Low-voltage transparent power distribution network construction method based on multi-stage collaborative architecture

Abstract

The invention discloses a low-voltage transparent power distribution network construction method based on a multi-stage collaborative architecture, and belongs to the technical field of power distribution network intellectualization. According to the low-voltage transparent power distribution network construction method based on the multi-stage collaborative architecture, total factor perception and intelligent regulation and control of a low-voltage power distribution network are realized by constructing a scheduling decision-edge collaboration-sensing perception three-stage collaborative architecture; the method comprises the following steps: firstly, acquiring whole-network electrical data by sensing layer equipment; secondly, the edge collaboration layer automatically identifies and updates a station-line-transformer-household topological relation based on space-time similarity calculation of multi-moment electrical parameters; and then the scheduling decision-making layer generates an optimization control strategy according to the panoramic data, and after the optimization control strategy is decomposed by the edge collaboration layer, the sensing layer executes accurate regulation and control to form considerable, measurable, controllable and adjustable closed-loop management. According to the method, the problems of difficult topology maintenance, poor regulation and control collaboration, high novel source load access risk and the like of the low-voltage distribution network are solved, and the power supply reliability and the operation and maintenance efficiency are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of power system distribution technology, specifically to a method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture, and in particular to a digital and intelligent construction and operation management method that achieves real-time data perception, automatic topology identification, intelligent control, and rapid fault handling of a low-voltage distribution network through three-level collaboration of the scheduling layer, edge collaboration layer, and sensing layer. Background Technology

[0002] As the "last mile" connecting the main power grid and end users, the distribution network is a vital public infrastructure for ensuring reliable power supply and serving people's livelihoods and economic development. With the large-scale integration of new power sources such as distributed photovoltaics and electric vehicle charging stations, low-voltage distribution networks are evolving from traditional unidirectional power supply networks to bidirectional interactive active networks. Their operating characteristics are becoming increasingly complex, placing higher demands on refined management and intelligent control.

[0003] However, the planning, construction, operation and maintenance management of low-voltage distribution networks still face the following prominent technical problems: 1. Weak asset data foundation and unclear topology. The low-voltage side has a large number of devices, diverse types, and frequent changes, making it difficult for the existing management system to dynamically and accurately update ledgers and network layout diagrams. Many newly added branch switches, miniature circuit breakers, and other equipment are not included in unified management, and the quality of data already entered into the system is inconsistent. The "station-line-transformer-customer" topology relies on manual maintenance with low accuracy, resulting in an unclear network infrastructure and an inability to provide reliable data support for operational analysis and control decisions.

[0004] 2. Insufficient professional collaboration and a disconnect between management mechanisms and technical support. Management barriers exist between professional lines such as planning, operation and maintenance, marketing, and scheduling. Business standards and collaborative processes are incomplete, and horizontal collaboration and vertical integration mechanisms are not functioning smoothly. At the same time, the lack of a unified technical platform to integrate multi-source heterogeneous data and business needs leads to a widespread phenomenon of "unmanageable and inaccessible" situations, making it difficult to resolve problems in a closed loop.

[0005] 3. Low level of intelligence, lacking panoramic perception and proactive control capabilities. Existing low-voltage side monitoring has incomplete coverage and limited perception capabilities. Data acquisition is discrete and lacks timeliness, failing to achieve real-time panoramic perception of electrical quantities, equipment status, and environmental information. More importantly, the system lacks the ability to fuse and analyze massive amounts of data and closed-loop control methods, making it unable to effectively support advanced applications such as load monitoring, rapid fault isolation, voltage and reactive power regulation, and distributed energy group control. It also struggles to cope with problems such as overload, voltage exceeding limits, frequent faults, and slow response to customer requests.

[0006] Therefore, there is an urgent need for a systematic technical solution that can build a real-time sensing system covering all elements of the low-voltage distribution network, realize the automatic identification and dynamic updating of topological relationships, and rely on a hierarchical and collaborative control architecture to solve technical problems such as data opacity, lack of coordination in control, and lack of closed-loop business, and ultimately build a transparent low-voltage smart distribution network that is "observable, measurable, controllable, and adjustable". Summary of the Invention

[0007] To address the aforementioned issues, this invention proposes a method for analyzing test voltage and insulation withstand levels. This method constructs a three-tiered collaborative technology system: "scheduling decision-making - edge collaboration - sensing perception." At the sensing layer, comprehensive electrical data is collected through intelligent terminals and sensing devices. At the edge layer, collaborative units with local computing capabilities are deployed to handle data aggregation, automatic topology identification, and regional autonomous control. At the scheduling layer, panoramic visualization and global optimization decisions are achieved based on multi-source data fusion and advanced algorithms. Through this three-tiered collaboration, this invention enables automatic identification and dynamic updating of low-voltage distribution network topology, supporting advanced applications such as precise load monitoring, rapid fault isolation, and distributed photovoltaic group control. Ultimately, it establishes a transparent low-voltage smart distribution network that is "observable, measurable, controllable, and adjustable," significantly improving operation and maintenance efficiency and power supply reliability.

[0008] The technical solution adopted in this invention is as follows: The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture is characterized by the following steps: Step 1, System Architecture Design and Deployment: Adopt a multi-level collaborative architecture of scheduling decision-making, Hongrui intelligent procurement collaboration and sensor perception, integrate the low-voltage dispatch and command center organizational model, establish a linkage between technology and management systems, and realize the overall framework of low-voltage transparent distribution network; Step 2, Data Acquisition and Acquisition: Using the Hongrui terminal, RS485 communication and HPLC carrier are used to collect household meter ledger information and real-time voltage, current, active power and reactive power data from devices such as distribution transformer monitoring terminals, incoming line switches, branch switches, miniature circuit breakers, and photovoltaic coordinators, and store them in the terminal. Step 3, Topology Relationship Identification and Maintenance: Based on the spatiotemporal vector judgment method, calculate the similarity matrix of voltage, current and power between devices, automatically identify and verify the station-line-transformer-customer link parameters, and draw a global relationship diagram by combining customer meter ledgers and GIS information to realize the updating and maintenance of topology relationships; Step 4, Control System Construction: Establish a three-level control system of global-regional-device. The scheduling layer generates control strategies, the edge layer Hongrui terminal coordinates regional devices to perform power allocation, and the sensing layer achieves precise control through photovoltaic coordinators and smart circuit breakers. In case of a fault, the region is isolated in real time and the load is transferred to shorten the handling time. Step 5, adjustment function implementation: Hongrui terminal collects real-time load and photovoltaic output data and sends it to the dispatch master station. The latter integrates group dispatch and control algorithm to generate control strategy and sends it to the terminal. The photovoltaic coordinator flexibly controls the inverter output and dynamically optimizes and solves the problems of reverse overload and voltage over-limit.

