Method and device for optimizing configuration of a perception unit for power distribution grid edge device identification

Abstract

This application provides a method and apparatus for optimizing the configuration of sensing units for identifying edge devices in a distribution network, belonging to the field of smart grid technology. The method includes: collecting raw measurement data of the distribution network and preprocessing the data; establishing a multi-objective optimization model based on the collected data; solving the multi-objective optimization model using an improved NSGA-II algorithm to obtain an optimized solution; selecting the Pareto optimal solution set based on the optimized solution through fast non-dominated sorting and congestion calculation; constructing a state estimation model of the distribution network based on the Pareto optimal solution set using the weighted least squares method; and identifying edge device nodes by analyzing the power direction characteristics and power fluctuation characteristics of the nodes. This application enables the obtained solution set to more closely approximate the true frontier, has a wider distribution, and better convergence efficiency, reducing the probability of the configuration scheme of sensing units and edge units getting trapped in local optima, and achieving accurate sensing of user-side edge devices.

Description

Technical Field

[0001] This application relates to the field of smart grid technology, and in particular to a method and apparatus for optimizing the configuration of sensing units for identifying edge devices in a distribution network. Background Technology

[0002] With the continuous integration of distributed power sources and diverse loads, the operation mode of distribution networks has undergone tremendous changes. It has shifted from the traditional "passive" distribution network, where power flows directly from the power generation source to end users, to the current "active" distribution network, where power is supplied from various distributed power sources to end users and the main grid. Furthermore, current no longer flows in a simple unidirectional direction but in multiple directions, resulting in a "multi-source" distribution network. Simultaneously, with the trend of multi-source development in distribution networks, diverse edge devices such as low-voltage smart circuit breakers, distributed photovoltaics, energy storage systems, and charging piles will be widely deployed. However, the large-scale integration of edge devices into the distribution network has a negative impact: the inconsistent protocols and data formats used by equipment manufacturers for edge devices lead to poor measurement accuracy and reliability for different types of edge devices, and make identification difficult. In addition, the economic efficiency of sensing unit configuration must be considered. With the increasing complexity of distribution network structures, it is currently difficult to configure smart meters (SMs) and feeder terminal units (FTUs) at every power injection node or feeder in the distribution network. This makes it difficult for the real-time status data acquired by the Supervisory Control and Data Acquisition (SCADA) system to meet the requirements of dispatchers. In summary, the increasingly complex distribution network topology, the bidirectional power flow caused by diverse loads, and the access of various edge devices bring strong time-related uncertainties to the system operation, posing significant challenges to distribution network status detection.

[0003] In the actual operation of distribution networks, real-time monitoring of the distribution network system plays a crucial role in system scheduling and control, risk monitoring, and fault location. Therefore, it is necessary to install sensing units in the distribution network system to achieve the monitoring purpose. However, with the large-scale integration of high-proportion distributed renewable energy and power electronic equipment, the distribution network system is changing towards an active distribution network environment with AC / DC interconnection, which further deepens the complexity of the system environment, making the status monitoring of nodes and the identification of user-side terminal equipment increasingly challenging.

[0004] To address the aforementioned issues, improve the overall system performance, enhance resilience, and reduce operation and maintenance costs, the State Grid Corporation of China proposed the concept of "minimized acquisition." This involves optimizing acquisition strategies, equipment configurations, and technical methods during power system data acquisition and monitoring to collect necessary information with minimal cost (e.g., the fewest acquisition devices, the narrowest acquisition channels, and the lowest energy consumption), while ensuring the safe, stable, reliable, and economical operation of the power system. Related technologies utilize lightweight edge units with inference capabilities at the edge computing layer. This allows for feature extraction and preliminary analysis of raw data from the data source via acquisition terminals. Furthermore, leveraging the characteristics of edge units, data processing, analysis, and decision-making functions are decentralized to the network edge. This on-site data processing reduces latency, improves system real-time performance, and avoids reliance on centralized cloud services.

