering different combinations of communication parameter templates and provides version management, signature verification, canary release, and atomic rollback mechanisms to ensure that parameter and policy adjustments are safe, stable, and controllable. All parameter changes and execution processes are recorded through audit logs, achieving end-to-end traceability and compliance. All parameter combinations and policy adjustments are subject to signature verification and are gradually rolled out through canary release to ensure the stability of parameter adjustments. When performance degradation occurs, the system can quickly recover to the previous version through an atomic rollback mechanism. Simultaneously, all parameter change processes generate audit logs, achieving end-to-end traceability from policy formulation and distribution to execution.

[0058] Specifically, in this embodiment, the present invention establishes a versioned parameter library, performs signature verification, canary release and atomic rollback on policy changes, and generates full audit logs, making the policy execution process controllable and traceable, thereby effectively reducing operational risks and ensuring system stability.

[0059] Specifically, in this embodiment, during the allocation of time slots and parameter resources, the system prioritizes links with lower scores and allocates redundant resources to them to ensure basic reachability. Simultaneously, resource consumption limits are set for strong links with higher scores to avoid uneven resource allocation leading to a "stronger gets stronger, weaker gets weaker" phenomenon, thereby achieving overall coverage balance and fairness. In resource scheduling, fairness constraints are added to the system to ensure that weak links receive at least the minimum required transmission resources.

[0060]

[0061] in, For the first Resource allocation for each user's link. For minimum transmission resources, A set of weak links;

[0062] At the same time, a cap is set on strong links to prevent them from excessively consuming resources and to ensure overall system fairness.

[0063]

[0064] in, for, for, for.

[0065] Specifically, in this embodiment, the present invention introduces fairness constraints and a safety net mechanism into the scheduling mechanism, giving priority to allocating enhanced resources to links with lower scores and limiting the resource ratio of strong links, thereby achieving key protection for weak links and balance of overall coverage, and avoiding structural imbalance.

[0066] Specifically, in this embodiment, the electricity meter terminal acts as a link terminal, receiving time slot-level parameters sent by the edge node and reporting data. Through differentiated configuration strategies, if the transmission reliability of the link is improved, the overall first-time meter reading rate and end-to-end performance will be improved.

[0067] See Figure 2 According to another aspect of the present invention, the present invention provides a method for AI scoring and time slot-level parameter distribution of a low-voltage distribution transformer area meter reading link, wherein the method for AI scoring and time slot-level parameter distribution of a low-voltage distribution transformer area meter reading link includes the following steps:

[0068] S1. Collect the link status information of each user meter on the edge node side;

[0069] S2. The link status information is processed through a lightweight artificial intelligence model to obtain the probability score and confidence level of the link transmission success rate.

[0070] S3. Based on the probability score and confidence level of the link transmission success rate, generate dynamic parameter combinations and allocate the parameter combinations to specific time slots of the meter reading frame. Send the parameters to the energy meter terminal at the time slot boundary through the clock synchronization mechanism.

[0071] S4. Implement version control, canary release, and atomic rollback for parameter templates, and record audit logs.

[0072] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. All equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A distribution radio area meter reading link AI scoring and time slot-level parameter distribution system, characterized in that, This includes, in sequence, edge nodes, a link feature acquisition module, an AI scoring module, a time slot-level parameter distribution module, and an energy meter terminal; The link feature acquisition module is used to acquire link status information in real time at the edge node side; The AI ​​scoring module is used to process multi-dimensional features based on link status information using a lightweight artificial intelligence model, and output a probability score and confidence level for link transmission success rate. The time slot-level parameter distribution module is used to generate dynamic parameter combinations based on the probability score and confidence level of the link transmission success rate, and allocate the parameter combinations to specific time slots of the meter reading frame. The parameters are then distributed to the energy meter terminal at the time slot boundary through a clock synchronization mechanism.

2. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to claim 1, characterized in that, The link status information includes signal-to-noise ratio, bit error rate, received signal strength, retransmission count, spectral energy density distribution, harmonic interference components, and line impedance disturbance.

3. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to claim 2, characterized in that, The link state information represents link features in vector form, specifically: ； in, For the first Link status information for each user. For the first Signal-to-noise ratio of each user link For the first Bit error rate of individual user links, For the first The received signal strength of the individual user link, For the first Retransmission count for each user link For the first Spectral energy density distribution of individual user links, For the first Harmonic interference components of the user link For the first Line impedance disturbance of individual user links.

4. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to any one of claims 1-3, characterized in that, The lightweight artificial intelligence model includes one or more of logistic regression, decision tree, random forest, or lightweight Transformer network.

5. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to any one of claims 1-3, characterized in that, The AI ​​scoring module uses the Platt calibration method to calibrate the output link transmission success rate, specifically: ； in, To improve the success rate of link transmission after calibration, for, For the first Link status information for each user. For calibration parameters, This represents the link transmission success rate before calibration.

6. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to any one of claims 1-3, characterized in that, The AI ​​scoring module uses a temperature scaling method to calibrate the output link transmission success rate.

7. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to any one of claims 1-3, characterized in that, The parameter combination includes modulation scheme, coding redundancy, repetition count, and transmit power.

8. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to any one of claims 1-3, characterized in that, The allocation of parameter combinations to specific time slots in the meter reading frame, wherein the specific time slots do not exceed the frame capacity of the meter reading frame, specifically refers to: ； in, To be assigned to the The number of time slots for each user's link. This represents the total time slot capacity per frame. This represents the total number of links for each user.

9. The distribution area meter reading link AI scoring and time slot-level parameter distribution system according to any one of claims 1-3, characterized in that, The low-voltage distribution area meter reading link AI scoring and time slot-level parameter distribution system also includes a parameter library and policy governance module. The parameter library and policy governance module is used to store versioned parameter templates and provide signature verification, canary release, atomic rollback and audit logs for parameter distribution.

10. A method for distributing AI scoring and time slot-level parameter distribution for distribution area meter reading links, comprising the distribution area meter reading link AI scoring and time slot-level parameter distribution system according to any one of claims 1-7, characterized in that, Includes the following steps: S1. Collect the link status information of each user meter on the edge node side; S2. The link status information is processed through a lightweight artificial intelligence model to obtain the probability score and confidence level of the link transmission success rate. S3. Based on the probability score and confidence level of the link transmission success rate, generate dynamic parameter combinations and allocate the parameter combinations to specific time slots of the meter reading frame. Send the parameters to the energy meter terminal at the time slot boundary through the clock synchronization mechanism. S4. Implement version control, canary release, and atomic rollback for parameter templates, and record audit logs.

Images