[0009] Furthermore, in step 1, the dispatch decision-making layer relies on the dispatch center to integrate multi-source data such as distribution network GIS topology, ledgers, and loads to build a four-way control platform, which generates global optimization instructions, visualizes the overall operation status, and performs cross-professional collaborative dispatch. The Hongrui intelligent procurement and collaboration layer deploys Hongrui terminals with local computing, local autonomy, and dispatch collaboration capabilities, receives dispatch instructions, and optimizes power distribution and voltage regulation within the region. The sensing and perception layer installs miniature circuit breakers, distribution area branch circuit breakers, distribution area intelligent transformer terminals, and photovoltaic protocol converters with carrier communication to collect all data on equipment status, electrical parameters, and environmental data and execute instructions.

[0010] Furthermore, in step 2, the Hongrui terminal obtains the household meter ledger information and 15-minute voltage, current, and active power data from the distribution transformer monitoring terminal via RS485 communication, obtains real-time electrical data from the incoming line switch and branch switch, and obtains real-time data from the miniature circuit breaker, photovoltaic coordinator, and sectionalizing switch via HPLC carrier wave, thus achieving full data perception.

[0011] Furthermore, in step 3, a matrix based on time and device is established, the similarity of voltage, current and power is calculated, and the link parameters are verified by grouping the similarity matrix and checking the global relationship matrix to realize the topological relationship identification of multiple types of sensing devices.

[0012] Furthermore, the similarity calculation in step 3 is based on: time... Calculated by formula , , Vector similarity: ; in, A time-series vector consisting of a certain electrical parameter at multiple moments; This is a time-series vector formed by the same electrical parameter of another device at the same multiple times; For vectors The model; For vectors The model; For vectors and The cosine similarity, with a range of . The closer the value is to 1, the more consistent the directions of the two vectors are, indicating that the two devices have a strong correlation in electrical parameters and are likely to be located in the same power supply branch or in an electrically adjacent position. The closer the value is to 0, the weaker the correlation is; the closer the value is to -1, the opposite the trend of change is. For the first The moment, the first Voltage values ​​of each device; For the first The moment, the first Current value of each device; For the first The moment, the first The active power value of each device; The above calculations generate a similarity matrix, which is then used for group calculations and global relation matrix verification to achieve automatic identification of topological relationships.

[0013] Furthermore, in step 4, when a fault occurs, the fault information is transmitted to the dispatch master station in real time, a handling work order is automatically generated, the Hongrui terminal is remotely controlled to isolate the fault area, and the load is transferred through the end-line interconnection technology, shortening the traditional fault handling process to fault discovery, work order issuance and on-site handling.

[0014] Furthermore, in step 5, for distributed photovoltaic access scenarios, the scheduling master station generates a photovoltaic output control strategy, and the Hongrui terminal flexibly controls the inverter output through the photovoltaic coordinator, dynamically optimizing according to source-load fluctuations.

[0015] Furthermore, this low-voltage transparent distribution network construction method based on a multi-level collaborative architecture realizes the updating and maintenance of topological relationships through the similarity matrix calculation method, and achieves real-time observability, measurability, controllability, and adjustability of the low-voltage power grid through the interaction between the Hongrui terminal and the dispatcher.

[0016] Furthermore, this low-voltage transparent distribution network construction method based on a multi-level collaborative architecture is adapted to the reverse overload, overvoltage, voltage sag, and operational safety issues caused by the access of new sources and loads such as distributed photovoltaics and electric vehicles.

[0017] A low-voltage transparent distribution network system for implementing the above method includes a dispatch decision layer, a Hongrui intelligent procurement collaboration layer, and a sensing and perception layer. The three layers are connected and work together through a communication network. The scheduling decision-making layer is configured to integrate distribution network GIS topology, ledgers and multi-source load data to build a four-way control platform, realizing panoramic visualization of the operation status, generation of global optimization instructions and cross-professional collaborative scheduling. The Hongrui Intelligent Procurement Collaboration Layer includes Hongrui terminals deployed in each low-voltage distribution area. The Hongrui terminals are configured to have local computing, local autonomy, and scheduling collaboration capabilities. They collect real-time voltage, current, active power, reactive power data, and household meter ledger information from lower-level sensing devices through RS485 communication and HPLC carrier. Based on spatiotemporal vector similarity calculation, they automatically identify and maintain the station-line-transformer-household topology relationship, receive control and adjustment commands issued by the scheduling decision layer, and optimize power distribution and voltage within the region. The sensing layer includes an incoming line switch, a branch switch, a sectionalizing switch, a miniature circuit breaker, a smart distribution transformer terminal for the transformer area, and a photovoltaic protocol converter. Each device is equipped with HPLC carrier communication function and is configured to upload voltage, current, active power, reactive power, equipment status, and environmental data in real time, and receive commands from the Hongrui terminal to execute circuit breaker opening and closing or photovoltaic inverter output adjustment. The Hongrui terminal is further configured to report fault information to the scheduling decision layer in real time when a fault is detected, and to receive remote control to quickly isolate the faulty area, and to realize load transfer of the non-faulty area through the interconnection of the terminal line; The scheduling decision layer also integrates a group scheduling and control algorithm, which is configured to generate a photovoltaic output control strategy based on the real-time load and photovoltaic output data sent by the Hongrui terminal, and send it to the Hongrui terminal. The Hongrui terminal then flexibly controls the inverter output through the photovoltaic protocol converter to dynamically eliminate reverse overload, overvoltage and voltage sag.

[0018] The low-voltage transparent distribution network construction method based on a multi-level collaborative architecture provided by this invention has the following significant advantages: 1. Achieve full transparency and accurate perception of low-voltage distribution networks: Through the full deployment of equipment and data collection in the "sensing and perception layer" and the fusion of multi-source data in the "edge collaboration layer", real-time, panoramic and accurate perception of electrical parameters, equipment status and topology relationships in low-voltage distribution networks is achieved, which completely solves the problem of "unclear assets and incomplete data" in traditional management and provides a reliable data foundation for intelligent decision-making.

[0019] 2. Significantly improve the efficiency and accuracy of topology relationship maintenance: By adopting an automatic topology identification method based on spatiotemporal vector similarity, the connection relationship between "station-line-transformer-customer" can be dynamically and automatically identified and updated. This avoids the problems of large workload, low efficiency and many errors caused by relying on manual on-site surveys and data entry, significantly reduces operation and maintenance costs, and ensures the real-time and accuracy of topology data.

[0020] 3. A highly efficient and collaborative control system has been established, improving response speed and reliability: Through a three-level collaborative control architecture of "scheduling-edge-sensing," closed-loop management has been achieved, from global optimization decision-making to regional autonomous execution and precise equipment control. Especially in fault scenarios, the system can automatically locate, remotely isolate, and quickly restore power supply, compressing the traditional multi-stage and time-consuming fault handling process into a minute-level response, greatly improving power supply reliability and customer satisfaction.

[0021] 4. Effectively supports the access of new sources and loads and active optimization and control: For the large-scale access of new loads such as distributed photovoltaic and electric vehicles, the group control algorithm and edge collaborative execution can dynamically adjust the photovoltaic output and balance the regional power, effectively suppress safety issues such as reverse overload and voltage over-limit, and improve the distribution network's ability to accept high proportions of renewable energy and its operational resilience.

