A power distribution network device information intelligent management system and method
By analyzing equipment attributes and using digital twin technology to assess static impact coefficients and dynamic perception levels, and combining multi-dimensional data, equipment categories are classified and differentiated management is implemented. This solves the problems of resource allocation imbalance and low maintenance efficiency in power distribution network equipment management, and achieves precise and efficient equipment management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HANGZHOU HANGGANG CLOUD COMPUTING DATA CENTER CO LTD
- Filing Date
- 2025-12-01
- Publication Date
- 2026-07-31
AI Technical Summary
The existing intelligent management system for power distribution network equipment lacks a comprehensive assessment of the static impact coefficient and dynamic perception level of equipment, resulting in a single judgment of equipment criticality, unbalanced resource allocation, rigid management mode, low maintenance efficiency, and an inability to achieve precise and efficient management of equipment status.
The equipment attribute analysis module is used to assess the static impact coefficient and dynamic perception level. The impact of failures is simulated through digital twin technology. Combined with multi-dimensional historical cost data, the equipment is divided into three categories: Category I key maintenance, Category II ordinary maintenance, and Category III routine maintenance. Differentiated health management and collaborative maintenance strategies are implemented to optimize resource allocation and maintenance tasks.
It enables accurate assessment and quantitative management of critical power distribution network equipment, optimizes resource allocation, improves the efficiency and reliability of equipment maintenance, reduces operation and maintenance costs, and enhances the predictive maintenance capability of equipment status.
Smart Images
Figure CN121599650B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power distribution network equipment management technology, and specifically to an intelligent management system and method for power distribution network equipment information. Background Technology
[0002] As the power grid continues to expand and the number of distribution network devices surges, traditional operation and maintenance models, which rely on manual experience and regular inspections, struggle to achieve precise and efficient management. Problems such as insufficient equipment status awareness, lack of differentiated maintenance strategies, and unscientific resource allocation are becoming increasingly prominent, necessitating the introduction of intelligent methods to achieve refined management of equipment throughout its entire lifecycle and improve the reliability and economy of power grid operation.
[0003] Existing or traditional intelligent management systems and methods for equipment information in power distribution networks have at least the following technical problems:
[0004] 1. Existing intelligent management systems and methods for equipment information in power distribution networks lack a comprehensive evaluation mechanism for the static impact coefficient and dynamic perception level of equipment. This results in a single dimension for judging the criticality of equipment. Traditional management models only classify equipment based on basic attributes such as equipment model or years of operation, failing to quantify and evaluate the actual impact of equipment failures on the reliability of power supply to users. At the same time, they ignore the differences in the real-time data acquisition capabilities of equipment, making it impossible to use maintenance resources on high-risk equipment. This leads to an imbalance in resource allocation, with insufficient maintenance of important load node equipment and excessive maintenance of non-critical equipment.
[0005] 2. Traditional intelligent management systems and methods for equipment information in distribution networks lack threshold adaptive optimization methods driven by multi-dimensional historical cost data. This leads to a disconnect between the perception level classification standards and the actual situation. Existing systems typically use fixed thresholds to judge equipment status, failing to dynamically adjust based on historical operation and maintenance data, such as over-maintenance costs, false alarm handling costs, and fault loss costs. Static threshold setting methods are difficult to adapt to the operating characteristics of distribution networks in different regions, resulting in misjudgments and missed judgments due to unreasonable threshold settings.
[0006] 3. Existing intelligent management systems and methods for equipment information in power distribution networks lack differentiated health management strategies for different equipment categories, resulting in rigid and inefficient maintenance models. Traditional methods often adopt a uniform monitoring frequency and maintenance cycle for all equipment, lacking hierarchical management based on the criticality of the equipment. This results in critical equipment not receiving sufficient monitoring and maintenance attention, while secondary equipment occupies too many operation and maintenance resources, thus failing to achieve early warning of equipment condition deterioration.