[0005] However, due to the large number of nodes and complex branches in the distribution network system, the optimal configuration of sensing units and edge units in the distribution network is a multi-objective optimization problem with a high degree of difficulty. Therefore, there is an urgent need for a sensing unit optimization configuration method that can achieve accurate sensing of user-side end devices. Summary of the Invention

[0006] This application provides a method and apparatus for optimizing the configuration of sensing units for identifying edge devices in a distribution network. This method enables the obtained solution set to more closely approximate the true frontier, has a wider distribution, and better convergence efficiency. It reduces the probability of the configuration schemes of sensing units and edge units getting trapped in local optima and achieves accurate sensing of user-side edge devices. The technical solution includes at least the following: Firstly, a method for optimizing the configuration of sensing units for identifying edge devices in a distribution network is provided, comprising: collecting raw measurement data of the distribution network and preprocessing the data; establishing a multi-objective optimization model based on the collected data; solving the multi-objective optimization model using an improved NSGA-II algorithm (Non-dominated sorting genetic algorithm II) to obtain an optimized solution, wherein the improved NSGA-II algorithm introduces a directional mutation strategy and a historical archiving mechanism, wherein the directional mutation strategy guides the solution to move towards the normal direction of the Pareto optimal frontier, and the historical archiving mechanism saves high-quality solutions from previous iterations to perform directional hybridization with new solutions in subsequent iterations; selecting the Pareto optimal solution set based on the optimized solution through fast non-dominated sorting and congestion calculation to obtain the optimal configuration scheme of sensing units and edge units; constructing a state estimation model of the distribution network based on the Pareto optimal solution set using the weighted least squares (WLS) method; and identifying edge device nodes by analyzing the power direction characteristics and power fluctuation characteristics of the nodes.

[0007] Optionally, the multi-objective optimization model considers minimizing configuration costs and improving reliability. The objective function of the multi-objective optimization model includes a cost function and an average coverage function. The formula for calculating the cost function is as follows:

[0008] in, For configuration cost, The unit price for the sensing unit. The number of sensing units. The purchase price of the edge unit. The number of edge units, The cost per unit length of communication channel, The total length of the communication channel to be laid; The formula for calculating the average coverage function is as follows:

[0009] in, This represents the average perceived coverage rate. The number of system nodes. For the number The number of nodes covered by the sensing unit when a node fails.

[0010] Optionally, the directional mutation strategy uses the covariance matrix eigenvector estimation method to guide the solution to move towards the normal direction of the Pareto optimal frontier. The formula for calculating the eigenvector is:

[0011] in, Let be the eigenvector corresponding to the smallest eigenvalue of the covariance matrix. For the neighborhood solutions in the solution set, This represents eigenvalue decomposition.

[0012] Optionally, the congestion calculation is performed according to the following formula:

[0013] in, For the first The crowding of each solution The number of objective functions, To find the value in the target space, This represents the distance between adjacent solutions in the target dimension.

[0014] Optionally, the state estimation model of the distribution network based on the weighted least squares method is calculated according to the following formula:

[0015] in, For state estimation, For Jacobian matrices, For weighted matrices, This is the measurement vector.

[0016] Optionally, the step of identifying edge device nodes by analyzing the power direction characteristics and power fluctuation characteristics of the nodes includes: if the node power is negative for a certain period of time and the ratio of the duration of negative power to the total monitoring time is greater than 40%, it is determined to be an edge device node; if the standard deviation of the node power is greater than 30W, it is considered to have fluctuating edge loads.

[0017] Secondly, a sensing unit optimization configuration device for identifying edge devices in a distribution network is provided, comprising: a data processing module for collecting raw measurement data of the distribution network and preprocessing the data; a modeling module for establishing a multi-objective optimization model based on the collected data; a solution module for solving the multi-objective optimization model using an improved NSGA-II algorithm to obtain an optimized solution, wherein the improved NSGA-II algorithm introduces a directional mutation strategy and a historical archiving mechanism, wherein the directional mutation strategy guides the solution to move towards the normal direction of the Pareto optimal frontier, and the historical archiving mechanism saves high-quality solutions from previous iterations for directional hybridization with new solutions in subsequent iterations; a calculation module for selecting the Pareto optimal solution set based on the optimized solution through fast non-dominated sorting and congestion calculation to obtain the optimal configuration scheme of sensing units and edge units; a construction module for constructing a state estimation model of the distribution network based on the Pareto optimal solution set using the weighted least squares method; and an identification module for identifying edge device nodes by analyzing the power direction characteristics and power fluctuation characteristics of the nodes.