[0022] 5. Promoting the digital transformation and cross-professional collaboration of management models: This approach integrates multi-professional data and business through a unified technical platform, breaking down professional barriers and realizing cross-departmental collaborative scheduling and closed-loop management. It promotes the transformation of low-voltage distribution networks from a "passive response" to a "proactive early warning and intelligent control" digital operation and maintenance model.

[0023] 6. Excellent scalability and adaptability: The multi-level collaborative architecture and standardized communication interface design facilitate the flexible access and expansion of new sensing devices, edge computing units, and advanced application functions, adapting to the future evolution of power distribution network structure and technology upgrade needs, and possessing long-term applicability and investment protection value.

[0024] In summary, this invention not only significantly improves the real-time sensing capability, operation and maintenance efficiency, and power supply reliability of low-voltage distribution networks, but also provides a systematic technical path for their development towards intelligence, flexibility, and sustainability, and has significant engineering application value and promotion prospects. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0026] Figure 1 This is a flowchart illustrating the low-voltage transparent distribution network construction method based on a multi-level collaborative architecture according to the present invention. Figure 2 This is a schematic diagram of the three-level collaborative system architecture and hardware deployment of the present invention; Figure 3This is a schematic diagram illustrating the principle of automatic topology recognition based on similarity calculation in this invention; Figure 4 This is a schematic diagram of the fault rapid isolation and load transfer process of the present invention; Figure 5 This is a schematic diagram of the group control and regulation implementation process for distributed photovoltaic power generation according to the present invention; Figure 6 This is a structural block diagram of the low-voltage transparent power distribution network system of the present invention. Detailed Implementation

[0027] 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 some embodiments of the present invention, and not all embodiments. 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.

[0028] To address the technical challenges of existing systems such as opaque data, uncoordinated regulation, and fragmented business processes, this embodiment provides a method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture. This embodiment uses a typical residential community in a prefecture-level city in southern China, powered by a 400kVA box-type transformer, as an example to fully illustrate the entire process of construction, commissioning, and operation of this low-voltage transparent distribution network based on a multi-level collaborative architecture. The power supply structure of this area includes 8 branch boxes, 512 residential users, and 98kW of distributed rooftop photovoltaic power.

[0029] like Figure 1 As shown, the construction method for a low-voltage transparent distribution network based on a multi-level collaborative architecture specifically includes the following steps: Step 1 aims to construct a three-tiered collaborative technical architecture of "scheduling decision-making - Hongrui intelligent procurement collaboration - sensing perception", complete the hardware deployment and software configuration at each level, and establish an effective linkage between the technical system and the management system.

[0030] like Figure 2 As shown, the system is deployed as follows: 1. Upper layer: Deployment of the scheduling and decision-making layer; Deployment location: OCS dispatch center of municipal power supply company.

[0031] Hardware Platform: Computing Servers: Two Huawei TaiShan 2280 V2 servers, each equipped with a 128-core Kunpeng 920 processor, 512GB of memory, and an 8TB SSD (configured as a RAID 5 disk array), providing high-performance computing power. Demonstration Terminals: Two Huawei CloudLink Board 86-inch touchscreen all-in-one machines, used for panoramic visualization of the operational status and human-computer interaction.

[0032] Network Security: Vertical Isolation: Deploy NARI SAFE-3000 forward isolation devices to achieve secure isolation between the scheduling data network and external networks. Lateral Protection: Deploy Sangfor AF-7500 power-specific firewalls to strengthen internal network security protection within the platform.

[0033] Software platform: Operating system: Kylin V10 SP3.

[0034] Core Application: A self-developed "Low-Pressure Four-Dimensional (Observable, Measurable, Controllable, Adjustable) Management Platform" based on a microservice architecture. Frontend: Utilizes the Vue3 framework and ECharts charting library to build the visualization interface. Backend: Employs the Spring Boot framework to build business logic and API services. Data Storage and Analysis: Uses the StarRocks OLAP database to support rapid querying and analysis of massive amounts of multi-source data.

[0035] Data integration: The platform integrates heterogeneous data from multiple sources, including distribution network GIS topology, equipment asset ledgers, and real-time and historical load data.

[0036] Communication method: A secure and reliable remote connection is established with the Hongrui Smart Procurement Collaboration Layer equipment in each of the subordinate transformer substations through a dedicated power communication channel consisting of "power wireless private network (4G / 5G) + APN + VPN". The communication heartbeat interval between systems is kept no more than 3 seconds.

[0037] 2. Middle Layer: Deployment of the Hongrui Intelligent Procurement Collaboration Layer: The device model is "Hongrui Intelligent Procurement Low-Voltage Edge Computing Terminal HR-ECU1000", which serves as the core edge computing unit.

[0038] Installation location: Inside the low-voltage cabinet of the box-type transformer, with a physical distance of no more than 5 meters from the distribution transformer monitoring terminal.

[0039] Hardware specifications: Structural dimensions: 260mm×180mm×90mm, using standard guide rail mounting.

[0040] Core processor: Rockchip RK3568 chip, quad-core ARM Cortex-A55 architecture, 2.0GHz clock speed, integrated neural network processing unit (NPU) with 0.8 TOPS computing power, supporting lightweight AI computing on the edge.

[0041] Memory and storage: 4GB LPDDR4 memory; 64GB eMMC internal storage; also equipped with a 128GB industrial-grade TF card for cyclic storage of local historical data for no less than 90 days.

[0042] Communication interfaces: Two independent RS485 interfaces with 1200V opto-isolation. One HPLC+HRF dual-mode carrier master station module, theoretically supporting up to 2048 concurrent communication nodes. One gigabit Ethernet port, one 4G communication module supporting full network compatibility, with an optional 230MHz powerline private network module as a backup. Integrated BeiDou / GPS dual-mode timing module, with timing accuracy ≤30 nanoseconds.

[0043] Power supply and environmental adaptability: Wide voltage input AC 85-265V, directly taken from the low voltage bus of the distribution transformer, and powered to the system after industrial-grade isolated DC / DC conversion; operating temperature range -40℃ to 85℃; protection level IP51.

[0044] 3. Deployment of the lower-level sensing and perception layer: like Figure 2 As shown, the sensing layer consists of intelligent devices deployed at key nodes of the low-voltage distribution network, responsible for the full acquisition of basic data and the execution of control commands, as detailed in the table below: Equipment Name Deployment quantity Communication methods Core Functions Model Example / Remarks Distribution transformer monitoring terminal 1 unit RS485 (Master Mode) Monitor electrical quantities such as three-phase voltage, current, and power on the low-voltage side of the transformer; store and manage meter file information for the transformer substation area; generate and store 15-minute frozen data. NARI PD-681, Dongfang Electronics DTZ-6863, etc. Main incoming line intelligent switch 1 unit RS485 Equipped with a high-precision current transformer, it monitors the total active and reactive power, voltage, and current of the transformer area; it supports remote electric opening and closing operations. Schneider ComPact NSX100, etc. Branch intelligent circuit breaker 8 units RS485+HPLC Dual Mode Molded case circuit breakers with measurement functions monitor the electrical parameters of each branch circuit; they also support remote opening and closing and fault alarms. ABB Tmax XT3+Ekip module, etc. Section switch 3 units HPLC Pole-mounted load switches have measurement and remote control functions and are used for line segmentation and fault isolation. Xu Ji WSK-12, etc. Miniature intelligent circuit breaker 512 units HPLC Monitors single-phase voltage, current, leakage status, and temperature on the user side; supports remote tripping and closing. Zhejiang Deling Intelligent Breakdown DZ47X series, etc. Photovoltaic protocol converter 28 units HPLC (Modbus front-end) It enables the conversion between the Modbus communication protocol and the HPLC carrier protocol of photovoltaic inverters, and supports real-time reading and flexible adjustment of the active / reactive power output of the inverters. HR-PVI100 All deployed sensing layer devices have passed type tests and communication protocol consistency tests organized by the State Grid Corporation of China or China Southern Power Grid Company to ensure interconnectivity and reliable performance.