[0007] 4. Traditional intelligent management systems and methods for equipment information in power distribution networks lack a multi-category maintenance task collaboration mechanism based on spatiotemporal correlation, resulting in low on-site operation efficiency and insufficient resource utilization. Existing systems often generate and execute maintenance work orders separately according to equipment category, without considering the possibility of collaboration when maintenance tasks of different categories of equipment overlap in time windows or are geographically close. The isolated scheduling mode causes maintenance personnel to frequently travel to different sites, resulting in low utilization of dedicated testing equipment, increased travel time and costs, and reduced overall operation and maintenance efficiency, failing to maximize the value of limited maintenance resources. Summary of the Invention
[0008] To address the aforementioned shortcomings of existing technologies, this invention provides an intelligent management system and method for equipment information in power distribution networks, which can effectively solve the problems of unscientific resource allocation and low maintenance efficiency in the background technology.
[0009] To solve the above technical problems, the present invention adopts the following technical solution: The present invention provides an intelligent management system for equipment information in a power distribution network, comprising:
[0010] The equipment attribute analysis module is used to obtain the management equipment in the distribution network corresponding to the management area, evaluate the static impact coefficient of each management equipment on the distribution network when it fails, and classify the dynamic perception level of each management equipment.
[0011] The management category delineation module is used to classify each device to be managed based on its static impact coefficient and dynamic perception level.
[0012] The equipment health management module is used to perform health management on each piece of equipment in each management category according to the classification results of the management categories corresponding to each piece of equipment to be managed.
[0013] The collaborative maintenance module is used to perform collaborative health maintenance on each device under management based on the health management results of each device.
[0014] Preferably, the process of evaluating the static impact coefficient of each managed device on the distribution network during a fault is as follows: obtaining the number, type, and topological relationship of each managed device, and obtaining the total number of users in the distribution network of the managed area.
[0015] Set the importance level coefficient of the load supplied by each device to be managed, and obtain the number of users in the corresponding distribution network when each device to be managed fails and stops.
[0016] By combining the total number of users in the distribution network under management, the importance level coefficient of the load supplied by each device under management, and the number of users affected by the corresponding distribution network when each device under management stops due to a fault, the static impact coefficient of each device under management on the distribution network during a fault is calculated.
[0017] Preferably, the process of dividing each managed device into dynamic sensing levels is as follows: the dynamic sensing levels are divided into full sensing level, basic sensing level and no sensing level.
[0018] The data update frequency of each device under management is obtained from the data monitoring points. Combined with the preset fast data update frequency threshold and slow data update frequency threshold, the dynamic perception level of each device under management is classified.
[0019] Based on the divided devices to be managed, corresponding level thresholds are set for the full perception level, basic perception level, and no perception level of each device to be managed, which are respectively denoted as the full perception level threshold, basic perception level threshold, and no perception level threshold.
[0020] Preferably, the process of setting the fast data update frequency threshold, slow data update frequency threshold, full perception level threshold, basic perception level threshold, and no perception level threshold is as follows: constructing a threshold dataset, constructing each set of threshold datasets, and dynamically classifying each device to be managed based on each set of threshold datasets.
[0021] Obtain the historical over-maintenance costs, historical false alarm handling costs, and historical failure loss costs for each device to be managed.
[0022] By combining the average historical over-maintenance cost, average historical false alarm handling cost, and average historical failure loss cost, as well as the set over-maintenance weight, false alarm weight, and failure loss weight, the comprehensive loss value corresponding to each group of threshold datasets is calculated.
[0023] Based on the comprehensive loss value corresponding to each set of threshold datasets, the values of fast data update frequency threshold, slow data update frequency threshold, full perception level threshold, basic perception level threshold, and no perception level threshold are obtained.
[0024] Preferably, the process of dividing the devices to be managed is as follows: combining the static influence coefficient and dynamic perception level of each device to be managed, the comprehensive criticality index corresponding to each device to be managed is calculated.
[0025] Based on the comprehensive criticality index corresponding to each device to be managed, the devices to be managed are divided into categories.
[0026] Preferably, the health management of each piece of equipment to be managed in each management category is carried out in the following specific process: the health management of each piece of equipment to be managed includes the health management of Class I key maintenance equipment, the health management of Class II ordinary maintenance equipment, and the health management of Class III routine maintenance equipment.