[0018] Thirdly, a computer device is provided, the computer device comprising: a memory and a processor, the memory storing at least one computer program, the at least one computer program being loaded and executed by the processor to implement the aforementioned method for optimizing the configuration of sensing units for identifying distribution network edge devices.

[0019] Fourthly, a computer-readable storage medium is provided, wherein at least one computer program is stored in the computer-readable storage medium, the at least one computer program being loaded and executed by a processor to implement the aforementioned method for optimizing the configuration of sensing units for identifying distribution network edge devices.

[0020] Fifthly, a computer program product is provided, including a computer program / instruction, which, when executed by a processor, implements the aforementioned method for optimizing the configuration of sensing units for identifying distribution network edge devices.

[0021] The beneficial effects of the technical solution provided in this application include at least the following: In this embodiment, an improved NSGA-II algorithm is used to solve a multi-objective optimization model to obtain an optimal solution. Then, based on the optimal solution, a Pareto optimal solution set is selected through fast non-dominated sorting and congestion calculation, yielding the optimal configuration scheme for sensing units and edge units. The improved NSGA-II algorithm introduces a directional mutation strategy and a historical archiving mechanism. The directional mutation strategy guides the solution towards the normal direction of the Pareto optimal frontier. The historical archiving mechanism saves high-quality solutions from previous iterations for directional hybridization with new solutions in subsequent iterations, thus providing biased guidance. The synergistic effect of the directional mutation strategy and the historical archiving mechanism makes the obtained solution set closer to the true frontier, with a wider distribution and better convergence efficiency, reducing the probability of the configuration scheme of sensing units and edge units getting trapped in local optima. Then, based on the Pareto optimal solution set, a state estimation model of the distribution network is constructed using the weighted least squares method. Considering the environment of an active distribution network, edge device node identification is performed by analyzing the power direction characteristics and power fluctuation characteristics of nodes, thereby achieving accurate sensing of user-side edge devices. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of a sensing unit optimization configuration method for identifying edge devices in a power distribution network, provided in an embodiment of this application. Figure 2 This is a schematic diagram of the Pareto front in the first example analysis of this application embodiment; Figure 3 This is a schematic diagram of the Pareto front in the fourth example analysis in the embodiments of this application; Figure 4 This is a schematic diagram of the edge device identification report provided in an embodiment of this application; Figure 5 This is a schematic diagram of the state estimation results of the distribution network provided in the embodiments of this application; Figure 6 This is a schematic diagram of the voltage amplitude distribution of the IEEE 33 system nodes provided in an embodiment of this application; Figure 7 This is a structural block diagram of a sensing unit optimization configuration device for identifying edge equipment in a power distribution network, provided in an embodiment of this application. Figure 8 This is a schematic diagram of the structure of the computer device provided in the embodiments of this application. Detailed Implementation

[0024] Unless otherwise defined, the technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this application pertains. The terms “first,” “second,” “third,” and similar terms used in this patent application specification and claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “an” or “a” and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms “comprising” or “including” and similar terms mean that the elements or objects preceding “comprising” or “including” encompass the elements or objects listed following “comprising” or “including” and their equivalents, and do not exclude other elements or objects.

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0026] Figure 1 This is a flowchart illustrating a method for optimizing the configuration of sensing units for identifying edge devices in a power distribution network, as provided in an embodiment of this application. Figure 1 As shown, the method includes: Step S1: Collect raw measurement data of the power distribution network and preprocess the data.

[0027] Step S2: Establish a multi-objective optimization model based on the collected data.

[0028] Step S3: Solve the multi-objective optimization model using the improved NSGA-II algorithm to obtain the optimal solution.

[0029] Among them, the improved NSGA-II algorithm introduces a directional mutation strategy and a history archiving mechanism. The directional mutation strategy is used to guide the solution to move towards the normal direction of the Pareto optimal frontier, and the history archiving mechanism saves the high-quality solutions in the previous iterations so that they can be directionally hybridized with new solutions in subsequent iterations.