[0045] Step 2, Data Collection and Acquisition: This step details how the Hongrui Intelligent Procurement Collaboration Layer uses dual-channel communication technology to achieve efficient collection, fusion, and local storage of full data from different types of equipment in low-voltage distribution networks, providing a unified and high-quality data foundation for subsequent topology identification and intelligent control.

[0046] 1. Millisecond-level data reading from RS485 bus devices: The Hongrui terminal is equipped with two independent RS485 interfaces, which are used to connect different types of devices to achieve differentiated and efficient data acquisition.

[0047] 1.1. Interface 1: Data Acquisition from Distribution Transformer Monitoring Terminal: Connection equipment: 1 distribution transformer monitoring terminal.

[0048] Communication protocol: Extended version of the power industry standard "DL / T 645-2007 Communication Protocol for Multifunctional Energy Meters".

[0049] Data collection tasks and schedule: Household form information: Every 15 minutes: that is, at the top of the hour, 15 minutes, 30 minutes, and 45 minutes, a freeze read is performed to obtain the complete household form ledger information of 512 households in the entire district (including household number, form address, multiplier, etc.) to ensure the synchronous update of basic files.

[0050] Total electrical quantities of the distribution transformer: Polling is performed every 5 seconds to read the three-phase voltage on the low-voltage side of the distribution transformer. , , Three-phase current , , Total active power Total reactive power and power factor Key operating parameters are monitored to enable real-time monitoring of the transformer's operating status.

[0051] 1.2. Interface 2: Data Acquisition for Incoming and Branch Switches: Connection equipment: 1 main incoming line intelligent switch and 8 branch intelligent circuit breakers, for a total of 9 devices.

[0052] Communication protocol: Modbus-RTU protocol commonly used in the industrial field.

[0053] Data Acquisition Strategy: A master-slave half-duplex polling method is adopted. The Hongrui terminal acts as the master station, sequentially accessing slave devices with addresses 1 to 9. Address 1 is the default main incoming switch, and addresses 2-9 are the eight branch switches. Function code 03 is used to read the holding registers and access the real-time operating data register groups of each switch device.

[0054] Data collected includes: phase voltage, line current, active power, reactive power, switch open / closed position status, and internal temperature of each switchgear.

[0055] Performance metrics: The total time to complete a full data reading of all 9 devices does not exceed 650 milliseconds, meeting the real-time requirements for rapid changes in electrical quantities at key nodes.

[0056] 2. Wide-area data reading using high-speed carrier network for HPLC: The Hongrui terminal uses a built-in HPLC high-speed power line carrier master station module to build and manage a wireless communication network covering all terminal devices in the entire station area, solving the access problem of a large number of dispersed devices.

[0057] Network architecture: Automatic networking: After the system is powered on, it automatically builds a two-level optimized network topology with Hongrui terminal as the central coordinator (CCO) and some high-performance branch switches or concentrators as proxy coordinators (PCOs), thereby enhancing network coverage and reliability.

[0058] Optimize the copying strategy: Network-wide time synchronization: A precise time synchronization signal is broadcast to the entire network every 3 seconds to ensure the consistency of data timestamps across all end devices, laying the foundation for subsequent spatiotemporal correlation analysis.

[0059] Time-based and tiered data reading: Priority allocation: The reading cycle is divided into multiple time slots. The photovoltaic coordinator, involved in real-time control, is assigned to the first time slot and enjoys the highest reading priority. Load balancing: The remaining numerous devices, such as the 512 miniature circuit breakers, are evenly distributed to subsequent time slots. For example, divided into 8 time slots, with approximately 64 nodes read per slot, network load balancing is achieved, avoiding channel congestion.

[0060] Hybrid communication mode: This mode combines broadcast commands with unicast follow-up. First, a read command is broadcast, and data is received in batches from devices that respond successfully. For devices that do not respond, unicast follow-up is then performed to improve overall read efficiency.

[0061] Data Acquisition Equipment and Content: This network collects real-time voltage, current, power, and status data such as leakage and tripping from devices such as miniature circuit breakers, photovoltaic coordinators, and sectionalizing switches.

[0062] Communication performance: Based on actual deployment tests, under normal conditions, the average success rate of data reading across the entire network is as high as 99.996%; under extreme weather conditions such as thunderstorms, high temperature and high humidity, the success rate of reading can still be maintained at a level of no less than 99.97%, which proves the robustness of the network.

[0063] 3. Local multi-mode hybrid data storage and management: To achieve efficient management and rapid application of massive amounts of diverse data, Hongrui Terminal adopts a hybrid database architecture for local storage.

[0064] SQLite, a relational database, stores static device attribute information such as device ID, model, and installation location; network topology relationships automatically identified by the system or maintained manually; and user-transformer correspondences—all structured, non-time-varying data. SQLite provides fast relational data access support for data querying, topology visualization, and device management.

[0065] TDengine Time Series Database: TDengine stores timestamped time series monitoring data uploaded by all devices, including voltage, current, power, and temperature. It employs a tiered storage strategy: for the high-precision data layer, it saves raw sampling data at the second or millisecond level for 7 days. This is primarily used for real-time monitoring, fault recording, and refined analysis. For the aggregated data layer, it saves aggregated data such as mean, maximum, and minimum values ​​at the minute level for 90 days. This is primarily used for medium- to long-term trend analysis, statistical report generation, and big data mining. TDengine leverages the high compression rate, high-speed write, and query performance of time series databases to achieve efficient storage and retrieval of massive amounts of monitoring data.

[0066] Through the above design, Hongrui Terminal has achieved "full-volume, all-time, and all-element" data perception and localized preprocessing of low-voltage distribution networks, from transformers to user meters, and from primary equipment to distributed power sources, thus building a solid data foundation.

[0067] Step 3, Topology Relationship Identification and Maintenance: This step is the core innovation of this invention, such as... Figure 3 As shown, its core lies in the adoption of an automatic identification method for low-voltage distribution network topology based on multi-parameter spatiotemporal vector cosine similarity clustering. This method utilizes the spatiotemporal correlation of electrical parameters between devices to automatically and accurately identify and dynamically update the network topology without manual intervention.

[0068] 1. Data preprocessing and feature extraction: Analysis time window selection: The past 24 hours is used as a complete analysis period, and 96 equally spaced hourly times (00:00, 00:15, …, 23:45) are selected as analysis sample points.