[0027] Based on the set data collection time points of Category I, the status parameters of each managed device belonging to Category I are collected, and the normalized values of each status parameter of each managed device are obtained through function analysis.
[0028] By combining the normalized values of the status parameters of each device under management with the corresponding weighting factors, the health assessment index of each device under management is calculated.
[0029] By combining the health assessment index corresponding to each device to be managed, and the health assessment index of each device at the current data collection time and the historical data collection time, the predicted health assessment index of each device to be managed at the future data collection time is calculated.
[0030] Health management is implemented for Class I key maintenance equipment by combining health assessment index thresholds and predicted health assessment index thresholds.
[0031] Preferably, the health management of the Class II general maintenance equipment is carried out as follows: based on the set Class II data collection time point, key parameters of each Class II equipment to be managed are collected.
[0032] Based on the weighting factors corresponding to the key parameters of each managed device belonging to Category II, calculate the device status score of each managed device belonging to Category II.
[0033] Obtain the conservative basic maintenance cycle for each managed device belonging to Category II, and calculate the maintenance time interval between each managed device and the next maintenance by combining the device status score of each managed device belonging to Category II and the set adjustment coefficient.
[0034] Based on the runtime of each device to be managed, the maintenance interval before the next maintenance, the conservative basic maintenance cycle, and the Class II device status scoring threshold, health management is performed on Class II ordinary maintenance devices.
[0035] Preferably, the health management of the Class III routine maintenance equipment is carried out as follows: based on the set Class III data collection time point, the communication status of each Class III device to be managed is collected.
[0036] When the communication status of a managed device is interrupted and the interruption duration exceeds the set time threshold for distinguishing between instantaneous communication fluctuations and actual device failures, the fault determination condition is met, and a fault warning is triggered.
[0037] After triggering a fault warning, the number of affected users in the distribution network corresponding to the area to be evaluated is obtained. Based on the threshold range of the number of affected users corresponding to each response level, health management of Class III routine maintenance equipment is performed.
[0038] Preferably, the collaborative health maintenance of each device to be managed is carried out as follows: Based on the maintenance work orders of type I, type II and type III, the corresponding maintenance execution time points are obtained. When the maintenance execution time points corresponding to two or three types of maintenance work orders are on the same working day or within a preset time window, the various types of maintenance work orders are grouped into a candidate collaborative task set, and then collaborative health maintenance is carried out on the candidate collaborative task set.
[0039] In a second aspect, the present invention provides an intelligent management method for equipment information in a power distribution network, comprising:
[0040] S1. Obtain the management equipment in the distribution network corresponding to the management area, evaluate the static impact coefficient of each management equipment on the distribution network during a fault, and classify the dynamic perception level of each management equipment.
[0041] S2. Based on the static impact coefficient and dynamic perception level of each device to be managed, classify the devices to be managed.
[0042] S3. Based on the management category classification results of each device to be managed, perform health management on each device to be managed in each management category.
[0043] S4. Based on the health management results of each device to be managed, conduct collaborative health maintenance on each device to be managed.
[0044] The technical solution provided by this invention has the following advantages compared with the known prior art:
[0045] 1. In the process of equipment attribute analysis, the embodiments of the present invention introduce a dual evaluation system of static influence coefficient and dynamic perception level, and use digital twin technology to simulate the impact of equipment failure. This is conducive to achieving accurate and quantitative evaluation of the criticality of distribution network equipment. The multi-dimensional evaluation method changes the traditional extensive management model that relies on single experience, and provides a scientific data foundation for the subsequent formulation of differentiated operation and maintenance strategies.
[0046] 2. In the process of classifying management categories, this invention obtains a comprehensive criticality index by multiplying the static impact coefficient by the dynamic perception level, and uses a fixed ratio method to classify equipment into three categories: Category I key maintenance, Category II ordinary maintenance, and Category III routine maintenance. This is conducive to building a clear and focused equipment management team. The refined classification system ensures that management resources can be accurately allocated according to the actual importance and risk level of the equipment, avoiding the drawbacks of "one-size-fits-all" management.