[0030] Step S4: Based on the optimized solution, the Pareto optimal solution set is selected through fast non-dominated sorting and crowding calculation to obtain the optimal configuration scheme of sensing units and edge units.

[0031] Step S5: Construct a state estimation model for the distribution network based on the Pareto optimal solution set and the weighted least squares method.

[0032] Step S6: Identify edge device nodes by analyzing the power direction characteristics and power fluctuation characteristics of the nodes.

[0033] In this embodiment, an improved NSGA-II algorithm is used to solve a multi-objective optimization model to obtain an optimal solution. Then, based on the optimal solution, a Pareto optimal solution set is selected through fast non-dominated sorting and congestion calculation, yielding the optimal configuration scheme for sensing units and edge units. The improved NSGA-II algorithm introduces a directional mutation strategy and a historical archiving mechanism. The directional mutation strategy guides the solution towards the normal direction of the Pareto optimal frontier. The historical archiving mechanism saves high-quality solutions from previous iterations for directional hybridization with new solutions in subsequent iterations, thus providing biased guidance. The synergistic effect of the directional mutation strategy and the historical archiving mechanism makes the obtained solution set closer to the true frontier, with a wider distribution and better convergence efficiency, reducing the probability of the configuration scheme of sensing units and edge units getting trapped in local optima. Then, based on the Pareto optimal solution set, a state estimation model of the distribution network is constructed using the weighted least squares method. Considering the environment of an active distribution network, edge device node identification is performed by analyzing the power direction characteristics and power fluctuation characteristics of nodes, thereby achieving accurate sensing of user-side edge devices.

[0034] For example, in step S1, raw measurement data can be generated based on the existing basic measurement system and operation records of the distribution network. Here, the sources of raw measurement data include: SCADA, Remote Terminal Units (RTUs), electricity meters, Advanced Metering Infrastructure (AMI), Phasor Measurement Units (PMUs), and other sensing units, planning and design data, and manual inspection records. Raw measurement data includes node voltage, load data, and communication link data.

[0035] For example, after collecting the raw measurement data of the power distribution network, data preprocessing methods can be used to remove noise data, and data fusion technology can be used to integrate the data to ensure the comprehensiveness and accuracy of the data.

[0036] Alternatively, the data can be integrated using techniques such as weighted averaging or Kalman filtering. The formula for Kalman filtering is: (1) in, For state estimation using Kalman filtering, For measured values, For Kalman gain, This is the measurement matrix.

[0037] Optionally, in step S2, the multi-objective optimization model considers minimizing configuration costs and improving reliability. The objective function of the multi-objective optimization model includes a cost function and an average coverage function.

[0038] For example, the formula for calculating the cost function is: (2) in, For configuration cost, The unit price for the sensing unit. The number of sensing units. The purchase price of the edge unit. The number of edge units, The cost per unit length of communication channel, The total length of the communication channel to be laid.

[0039] For example, the unit price of the sensing unit can be provided by market research data, while the unit price of the edge unit can be determined based on actual procurement costs.

[0040] For example, the formula for calculating the average coverage function is as follows: (3) in, This represents the average perceived coverage rate. The number of system nodes. For the number The number of nodes covered by the sensing unit when a node fails. In this embodiment, the average sensing coverage rate is... It can be defined as: in The number is Single node failure and the rest The average sensing coverage rate of the sensing units under normal operating conditions of all nodes is used to measure the average performance of the overall sensing coverage level under the condition of failure of any single node. This can also be referred to as... Average coverage.

[0041] In this embodiment, parameters such as the candidate node set used in the cost function, the total length of the communication channel, the set of sensing unit types, the total number of nodes in the average coverage function, and the number of nodes covered in the fault scenario can all be determined through the above step S1. The multi-objective optimization model can ensure good economy and reliability in the optimized configuration of the distribution network sensing units by designing the cost function and the average coverage function.