[0069] Electrical feature vector construction: for each time step and each device Extract the electrical feature triples at that moment to form a feature vector. : ; In the formula, For equipment At any moment The voltage value; For equipment At any moment The current value; For equipment At any moment The active power value.

[0070] 2. Construction of time series vectors: To analyze the behavioral patterns of each device over time, for each device... Arrange the values ​​of the same type of electrical parameters at 96 time points in chronological order to construct three time-series vectors of length 96, represented as column vectors: ; ; . 3. Cosine similarity calculation: For any two devices and Calculate their voltage values ​​respectively. Current Active power Cosine similarity of time-series vectors across three dimensions. This value is used to quantify the consistency of the trends in variation of the same electrical parameters between two devices.

[0071] The calculation formula, taking voltage similarity as an example: ; In the formula, dot product of vectors ; The Euclidean norm (modulus) of a vector: Similarly, it can be calculated and .

[0072] Significance of similarity: The range of similarity values ​​is... A value close to 1 indicates that the two devices exhibit highly consistent trends in this electrical quantity, indicating a strong electrical correlation. A value close to 0 indicates that the trends are unrelated, indicating a weak electrical correlation. A value close to -1 indicates that the trends are opposite.

[0073] 4. Weighted Comprehensive Similarity Calculation Since different electrical parameters have varying abilities to characterize topological relationships, a weighted approach is used to calculate the overall similarity. This allows for a more comprehensive assessment of the correlation strength between devices.

[0074] ; In this embodiment, after training and verification with data from hundreds of actual transformer substations, the optimal weights were determined as follows: ; Weighting criteria: Voltage weight 0.45: In low-voltage distribution networks, the voltage amplitude and fluctuation of equipment on the same power supply branch are highly similar, making it the most stable indicator for judging electrical proximity. Current weight 0.40: Current paths are clear and can effectively reflect upstream and downstream power supply relationships, but it is slightly affected by individual differences in user loads. Active power weight 0.15: Active power is significantly affected by the randomness of user electricity consumption behavior, and its indicative role in topological relationships is relatively weak, but it still contains some correlation information.

[0075] 5. Association Determination and Hierarchical Clustering: Threshold setting: ≥0.88: Determined to be a strong correlation, with a very high probability that they belong to the same power supply branch or are directly connected. 0.75≤ <0.88: The correlation is considered moderate, but further confirmation with more information or subsequent data is needed. <0.75: The correlation is weak or there is no direct association.

[0076] Bottom-up hierarchical clustering algorithm flow: Initialization: Treat each user table as an independent cluster.

[0077] The first round of clustering, micro-circuit breaker (MCB) assignment: All micro-circuit breakers (MCBs) are traversed, and their average comprehensive similarity to existing customer meter clusters is calculated. If the average similarity between a micro-circuit breaker and a certain cluster is ≥0.88, then the micro-circuit breaker is assigned to that cluster. This step determines the specific set of customers served by each micro-circuit breaker.

[0078] The second round of clustering, branch clustering: It iterates through all branch switches and calculates their average comprehensive similarity to the clusters formed in the first round that already include customer meters and micro-terminals. Again, based on a threshold, the branch switch is assigned to the cluster with the highest similarity. This step establishes the affiliation relationship between the branch switches and the downstream load groups.

[0079] The third round of clustering, main incoming line aggregation: calculates the average similarity between the main incoming line switch and all branch clusters, completes the final aggregation, and forms a complete tree-like topology structure of "main incoming line switch → branch switch → miniature circuit breaker → household meter".

[0080] 6. Verification, maintenance, and visualization: Automatic verification: The topology results automatically identified by the algorithm are compared with the "household-transformer relationships" already registered in marketing management systems such as electricity information collection systems. If the proportion of inconsistent nodes exceeds 5%, the system automatically generates a "topology relationship anomaly verification work order".

[0081] Manual intervention and closed-loop correction: Maintenance personnel carry mobile work terminals to the site to conduct on-site verification of anomalies indicated by work orders, such as scanning codes to confirm devices. The correct topology relationship after verification will be written back to the marketing management system and the local database of the Hongrui terminal, realizing data consistency and self-correction between systems.

[0082] Dynamic update mechanism: The above-mentioned automatic topology identification and verification process is set to be executed automatically once a day during the low-load period in the early morning, such as 02:00-03:00, with the time taken within 6 minutes. This mechanism ensures that the topology relationship can be updated in near real-time with changes in network structure such as power grid upgrades and user installation applications.

[0083] Multi-dimensional visualization: The Hongrui terminal uploads the final topology data and device coordinates to the dispatch master station in standard GeoJSON format. The master station platform uses a Geographic Information System (GIS) engine, such as calling the ArcGIS API, to automatically generate and provide the following views: Geographic wiring diagram: Accurately draw the location of equipment and the route of wiring on the satellite image base map.

[0084] Electrical Single-Line Diagram: A clear single-line diagram of the system, supporting interactive operations: zoom, drag, and click to view real-time data.

[0085] Hierarchical topology tree diagram: intuitively displays the hierarchical affiliation and connection relationship of "transformer - branch - user".

[0086] Through this step, the present invention completely changes the traditional low-voltage topology management mode that relies on manual input and maintenance, and realizes intelligent and automated identification and high-precision dynamic management of topology relationships, laying a core foundation for the transparent operation of the distribution network.

[0087] Step 4: Construction of a three-level collaborative intelligent control system and rapid fault handling method This step aims to establish and implement a three-tiered collaborative intelligent control system: scheduling layer, edge layer, and device layer. For example... Figure 4 As shown, taking a real residential electrical leakage fault as an example, the complete minute-level closed-loop handling process from fault occurrence, location, isolation, power transfer to restoration is demonstrated.

[0088] 1. Three-tier collaborative control architecture: Scheduling layer: Policy generation and global coordination. Functional Positioning: Serving as the brain of control decisions, it performs global optimization analysis based on panoramic data (load, photovoltaic output, voltage quality, etc.). Control Output: Generates network-wide regulation strategies, such as preventative load transfer, voltage and reactive power optimization commands, and optimal isolation and power transfer schemes under fault conditions.

[0089] Edge layer, instruction decomposition and region autonomy: Functional Positioning: Centered on the Hongrui terminal, it acts as the receiver of dispatch instructions and the coordinator of regional execution. Control Role: It receives macro-level strategies issued by the dispatch layer, combines them with the real-time operating status of the local area, decomposes, optimizes, prioritizes, and performs security checks on instructions, and is responsible for coordinating the collaborative actions of various types of equipment within the area.

[0090] At the device level, precise execution and status feedback: Functional Positioning: Composed of various intelligent circuit breakers and photovoltaic coordinators, it is the final execution unit of control commands. Control Execution: It accurately executes specific operations such as opening and closing circuit breakers and power regulation from the edge layer, and provides real-time feedback on the execution results and its own status.

[0091] 2. Rapid Fault Isolation and Recovery Process: The following example, using a real-life residential electrical leakage fault, details the closed-loop handling process of the three-level control system. Case background: In a residential community, a user connected to a miniature circuit breaker (address 0xA7F3) experienced an electrical leakage.