[0047] 3. In the equipment health management process, the embodiments of the present invention implement distinctly different management strategies for different categories of equipment. For Class I equipment, real-time status parameter collection, health index calculation, and future trend prediction are performed. For Class II equipment, the maintenance cycle is dynamically adjusted based on key parameter scores. For Class III equipment, the communication status is mainly monitored and a response is given after a fault timeout. This facilitates the upgrade from preventive maintenance to predictive maintenance and status-based maintenance, and helps to significantly optimize the operation and maintenance costs of ordinary equipment while ensuring the extremely high reliability of critical equipment.
[0048] 4. In the collaborative maintenance decision-making process, the embodiments of the present invention integrate various maintenance work orders according to their execution time points and evaluate whether they meet the requirements of geographical distance and sharing of human resources and equipment resources within a preset time window. This helps to optimize the originally isolated maintenance tasks into efficient collaborative work groups, which helps to minimize the repeated trips of the operation and maintenance team and the repeated start-ups and shutdowns of equipment, thereby improving the overall operation and maintenance efficiency and reducing the scheduling costs of manpower and materials. Attached Figure Description
[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.
[0050] Figure 1 This is a schematic diagram of the system structure connection of the present invention.
[0051] Figure 2 This is a schematic diagram of the implementation steps of the present invention. Detailed Implementation
[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0053] The present invention will be further described below with reference to embodiments.
[0054] Please see Figure 1 As shown, an intelligent management system for equipment information in a power distribution network includes at least:
[0055] The equipment attribute analysis module is used to obtain the management equipment in the distribution network corresponding to the management area, evaluate the static impact coefficient of each management equipment on the distribution network when it fails, and classify the dynamic perception level of each management equipment.
[0056] In a specific embodiment, the process of evaluating the static impact coefficient of each managed device on the distribution network during a fault is as follows: Based on the standard data interface, the number, type and topology of each managed device are obtained from the database of the managed area, and the total number of users in the distribution network of the managed area is obtained.
[0057] Based on each user profile, an importance level coefficient is set for the load supplied by each managed device. A digital twin of the distribution network in the managed area is constructed. The failure and shutdown of each managed device are simulated in the digital twin to calculate the number of affected users in the corresponding distribution network.
[0058] Combined with the total number of users in the distribution network of the area to be managed Importance level coefficient of the load supplied by each device to be managed And the number of affected users in the distribution network corresponding to each managed device that fails and stops operating. Through the calculation formula: The static impact coefficients of each managed device on the distribution network during a fault were calculated. ,in Assign a number to each device to be managed. and These are respectively represented as the weighting factor for the number of users and the weighting factor for the importance level coefficient of the load supplied by the device.
[0059] It should be noted that, The values are directly based on the classification of power users in documents such as the "Technical Standard for Configuration of Power Supply and Self-provided Emergency Power Supply for Important Power Users" issued by the industry. In this scheme, they are used to assess the static impact of equipment failure on the distribution network. Combined with the number of users, the static impact coefficient is calculated to identify and prioritize the management of equipment that has a greater impact on the operation of the power grid after failure. and The values range from 0 to 1.
[0060] It should be noted that, and The setting process is based on existing multi-criteria decision analysis or expert evaluation methods, and is used to balance the relative contributions of the number of users and the importance of the load in the overall evaluation. It is determined through the analytic hierarchy process, the Delphi method or historical data statistical analysis, and will not be elaborated further here.
[0061] It should be noted that the construction of a digital twin of the distribution network in the area to be managed is based on the existing physical topology of the distribution network, equipment parameters, and user connection relationships. In this scheme, it is used to simulate and evaluate the state of the distribution network and the state of the equipment to be managed after equipment failure. By modeling the connection relationships between nodes (buses, switches) and edges (lines, transformers) in the power grid, and associating user profiles with their power supply nodes, a virtual power grid model is formed. When a managed device (such as a feeder or a transformer) fails and goes out of service in the digital twin, the system will automatically traverse and identify all user nodes that have lost electrical connection with the unique power source in the area due to this failure based on connectivity analysis algorithms in graph theory (such as breadth-first search or depth-first search), and then count the number of affected users.