[0042] For example, in step S3, the directional mutation strategy uses the covariance matrix eigenvector estimation method to guide the solution to move towards the normal direction of the Pareto optimal frontier. The formula for calculating the eigenvector is: (4) in, Let be the eigenvector corresponding to the smallest eigenvalue of the covariance matrix. For the neighborhood solutions in the solution set, This represents eigenvalue decomposition.

[0043] In this embodiment, the covariance matrix in the directed mutation strategy is calculated from the decision vector samples of the current generation of non-dominated elite solutions. Its eigenvalue decomposition gives the degree of dispersion in each direction. The eigenvector corresponding to the smallest eigenvalue can be selected as the normal direction estimate of the Pareto front. When performing mutation, a certain intensity of random perturbation is superimposed with this eigenvector as the direction, thereby achieving directed mutation along the normal direction.

[0044] For example, the historical archiving mechanism can be achieved by introducing an additional independent historical high-quality solution archive set A. This set contains representative non-dominated Pareto solutions selected during successive iterations, instead of being limited to individuals from the current parent and child generations as in the standard NSGA-II algorithm. When generating the next generation, parent individuals are selected from set A with a certain probability and cross-mutated with current population individuals through directed mutation. This allows the Pareto front information from previous generations to guide the current search direction. When the population diversity is detected to be below a threshold for several consecutive generations, some individuals are reintroduced from set A to replace the inferior individuals in the current population. This "rebooting and strengthening" of the population using historical high-quality solutions reduces the risk of getting trapped in local optima.

[0045] In this embodiment, compared with the standard NSGA-II algorithm, the historical archiving mechanism improves upon the previous Pareto frontier information by archiving high-quality solutions across generations, utilizing archived individuals to participate in crossover and directional mutation, and restarting based on archives when diversity is insufficient. This allows the Pareto frontier information of each generation to continuously participate in subsequent iterations, which is beneficial to improving global convergence while maintaining the diversity of solution sets.

[0046] For example, in step S4, the congestion calculation is performed according to the following formula: (5) in, For the first The crowding of each solution The number of objective functions, To find the value in the target space, This represents the distance between adjacent solutions in the target dimension.

[0047] For example, in step S5, the state estimation model of the distribution network based on the weighted least squares method is calculated according to the following formula: (6) in, For state estimation, For Jacobian matrices, For weighted matrices, This is the measurement vector.

[0048] In this embodiment, the measurement set used for state estimation, the composition of the measurement vectors, the row structure of the Jacobian matrix, and the dimensions and diagonal elements of the weighting matrix can be determined through the optimal configuration scheme of the sensing units and edge units in step S4 above. When constructing the specific values ​​of the Jacobian matrix and weighting matrix, they can be calculated using the optimal configuration scheme of the sensing units and edge units in step S4 combined with the data collected in step S1.

[0049] For example, during the state estimation process, node voltages can be used as basic state variables, and the voltage and power flow information of each node can be estimated using the least squares method.

[0050] Optionally, in step S6, edge device node identification is performed by analyzing the power direction characteristics and power fluctuation characteristics of the node, including: if the node power is negative for a certain period of time and the ratio of the duration of negative power to the total monitoring time is greater than 40%, it is determined to be an edge device node; if the standard deviation of the node power is greater than 30W, it is considered to have fluctuating edge load.

[0051] For example, the formula for calculating the standard deviation of node power is: (7) in, The standard deviation of power, For the number of time points, The power value at each time point, This represents the average power.

[0052] Optionally, after completing the identification of edge device nodes, the method further includes: outputting the state estimation results of the distribution network and the edge device identification report.

[0053] For example, the state estimation results of the distribution network include node voltage distribution maps, and the edge device identification report includes the identification results of edge device nodes and the power characteristic analysis report of edge devices. Here, the power characteristic analysis report of edge devices mainly targets edge device nodes and is used to summarize and visualize the power behavior of edge device nodes. The content of the power characteristic analysis report of edge devices includes, but is not limited to: active power time series curves for typical periods, power sign distribution and the proportion of negative power duration (i.e., the ratio of negative power duration to total monitoring time), power standard deviation, peak-valley difference and other statistical quantities characterizing volatility, as well as maximum, minimum, average and peak-valley difference statistics on daily or weekly scales.