[0092] Phase 1: Local fault detection and instantaneous reporting; Time: 14:26:33.000; Event: User's air conditioner is leaking current, with a leakage current of 52mA. Action: The miniature intelligent circuit breaker detected the anomaly, immediately activated its built-in protection logic, automatically tripped locally, and cut off the power supply to the fault point to prevent the accident from escalating.

[0093] Time: 14:26:33.412; Event: The micro-circuit transmitted information packets containing the "leakage protection action" event, its own address, and fault parameters to the Hongrui terminal in this area via the HPLC carrier network, with a reporting delay of only 412 milliseconds.

[0094] Phase Two: Edge Intelligent Positioning and Information Fusion; Time: 14:26:33.789; Action: After receiving the fault event, the Hongrui terminal immediately started the analysis program. It queried the latest topology database stored locally and quickly located the device ID (0xA7F3) of the faulty micro-circuit. Based on the topology, it automatically determined that the micro-circuit directly affected 3 households downstream.

[0095] Time: 14:26:34.105 Action: Hongrui Terminal sends a structured fault report, including the faulty device ID, geographical / electrical location, number of affected households, and associated branch switch information, to the dispatch master station.

[0096] Phase Three: Intelligent Decision-Making and Remote Isolation at the Main Station; Time: 14:26:34.623 Action: After receiving the fault report, the dispatch master station "Low-voltage Four-way Control Platform" automatically generates a standard "Low-voltage Fault Repair Work Order" based on the fault location and topology. Through Geographic Information System (GIS) matching, the work order is automatically dispatched to the mobile terminal of the repair personnel closest to the fault point, a distance of approximately 1.2 kilometers.

[0097] Time: 14:26:35.217; Action: The dispatch master station synchronously issued a remote control command to the Hongrui terminal, ordering it to remotely isolate the nearest upstream device to the fault point—the No. 4 branch intelligent circuit breaker.

[0098] Time: After 14:26:35.217; Action: Hongrui terminal executes command to control the No. 4 branch switch to trip. At this point, the faulty section is completely isolated, and the affected area is precisely reduced from 120 households under this branch to 3 households under the faulty micro-circuit.

[0099] Phase 4: Rapid load transfer to non-faulty areas Time: 14:26:37.890; Action: After confirming that the faulty section has been isolated, the dispatch master station automatically activated the preset load transfer scheme and remotely closed the end interconnection switch connecting this transformer area with the adjacent healthy transformer area. Result: The 117 non-faulty users connected to branch switch No. 4, excluding the faulty branch, completed the load transfer within approximately 3 seconds, and power was restored from the adjacent transformer area. The users' perception of power outage was close to zero.

[0100] Phase 5: On-site inspection and power restoration Time: 14:29:10; Action: Repair personnel arrived at the scene and quickly located the fault point based on the precise positioning information on the mobile terminal, and handled the leakage fault. Restoration: After the fault was resolved, the repair personnel remotely issued a "allow closing" command through the mobile terminal, and the Hongrui terminal remotely closed the No. 4 branch switch and the user micro-circuit in sequence, restoring power to the faulty section.

[0101] 3. Comparison of implementation results: The entire process of this method took 4 minutes and 37 seconds from the occurrence of the fault (14:26:33) to the restoration of power to the non-faulty area (14:26:37.890). The traditional handling process took approximately 1 hour and 12 minutes, including approximately 15 minutes for user reports, approximately 30 minutes for manual on-site inspection and location, approximately 15 minutes for reporting and requesting a power outage through hierarchical channels, and approximately 12 minutes for fault handling and restoration. This represents a significant improvement in efficiency: handling time was reduced by over 93%, and the number of affected users decreased by 97.5%. This method achieves a fundamental shift from "passive response and manual handling" to "proactive perception, automatic isolation, rapid power transfer, and precise emergency repair," greatly improving power supply reliability and customer service levels.

[0102] Step 5: Implement flexible control and operation optimization functions to adapt to new source-load access: This step focuses on addressing the technical challenges brought about by new source-load integration technologies such as distributed photovoltaics, etc. Figure 5 As shown, the "schedule-edge" collaborative group control mechanism enables minute-level flexible regulation of photovoltaic power output, effectively solving the voltage over-limit problem.

[0103] 1. Group dispatch and control process for high-penetration photovoltaic systems: 1.1. Real-time accurate data collection and aggregation: The data acquisition entity is the Hongrui terminal. The acquisition cycle is once every 5 seconds. The core data collected includes: source-side data, obtained through each photovoltaic coordinator, capturing the real-time active power output of each inverter. and doing nothing ; Grid-side data, collecting the total active power of the low-voltage busbars of distribution transformers. Total reactive power Voltage quality data, collecting three-phase line voltages of the bus. , , .

[0104] 1.2. Data upload and overall optimization of the main site: Upload cycle and content: Hongrui terminal packages and uploads the aggregated data of the distribution area to the dispatch master station every minute.

[0105] Main site optimization decisions: Algorithm core: Multi-objective rolling optimization is performed using a model predictive control (MPC) framework.

[0106] Optimization goal: Minimize To achieve the best balance between voltage quality and economic benefits.

[0107] Operating constraints: Voltage safety constraint: 380V≤U≤418V. Equipment regulation capability constraint: The active power output regulation range of each inverter is 0% to 100%. Regulation fairness constraint: The difference in power reduction ratio between any two inverters shall not exceed 20%.

[0108] Solution and cycle: High-performance optimization solvers such as Gurobi are used to solve the optimal control sequence for the future time period every 15 minutes, with a single calculation time of ≤8 seconds.

[0109] 1.3. Issuance and rapid execution of control strategies at the edge: Command issuance: The dispatch master station will issue the optimized and calculated specific control commands to the Hongrui terminals of the corresponding transformer areas.

[0110] Local execution: After Hongrui terminal receives the instruction: The corresponding photovoltaic coordinator is addressed using an HPLC carrier network.

[0111] The target active and reactive power setpoints are written into the control register of the photovoltaic inverter via the Modbus-RTU protocol. For example, register 40010 corresponds to the active power setting.

[0112] The photovoltaic inverter responds and adjusts its output within 5 seconds.

[0113] 1.4. Verification of the regulatory effect: Scenario: Midday peak solar power generation period. Before regulation: The highest voltage on the transformer substation bus reached 412V, exceeding the limit, with a voltage compliance rate of only 78%. After regulation: Through flexible regulation, the bus voltage stabilized at 398V, within the acceptable range, and the voltage compliance rate improved to 99.8%. Core value: Under the premise of minimal curtailment, the problem of voltage exceeding the limit caused by solar backfeeding was solved, significantly improving the distribution network's carrying capacity and absorption rate for distributed solar power.

[0114] 2. Overall beneficial effects and implementation effectiveness of the method: 2.1. Comprehensive benefits brought about by core technology innovation: The low-voltage transparent distribution network construction method based on a multi-level collaborative architecture proposed in this embodiment achieves the following core values ​​through two core technologies: "automatic topology identification" and "source-load coordinated regulation": Achieving holographic perception and transparent management: By calculating the similarity matrix, the topology relationship is automatically updated and maintained. Combined with the real-time interaction between the Hongrui terminal and the dispatcher, the low-voltage power grid achieves a level of transparency that is "real-time observable, measurable, controllable, and adjustable".