[0062] In a specific embodiment, the process of dividing each managed device into dynamic sensing levels is as follows: the dynamic sensing levels are divided into full sensing level, basic sensing level and no sensing level.
[0063] The system acquires the data update frequency of each monitored data point of the managed device. Based on preset fast and slow data update frequency thresholds, if the data update frequency of a managed device is greater than the fast data update frequency threshold, the dynamic perception level of the managed device is determined to be the full perception level. If the data update frequency of a managed device is less than or equal to the fast data update frequency threshold and greater than or equal to the slow data update frequency threshold, the dynamic perception level of the managed device is determined to be the basic perception level. If the data update frequency of a managed device is less than the slow data update frequency threshold, the dynamic perception level of the managed device is determined to be the no perception level.
[0064] Based on the full perception level, basic perception level, and no perception level, corresponding level thresholds are set, which are respectively denoted as the full perception level threshold, basic perception level threshold, and no perception level threshold.
[0065] It should be noted that obtaining the data update frequency of each data monitoring point of the managed equipment refers to querying the timestamp sequence of data packets within a preset time window from the database connected to the equipment data monitoring point (such as a time series database deployed on the power distribution IoT platform), and determining it by calculating the average interval between adjacent timestamps; for example, for a transformer oil temperature monitoring point, if the database shows that 12 data points with precise timestamps were recorded in the most recent hour, and the time interval between adjacent data points is 5 minutes, then the current data update frequency of the monitoring point can be determined to be 5 minutes / time.
[0066] In a specific embodiment, the process of setting the data update fast frequency threshold, data update slow frequency threshold, full perception level threshold, basic perception level threshold and no perception level threshold is as follows: construct a threshold dataset, which includes the data update fast frequency threshold, data update slow frequency threshold, full perception level threshold, basic perception level threshold and no perception level threshold.
[0067] By optimizing the algorithm, threshold datasets are generated for each group. Based on these threshold datasets, dynamic perception levels are assigned to each device to be managed.
[0068] Based on the dynamic perception level of each device to be managed, the historical over-maintenance cost, historical false alarm handling cost, and historical failure loss cost corresponding to each device to be managed are obtained from the database. The average historical over-maintenance cost, average historical false alarm handling cost, and average historical failure loss cost corresponding to each device to be managed are calculated by averaging.
[0069] Through the loss function: Calculate the first The comprehensive loss value corresponding to the group threshold dataset ,in For each group of threshold datasets, The numbers corresponding to each threshold dataset are: The value is a positive integer. , and These are the corresponding over-maintenance weight, false alarm weight, and fault loss weight, respectively. , and These are respectively represented as average historical over-maintenance cost, average historical false alarm handling cost, and average historical failure loss cost.
[0070] Based on the autonomous decision-making mechanism of the optimization algorithm, the threshold dataset with the lowest comprehensive loss value among the threshold datasets is selected, and the threshold corresponding to this threshold dataset is used as the value of the fast data update frequency threshold, slow data update frequency threshold, full perception level threshold, basic perception level threshold, and no perception level threshold.
[0071] It should be noted that, , and The values are all greater than and less than , , and The settings are directly derived from the specific management strategies of the power grid operator. The process involves using expert decision-making methods such as the Delphi method to quantify qualitative management objectives into fixed weight values. For example, if the management strategy clearly prioritizes power supply reliability over cost control, the expert team will assign a weight to fault loss accordingly. A high value (e.g., 0.6) reduces over-maintenance weights. (e.g., 0.3) and false alarm weight (e.g., 0.1), thereby ensuring that the subsequent parameter optimization process always revolves around the core objective of minimizing power outage losses.