[0054] The effects of this application will be further illustrated below through computer simulation.

[0055] To verify the feasibility of the aforementioned method for optimizing the configuration of sensing units for identifying distribution network edge devices, a simulation was performed on an IEEE 33-bus system using MATLAB 2023b and MatPower 8.0 on a Windows 10 system. To avoid the model getting trapped in local optima, four simulations were conducted.

[0056] Figure 2 This is a schematic diagram of the Pareto front in the first example analysis of this application. Figure 3 This is a schematic diagram of the Pareto front in the fourth example analysis of this application. See also... Figure 2 and Figure 3 Analysis of the Pareto solution plane reveals that the feasible schemes calculated using the sensing unit optimization configuration method of this application are roughly evenly distributed on both sides of the Pareto front, which is consistent with the actual distribution of the Pareto front. Furthermore, the distribution is not uniform left-right in the plane, but concentrated in the target region of this study. This indicates that the improved NSGA-II algorithm used in this application, guided by a reasonable strategy, possesses strong adaptability and convergence in handling the direction of optimal configuration of sensing units in the distribution network.

[0057] Table 1 Model running results

[0058] As shown in Table 1, after four simulation examples, the optimized model yielded optimal configuration schemes in three directions. In the cost minimization direction, while ensuring a certain level of reliability, a system with only 11 sensing units and 2 edge units was achieved, reducing the cost to 46,150.99 yuan. In the reliability direction, configuring 22 sensing units and four edge units ensured sufficient reliability with an average coverage rate of 99.36%, but the cost of this configuration increased by more than 40,000 yuan compared to the cost minimization scheme, thus reducing its economic efficiency. The balance point scheme, which balances reliability and cost, increases the number of sensing units by two and changes the installation nodes of the sensing and edge units. Compared to the cost minimization scheme, it only increases the cost by 6,017.17 yuan while improving reliability by 0.43%, demonstrating strong cost-effectiveness.

[0059] Table 2 Computation Time and Computation Scale

[0060] As shown in Table 2, the improved NSGA-II algorithm in this application can help optimize the configuration scheme of the model to a certain extent, and improve the performance of the solution set to some extent. However, due to the inherent characteristics of the NSGA-II algorithm, the completely random perturbation brought by the non-directional mutation strategy cannot guarantee that each run will optimize in a positive direction, and there is a small probability that the performance will not change or will deteriorate. It may be possible to appropriately improve the performance by increasing the directional mutation probability. Overall, through continuous iterative calculations and learning from historical high-quality solutions, the performance of the scheme is continuously improved. Furthermore, the computation time is generally in the range of 60s to 70s, which is relatively fast, indicating that the method in this application has good performance in processing the optimized configuration of distribution network sensing units.

[0061] This application incorporates the derived cost-minimizing sensing unit configuration scheme into a distribution network state estimation model containing edge devices for edge device identification and configuration scheme performance testing. IEEE 33-bus system parameters are directly imported from the Matpower 8.0 toolbox.

[0062] Distributed photovoltaic energy storage was installed at nodes 13, 23, and 31 of IEEE 33 nodes through random generation and installation of the system. Figure 4 This is a schematic diagram of the edge device identification report provided in an embodiment of this application. For example... Figure 4 As shown, the identification strategy in this application, which analyzes the power direction characteristics and power fluctuation characteristics of nodes, can identify most of the edge devices of the nodes with an accuracy of more than 60%. Only a small number of nodes are not identified and need to be recalled for re-identification and detection.

[0063] Figure 5 This is a schematic diagram of the state estimation results of the distribution network provided in the embodiments of this application. Figure 6 This is a schematic diagram of the node voltage amplitude distribution in an embodiment of the IEEE 33 system provided in this application. By... Figure 5 and Figure 6 The comparison shows that the sensing unit optimization configuration method in this application provides a similar distribution of node voltage amplitude state estimation results to the actual situation. In the directions of the main feeder (directions 1-18), the voltage amplitude decreases, but the overall magnitude is within 10%, which does not exceed the specified ±10% allowable range, proving that this method is relatively successful in estimating the state of the system node voltage.