[0115] Enhancing the adaptability and resilience of the power grid: This method is particularly suitable for the complex operating conditions caused by the large-scale access of new sources and loads such as distributed photovoltaics and electric vehicle charging piles. It can effectively deal with safety issues such as "transformer reverse overload, local overvoltage, and voltage sag" and improve the safety and resilience of power grid operations.

[0116] 2.2. Empirical data for large-scale application: From January 2024 to June 2025, this method was implemented on a large scale in 1268 low-voltage distribution areas in a prefecture-level city. The key performance indicators (KPIs) are compared below based on an independent evaluation by the China Electric Power Research Institute: In summary, this method not only provides a complete technical solution but has also been tested in large-scale engineering practice. Its implementation can systematically solve the long-standing problems of "blindness, muteness, disorder, and poor performance" in low-voltage distribution networks, significantly improving operation and maintenance efficiency, power supply reliability, and renewable energy absorption capacity. It provides an effective path for building intelligent, flexible, and reliable distribution networks under new power systems. Those skilled in the art can proportionally increase or decrease equipment or fine-tune algorithm parameters according to different application scenarios; such equivalent substitutions and adaptive modifications fall within the protection scope of this invention.

[0117] In addition, such as Figure 6 As shown, this embodiment also provides a physical system corresponding to the aforementioned method, namely a low-voltage transparent distribution network system based on a multi-level collaborative architecture. This system consists of a dispatch decision-making layer, a Hongrui intelligent procurement collaboration layer, and a sensing and perception layer. These three layers are connected through a secure dedicated power communication network, forming a vertically integrated and horizontally collaborative organic whole.

[0118] System overall structure: The system consists of a scheduling decision layer, a Hongrui intelligent procurement collaboration layer, and a sensing and perception layer. The three layers are connected through a secure power-specific communication network, forming an organic whole that is vertically integrated and horizontally coordinated.

[0119] 1. Scheduling decision-making level: Deployment location: Dispatch center of municipal power supply company.

[0120] Core component: Low-voltage "observable, measurable, controllable, and adjustable" intelligent management and control platform.

[0121] Core functional configuration: Data fusion and panoramic visualization: The configuration integrates real-time data from various distribution areas, distribution network geographic information system (GIS) topology, equipment asset ledgers, and historical load data to build a unified digital twin model and realize a panoramic visualization display of the power grid's operating status.

[0122] Intelligent analysis and decision-making: Configured to run advanced analysis algorithms such as power flow calculation and situational awareness based on fused data, generating network-wide optimized operation strategies, fault handling plans and scheduling instructions.

[0123] Cross-disciplinary collaborative scheduling: Configured to provide a unified command interface, supporting collaborative operations and closed-loop management of different disciplines such as scheduling, operation and maintenance, and marketing based on the same data foundation and visualization scenario.

[0124] 2. Hongrui Intelligent Procurement Collaboration Layer: Deployment location: Each low-voltage distribution transformer area.

[0125] Core component: Hongrui terminals deployed in each distribution area serve as the core edge computing and collaborative control unit.

[0126] Core functional configuration: Multi-source data acquisition and aggregation: Configured to acquire real-time electrical quantities of all lower-level sensing devices through its RS485 interface and HPLC carrier master module: voltage, current, active power, reactive power, equipment status, environmental data, and household meter ledger information.

[0127] Automatic topology identification and maintenance: Configured with a built-in "spatiotemporal vector similarity calculation" algorithm, it automatically calculates, identifies and dynamically maintains the topological connection relationship of "station-line-transformer-household" based on the collected time series data.

[0128] Regional autonomy and collaborative control: Command execution: Configured to receive and parse control and regulation commands issued by the scheduling decision layer.

[0129] Local optimization: Configured to combine the real-time operating status of the local area, perform safety verification and optimization decomposition of instructions, and coordinate various switching equipment such as branch circuit breakers and regulating equipment such as photovoltaic coordinators within the control area.

[0130] Rapid Fault Handling: When a fault is detected or received, the fault location logic is immediately activated, the fault details are reported to the scheduling layer in real time, and remote control commands are accepted to quickly isolate the faulty section and assist or automatically perform load transfer in the non-faulty area.

[0131] Data caching and protocol conversion: Configured to have local data storage capabilities and complete the conversion between different device communication protocols such as DL / T645, Modbus, HPLC and scheduling layer platform protocols.

[0132] 3. Sensing Layer Deployment locations: key nodes in the low-voltage distribution network, including distribution transformer side, line branch points, user access points, and distributed power source access points.

[0133] The core equipment components and functions are as follows: Equipment Name Main Function Configuration Smart distribution transformer terminal in the distribution area Monitor the three-phase electrical parameters on the low-voltage side of the transformer, and store and upload the meter information. Main incoming line intelligent switch Monitors the total active and reactive power, voltage, and current of the transformer area, and supports remote switching on and off. Branch intelligent circuit breaker Monitor the electrical parameters of each branch circuit, and support remote opening and closing of circuit breakers and fault alarms. Section switch Used for line segmentation, it has measurement and remote control functions and supports fault isolation. Miniature intelligent circuit breaker Monitors single-phase electrical quantities, leakage current, and temperature status on the user side, and supports remote tripping and closing. Photovoltaic protocol converter Connect to a photovoltaic inverter to achieve the conversion between Modbus protocol and HPLC carrier protocol, and support the reading and setting of the inverter's active / reactive power output. Unified communication features: All of the above devices are equipped with HPLC carrier communication modules, and some devices also have RS485 interfaces, forming a broadband communication network based on power line media, enabling real-time data uploading and reliable reception and execution of control commands.

[0134] 4. Implementation of system-level collaborative functions: Coordinated regulation of new energy sources and loads: The group-based scheduling and control algorithm integrated in the scheduling decision layer is configured to calculate and generate photovoltaic output regulation strategies on a rolling basis based on the real-time load and photovoltaic output data periodically transmitted from the Hongrui terminal. After the strategy is sent to the Hongrui terminal, the terminal flexibly controls the output of the photovoltaic inverter through the photovoltaic protocol converter, dynamically solving problems such as reverse overload, overvoltage, and voltage sag caused by large-scale photovoltaic access.

[0135] The "monitoring, control, adjustment, and maintenance" closed loop: Through real-time data flow and control flow between the three layers, the system forms a complete management closed loop of "comprehensive status perception → intelligent analysis and decision-making → precise collaborative control → effect evaluation and feedback", realizing the fundamental transformation of low-voltage distribution networks from traditional passive operation and maintenance to modern proactive intelligent operation.