[0072] It should be noted that optimization algorithms (such as genetic algorithms and particle swarm optimization algorithms) are intelligent search strategies in existing technologies. In this solution, they are specifically applied as an automatic parameter optimization mechanism. First, an initial threshold dataset (i.e., multiple candidate solutions) representing different classification criteria is generated in a predefined parameter space. Then, each threshold is substituted into the dynamic perception level classification model for performance evaluation (i.e., calculating its corresponding historical comprehensive operation and maintenance loss value). Based on the evaluation results, a survival of the fittest process is simulated. Through iterative selection, crossover, and mutation operations, the system autonomously decides and converges to the optimal threshold dataset that minimizes the comprehensive loss function.
[0073] In the process of equipment attribute analysis, this invention introduces a dual evaluation system of static impact coefficient and dynamic perception level, and uses digital twin technology to simulate the impact of equipment failure. This facilitates accurate and quantitative evaluation of the criticality of distribution network equipment. The multi-dimensional evaluation method changes the traditional extensive management model that relies on single experience, and provides a scientific data foundation for the subsequent formulation of differentiated operation and maintenance strategies.
[0074] The management category delineation module is used to classify each device to be managed based on its static impact coefficient and dynamic perception level.
[0075] In a specific embodiment, the process of dividing each device to be managed is as follows: combining the static influence coefficient and dynamic perception level of each device to be managed, the static influence coefficient of each device to be managed is multiplied by the corresponding dynamic perception level to obtain the comprehensive criticality index of each device to be managed.
[0076] Based on the comprehensive criticality index corresponding to each piece of equipment to be managed, a fixed proportion method is used to classify them. The comprehensive criticality index corresponding to each piece of equipment to be managed is sorted in descending order. The top 20% of the equipment to be managed is classified as Class I key maintenance equipment, the bottom 30% of the equipment to be managed is classified as Class III routine maintenance equipment, and the equipment to be managed ranked between 20% and 70% is classified as Class II ordinary maintenance equipment.
[0077] In the process of classifying management categories, this invention obtains a comprehensive criticality index by multiplying the static impact coefficient by the dynamic perception level, and uses a fixed ratio method to divide equipment into three categories: Category I key maintenance, Category II ordinary maintenance, and Category III routine maintenance. This is conducive to building a clear and focused equipment management hierarchy. The refined classification system ensures that management resources can be accurately allocated according to the actual importance and risk level of the equipment, avoiding the drawbacks of a "one-size-fits-all" management approach.
[0078] The equipment health management module is used to perform health management on each piece of equipment in each management category according to the classification results of the management categories corresponding to each piece of equipment to be managed.
[0079] In a specific embodiment, the health management of each device to be managed in each management category is carried out as follows: the health management of each device to be managed includes the health management of Class I key maintenance devices, the health management of Class II ordinary maintenance devices, and the health management of Class III routine maintenance devices.
[0080] Based on the set data collection time points of Category I, the status parameters of each managed device belonging to Category I are collected, and the status parameters of each managed device belonging to Category I are recorded as follows: , This refers to the serial number of each piece of equipment to be managed belonging to Category I. The value of is a positive integer. For the first Each device to be managed corresponds to a number for its various status parameters. , The value of is a positive integer. This represents the total number of status parameters for each device to be managed.
[0081] Based on the status parameters of each managed device belonging to Category I Through the function: , obtained the The number of devices to be managed corresponds to the first Normalized values of the state parameters , and Represented as the first The number of devices to be managed corresponds to the first The upper and lower limits for safe operation of the status parameters.
[0082] Calculation formula: , obtained the Health assessment index corresponding to each device under management , For the first The number of devices to be managed corresponds to the first Weighting factors for the state parameters of an item.
[0083] Based on the health assessment index corresponding to each device to be managed, the calculation formula is as follows: , obtained the Each device to be managed corresponds to the future Predicted Health Assessment Index at Collection Time Point , The data collection time points are numbered. The value of is a positive integer. and Represented as the first The number of devices to be managed is in the first Each collection time point and Health assessment index at the time of data collection and These represent the number of historical data collection points set and the number of future data collection points set, respectively.
[0084] The health assessment index and predicted health assessment index of each device to be managed are compared with the corresponding health assessment index threshold and predicted health assessment index threshold, respectively. If the health assessment index of a device to be managed is less than the health assessment index threshold, or the predicted health assessment index is less than the predicted health assessment index threshold, a maintenance warning is triggered and a maintenance work order is generated; otherwise, no maintenance warning is triggered.