[0064] Table 3 Analysis of Interval State Estimation Results ( (Voltage over-limit indicator)

[0065] Table 3 shows that nodes at the end of the line, such as nodes 18 and 33, exhibited significant fluctuations in their estimated states. However, node 33, equipped with a sensing unit, had a certain level of measurement accuracy, thus its voltage did not exceed the limit. Node 18, on the other hand, did experience voltage exceedances. Nodes 13, 23, and 31, connected to edge devices, showed large voltage fluctuations, with some nodes exceeding their voltage limits. Overall, however, the voltage fluctuations at most nodes did not significantly exceed the limits or surpass reasonable thresholds. This demonstrates that the optimized sensing unit configuration scheme has relatively good monitoring performance.

[0066] The following are device embodiments of this application. For details not described in detail in the device embodiments, please refer to the above method embodiments.

[0067] Figure 7 This is a structural block diagram of a sensing unit optimization configuration device for identifying edge equipment in a power distribution network, provided in an embodiment of this application. Figure 7 As shown, the sensing unit optimization configuration device 100 includes: a data processing module 101, an establishment module 102, a solution module 103, a calculation module 104, a construction module 105, and an identification module 106.

[0068] The system comprises the following modules: Data Processing Module 101 collects raw measurement data of the distribution network and preprocesses the data; Modeling Module 102 establishes a multi-objective optimization model based on the collected data; Solution Module 103 uses an improved NSGA-II algorithm to solve the multi-objective optimization model to obtain the optimal solution. The improved NSGA-II algorithm introduces a directional mutation strategy and a history archiving mechanism. The directional mutation strategy guides the solution to move towards the normal direction of the Pareto optimal frontier, and the history archiving mechanism saves high-quality solutions from previous iterations for directional hybridization with new solutions in subsequent iterations; Calculation Module 104 selects the Pareto optimal solution set based on the optimal solution using fast non-dominated sorting and congestion calculation to obtain the optimal configuration scheme for sensing units and edge units; Construction Module 105 constructs a state estimation model of the distribution network based on the Pareto optimal solution set using the weighted least squares method; and Identification Module 106 identifies edge equipment nodes by analyzing the power direction characteristics and power fluctuation characteristics of the nodes.

[0069] It should be noted that the above embodiments of the sensing unit optimization configuration device for identifying distribution network edge equipment are only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the sensing unit optimization configuration device for identifying distribution network edge equipment provided in the above embodiments and the sensing unit optimization configuration method embodiments for identifying distribution network edge equipment belong to the same concept. The specific implementation process is detailed in the method embodiments and will not be repeated here.

[0070] The module division in this embodiment is illustrative and represents only one logical functional division. In actual implementation, other division methods are possible. Furthermore, the functional modules in each embodiment of this application can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0071] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a terminal device (which may be a personal computer, mobile phone, or communication device, etc.) or processor to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0072] Figure 8 This is a schematic diagram of the structure of the computer device provided in an embodiment of this application. Figure 8 As shown, the computer device 200 includes a processor 201 and a memory 202.

[0073] Processor 201 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 201 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). Processor 201 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 201 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content required to be displayed on the screen. In some embodiments, processor 201 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0074] The memory 202 may include one or more computer-readable storage media, which may be non-transitory. The memory 202 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 202 is used to store at least one instruction, which is executed by the processor 201 to implement the sensing unit optimization configuration method for identifying distribution network edge devices provided in this embodiment.

[0075] Those skilled in the art will understand that Figure 8 The structure shown does not constitute a limitation on the computer device 200, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0076] This application also provides a non-transitory computer-readable storage medium, which, when the instructions in the storage medium are executed by the processor of a computer device, enables the computer device to execute the sensing unit optimization configuration method for identifying distribution network edge devices provided in this application.

[0077] This application also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the sensing unit optimization configuration method for identifying distribution network edge devices provided in this application.