[0136] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture, characterized in that: Includes the following steps: Step 1, System Architecture Design and Deployment: Adopt a multi-level collaborative architecture of scheduling decision-making, Hongrui intelligent procurement collaboration and sensor perception, integrate the low-voltage dispatch and command center organizational model, establish a linkage between technology and management systems, and realize the overall framework of low-voltage transparent distribution network; Step 2, Data Acquisition and Acquisition: Using the Hongrui terminal, RS485 communication and HPLC carrier are used to collect household meter ledger information and real-time voltage, current, active power and reactive power data from devices such as distribution transformer monitoring terminals, incoming line switches, branch switches, miniature circuit breakers, and photovoltaic coordinators, and store them in the terminal. Step 3, Topology Relationship Identification and Maintenance: Based on the spatiotemporal vector judgment method, calculate the similarity matrix of voltage, current and power between devices, automatically identify and verify the station-line-transformer-customer link parameters, and draw a global relationship diagram by combining customer meter ledgers and GIS information to realize the updating and maintenance of topology relationships; Step 4, Control System Construction: Establish a three-level control system of global-regional-device. The scheduling layer generates control strategies, the edge layer Hongrui terminal coordinates regional devices to perform power allocation, and the sensing layer achieves precise control through photovoltaic coordinators and smart circuit breakers. In case of a fault, the region is isolated in real time and the load is transferred to shorten the handling time. Step 5, adjustment function implementation: Hongrui terminal collects real-time load and photovoltaic output data and sends it to the dispatch master station. The latter integrates group dispatch and control algorithm to generate control strategy and sends it to the terminal. The photovoltaic coordinator flexibly controls the inverter output and dynamically optimizes and solves the problems of reverse overload and voltage over-limit.

2. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture as described in claim 1, characterized in that: In step 1, the dispatch decision-making layer relies on the dispatch center to integrate multi-source data such as distribution network GIS topology, ledgers, and loads to build a four-capable management platform, which generates global optimization instructions, visualizes the overall operation status, and performs cross-professional collaborative dispatch. The Hongrui intelligent procurement and collaboration layer deploys Hongrui terminals with local computing, local autonomy, and dispatch collaboration capabilities, receives dispatch instructions, and optimizes power distribution and voltage regulation within the region. The sensing and perception layer installs miniature circuit breakers, distribution area branch circuit breakers, distribution area intelligent distribution transformer terminals, and photovoltaic protocol converters with carrier communication to collect all data on equipment status, electrical parameters, and environmental data and execute instructions.

3. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture as described in claim 1, characterized in that: In step 2, the Hongrui terminal obtains the household meter ledger information and 15-minute voltage, current and active power data from the distribution transformer monitoring terminal through RS485 communication, obtains real-time electrical data from the incoming line switch and branch switch, and obtains real-time data from the miniature circuit breaker, photovoltaic coordinator and sectionalizing switch through HPLC carrier wave, so as to realize full data perception.

4. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture according to claim 1, characterized in that: In step 3, a matrix based on time and device is established, the similarity of voltage, current and power is calculated, and the link parameters are verified by grouping the similarity matrix and checking the global relationship matrix to realize the topological relationship identification of multiple types of sensing devices.

5. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture according to claim 4, characterized in that: The similarity calculation in step 3 is based on the time interval. Calculated by formula , , Vector similarity: ； in, A time-series vector consisting of a certain electrical parameter at multiple moments; This is a time-series vector formed by the same electrical parameter of another device at the same multiple times; For vectors The model; For vectors The model; For vectors and The cosine similarity, with a range of . The closer the value is to 1, the more consistent the directions of the two vectors are, indicating that the two devices have a strong correlation in electrical parameters and are likely to be located in the same power supply branch or in an electrically adjacent position. The closer the value is to 0, the weaker the correlation is; the closer the value is to -1, the opposite the trend of change is. For the first The moment, the first Voltage values ​​of each device; For the first The moment, the first Current value of each device; For the first The moment, the first The active power value of each device; The above calculations generate a similarity matrix, which is then used for group calculations and global relation matrix verification to achieve automatic identification of topological relationships.

6. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture according to claim 1, characterized in that: In step 4, when a fault occurs, the fault information is transmitted to the dispatch master station in real time, a handling work order is automatically generated, the Hongrui terminal is remotely controlled to isolate the fault area, and the load is transferred through the end-line interconnection technology, shortening the traditional fault handling process to fault discovery, work order issuance and on-site handling.

7. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture according to claim 1, characterized in that: In step 5, for distributed photovoltaic access scenarios, the scheduling master station generates a photovoltaic output control strategy, and the Hongrui terminal flexibly controls the inverter output through the photovoltaic coordinator, dynamically optimizing according to source-load fluctuations.

8. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture according to claim 1, characterized in that: This low-voltage transparent distribution network construction method based on a multi-level collaborative architecture realizes the updating and maintenance of topology relationships through similarity matrix calculation, and achieves real-time observability, measurability, controllability, and adjustability of the low-voltage power grid through the interaction between Hongrui terminal and dispatch.

9. The method for constructing a low-voltage transparent distribution network based on a multi-level collaborative architecture according to claim 1, characterized in that: This low-voltage transparent distribution network construction method based on a multi-level collaborative architecture is adapted to the reverse overload, overvoltage, voltage sag, and operational safety issues caused by the access of new sources and loads such as distributed photovoltaics and electric vehicles.

10. A low-voltage transparent distribution network system for implementing the method as described in any one of claims 1-9, characterized in that: It includes a scheduling and decision-making layer, a Hongrui intelligent procurement collaboration layer, and a sensing and perception layer. The three layers are connected and work together through a communication network. The scheduling decision-making layer is configured to integrate distribution network GIS topology, ledgers and multi-source load data to build a four-way control platform, realizing panoramic visualization of the operation status, generation of global optimization instructions and cross-professional collaborative scheduling. The Hongrui Intelligent Procurement Collaboration Layer includes Hongrui terminals deployed in each low-voltage distribution area. The Hongrui terminals are configured to have local computing, local autonomy, and scheduling collaboration capabilities. They collect real-time voltage, current, active power, reactive power data, and household meter ledger information from lower-level sensing devices through RS485 communication and HPLC carrier. Based on spatiotemporal vector similarity calculation, they automatically identify and maintain the station-line-transformer-household topology relationship, receive control and adjustment commands issued by the scheduling decision layer, and optimize power distribution and voltage within the region. The sensing layer includes an incoming line switch, a branch switch, a sectionalizing switch, a miniature circuit breaker, a smart distribution transformer terminal for the transformer area, and a photovoltaic protocol converter. Each device is equipped with HPLC carrier communication function and is configured to upload voltage, current, active power, reactive power, equipment status, and environmental data in real time, and receive commands from the Hongrui terminal to execute circuit breaker opening and closing or photovoltaic inverter output adjustment. The Hongrui terminal is further configured to report fault information to the scheduling decision layer in real time when a fault is detected, and to receive remote control to quickly isolate the faulty area, and to realize load transfer of the non-faulty area through the interconnection of the terminal line; The scheduling decision layer also integrates a group scheduling and control algorithm, which is configured to generate a photovoltaic output control strategy based on the real-time load and photovoltaic output data sent by the Hongrui terminal, and send it to the Hongrui terminal. The Hongrui terminal then flexibly controls the inverter output through the photovoltaic protocol converter to dynamically eliminate reverse overload, overvoltage and voltage sag.

Images