[0085] It should be noted that, and Source: Technical specifications or industry standards provided by the equipment manufacturer.
[0086] It should be noted that, The values are all greater than and less than , The setting process is data-driven. It involves collecting historical data from a large amount of similar equipment (such as distribution transformers of the same model), including sequences of various state parameters and final fault records. Then, through statistical analysis (such as calculating the correlation coefficient between each parameter and the fault occurrence time, or using algorithms such as random forests to calculate feature importance), the contribution of each parameter to the prediction of equipment health degradation is quantified. Finally, these contributions are normalized so that their sum is 1, which yields the final weight. For example, if analysis reveals that the correlation between insulating oil chromatographic data and transformer faults is 1.5 times that of winding temperature, then the former is weighted 1.5 times that of the latter.
[0087] It should be noted that the setting of the health assessment index threshold is based on the process of statistical analysis of historical fault data. The specific steps are as follows: extract a large amount of historical health index data of similar equipment and their corresponding fault records from the management system, statistically analyze the health index distribution of all equipment that eventually developed into faults within a period of time (such as 24 hours) before the fault occurred, and calculate its statistical quantile (the calculation method is such as using the 10th percentile). This quantile is determined as the health assessment index threshold.
[0088] For example, if analysis reveals that 90% of the faulty devices had a health index below 0.62 before the failure, the initial health assessment index threshold is set to 0.6. Finally, this health assessment index threshold is fine-tuned by domain experts based on their operational experience to form the final health assessment index threshold.
[0089] It should be noted that the setting of the predicted health assessment index threshold is based on a quantitative process driven by historical model performance data. Specifically, the predicted health assessment index is first backtested by comparing historical predicted values with actual values to calculate the prediction error distribution. For example, if the test finds that 90% of the error predicted 7 days in advance falls within ±0.05, then a safety margin is determined based on this error distribution. For example, an upper limit of 0.05 is taken as the minimum protection boundary. This margin is then deducted from the current health assessment threshold. If the current threshold is set to 0.6, then the predicted threshold is calculated to be 0.6 - 0.05 = 0.55.
[0090] In a specific embodiment, the health management of the Class II general maintenance equipment is carried out as follows: based on the set Class II data collection time point, key parameters of each Class II equipment to be managed are collected to obtain various key parameters of each Class II equipment to be managed.
[0091] By combining the weighting factors corresponding to the key parameters of each managed device belonging to Category II, the device status score of each managed device belonging to Category II is obtained through weighted calculation.
[0092] Obtain the conservative basic maintenance cycle for each piece of equipment to be managed in Category II, combine it with the equipment status score of each piece of equipment to be managed in Category II, and the set adjustment coefficient, and calculate using the following formula: The distance to the next maintenance is calculated. Maintenance interval for each device under management , These are the device numbers for each piece of equipment under management that belongs to Category II. The value of is a positive integer. and For the first The conservative basic maintenance cycle and equipment status score of each piece of equipment to be managed. This is the set adjustment coefficient.
[0093] When the The runtime of each managed device and The sum ≥ , or the When the equipment status score of a device to be managed is less than or equal to the set threshold for Class II equipment status scores, then the first device will be generated. The maintenance work order corresponding to the equipment to be managed.
[0094] It should be noted that key parameters include, but are not limited to, contact point temperature, number of protection actions, mechanical characteristic data (such as opening and closing time), and visual inspection results (such as corrosion level).
[0095] It should be noted that the first The runtime of a managed device refers to the time from the first... The cumulative running time of a device under management since its last maintenance.
[0096] It should be noted that the weight factors for each key parameter of the managed equipment belonging to Category II are all greater than 0 and less than 1, and their setting process is the same as... The setting process is the same, so it will not be described in detail here. The weighted calculation process is as follows: after normalizing each key parameter by the maximum-minimum method, multiply it by its corresponding weight factor, and sum the weighted results of all parameters. This weighted fusion method is a mature technology in the field of data mining, so it will not be described in detail here.