[0078] The above description is merely an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for optimizing the configuration of sensing units for identifying edge equipment in a distribution network, characterized in that, include: Collect raw measurement data of the power distribution network and preprocess the data; A multi-objective optimization model is established based on the collected data; The improved NSGA-II algorithm is used to solve the multi-objective optimization model to obtain the optimal solution. The improved NSGA-II algorithm introduces a directional mutation strategy and a history archiving mechanism. The directional mutation strategy is used to guide the solution to move towards the normal direction of the Pareto optimal frontier. The history archiving mechanism saves the high-quality solutions in each iteration process so that they can be directionally hybridized with new solutions in subsequent iterations. Based on the optimized solution, the Pareto optimal solution set is selected through fast non-dominated sorting and crowding calculation to obtain the optimal configuration scheme of sensing units and edge units; Based on the Pareto optimal solution set, a state estimation model for the distribution network is constructed using the weighted least squares method. Edge device nodes are identified by analyzing the power direction and power fluctuation characteristics of the nodes.

2. The method for optimizing the configuration of sensing units for identifying edge equipment in a distribution network according to claim 1, characterized in that, The multi-objective optimization model considers minimizing configuration costs and improving reliability. The objective functions of the multi-objective optimization model include a cost function and an average coverage function. The formula for calculating the cost function is as follows: in, For configuration cost, The unit price for the sensing unit. The number of sensing units. The purchase price of the edge unit. The number of edge units, The cost per unit length of communication channel, The total length of the communication channel to be laid; The formula for calculating the average coverage function is as follows: in, This represents the average perceived coverage rate. The number of system nodes. For the number The number of nodes covered by the sensing unit when a node fails.

3. The method for optimizing the configuration of sensing units for identifying distribution network edge equipment according to claim 2, characterized in that, The directional mutation strategy uses the covariance matrix eigenvector estimation method to guide the solution to move towards the normal direction of the Pareto optimal frontier. The formula for calculating the eigenvector is: in, Let be the eigenvector corresponding to the smallest eigenvalue of the covariance matrix. For the neighborhood solutions in the solution set, This represents eigenvalue decomposition.

4. The method for optimizing the configuration of sensing units for identifying distribution network edge equipment according to claim 3, characterized in that, The congestion level is calculated according to the following formula: in, For the first The crowding of each solution The number of objective functions, To find the value in the target space, This represents the distance between adjacent solutions in the target dimension.

5. The method for optimizing the configuration of sensing units for identifying edge equipment in a distribution network according to claim 4, characterized in that, The state estimation model for the distribution network based on the weighted least squares method is calculated according to the following formula: in, For state estimation, For Jacobian matrices, For weighted matrices, This is the measurement vector.

6. The method for optimizing the configuration of sensing units for identifying distribution network edge equipment according to claim 5, characterized in that, The method of identifying edge device nodes by analyzing the power direction characteristics and power fluctuation characteristics of nodes includes: If a node's power is negative for a certain period of time and the ratio of the duration of negative power to the total monitoring time is greater than 40%, it is determined to be an edge device node. If the standard deviation of node power is greater than 30W, it is considered that there is a fluctuating edge load.

7. A sensing unit optimization configuration device for identifying edge equipment in a power distribution network, characterized in that, include: The data processing module is used to collect raw measurement data of the power distribution network and preprocess the data; A module is established to build a multi-objective optimization model based on the collected data; The solution module is used to solve the multi-objective optimization model using the improved NSGA-II algorithm to obtain the optimal solution. The improved NSGA-II algorithm introduces a directional mutation strategy and a history archiving mechanism. The directional mutation strategy is used to guide the solution to move towards the normal direction of the Pareto optimal frontier. The history archiving mechanism saves the high-quality solutions in the previous iterations so that they can be directionally hybridized with new solutions in subsequent iterations. The calculation module is used to filter the Pareto optimal solution set based on the optimization solution through fast non-dominated sorting and crowding calculation, and obtain the optimal configuration scheme of the sensing unit and the edge unit. The module is used to construct a state estimation model of the distribution network based on the Pareto optimal solution set using the weighted least squares method. The identification module is used to identify edge device nodes by analyzing the power direction characteristics and power fluctuation characteristics of the nodes.

8. A computer device, characterized in that, The computer device includes a memory and a processor, wherein the memory stores at least one computer program, which is loaded and executed by the processor to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to implement the method according to any one of claims 1 to 6.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the method described in any one of claims 1 to 6.

Images