[0097] It should be noted that the adjustment coefficient The acquisition process, based on existing strategy mapping and parameter fixing methods, is used in this solution to dynamically adapt operation and maintenance strategies (such as extending equipment service life) to the specific equipment maintenance interval formula. The specific values are preset and fixed in the database by analyzing the correlation between historical maintenance data and equipment status, combined with expert experience; for example, when the operation and maintenance strategy is set to moderately extend the maintenance cycle to save resources, then... The value is 0.1, which makes the formula... The item undergoes minor adjustments, thus postponing the next maintenance. In this scheme, the value range is from 0 to 2. When the value is 0, the maintenance cycle will depend entirely on the conservative base value; while when When the value is 0.2, the equipment status score is allowed to positively adjust the cycle by up to 20%.
[0098] In a specific embodiment, the health management of the Class III routine maintenance equipment is carried out as follows: Based on the set Class III data collection time point, the communication status of the managed equipment belonging to Class III is collected. When the communication status of a managed equipment is interrupted and the interruption duration exceeds the set time threshold for distinguishing between instantaneous communication fluctuations and actual equipment failures, it is determined that the fault judgment condition is met and a fault warning prompt is triggered.
[0099] After a fault warning is triggered, the number of affected users in the distribution network corresponding to the area to be evaluated is obtained. The number of affected users is compared with the threshold range of the number of affected users corresponding to each response level. If the number of affected users is within the threshold range of the number of affected users corresponding to a certain response level, it indicates that the response level of the device to be managed is the response level corresponding to the threshold range of the number of affected users. Based on the response level of each device to be managed, a corresponding maintenance work order is generated.
[0100] It should be noted that the setting of the data collection time points for Class I, Class II, and Class III is based on the aforementioned equipment classification (Class I, Class II, and Class III) and their dynamic sensing levels, and is preset with the principle of achieving optimal allocation of sensing resources. For Class I key maintenance equipment that requires high-density monitoring, a short data collection interval is set (e.g., every 15 minutes); for Class II ordinary maintenance equipment that receives ordinary monitoring, a medium data collection interval is set (e.g., once a day); and for Class III routine maintenance equipment that only needs basic status understanding, a long data collection interval is set (e.g., once a week).
[0101] It should be noted that the process of setting the duration threshold for distinguishing between instantaneous communication fluctuations and actual equipment failures is based on a retrospective analysis of all communication interruption events in historical operation and maintenance data. The analysis statistically determines the duration of temporary interruptions caused by network lag and signal interference, taking the upper limit of the statistical data (such as the 95th percentile). For example, if the analysis finds that 95% of instantaneous communication fluctuations in a certain distribution network area recover automatically within 5 minutes, while interruptions exceeding 5 minutes are caused by actual failures such as equipment power failure and communication module hardware damage, then the duration threshold is set to 5 minutes.
[0102] In the equipment health management process, this invention implements distinctly different management strategies for different categories of equipment. For Category I equipment, real-time status parameter collection, health index calculation, and future trend prediction are performed. For Category II equipment, maintenance cycles are dynamically adjusted based on key parameter scores. For Category III equipment, communication status is primarily monitored, and responses are initiated after fault timeouts. This approach facilitates an upgrade from preventative maintenance to predictive and status-based maintenance, and helps to significantly optimize the operation and maintenance costs of ordinary equipment while ensuring extremely high reliability of critical equipment.
[0103] The collaborative maintenance module is used to perform collaborative health maintenance on each device under management based on the health management results of each device.
[0104] In a specific embodiment, the collaborative health maintenance of each managed device is carried out as follows: Based on the generated Type I, Type II, and Type III maintenance work orders, the corresponding maintenance execution time points are obtained. When the maintenance execution time points corresponding to two or three types of maintenance work orders are on the same working day or within a preset time window, the various types of maintenance work orders are grouped into a candidate collaborative task set. Then, the candidate collaborative task set is evaluated to see if it meets the distance and resource requirements. If it does, the candidate collaborative task set is jointly maintained; otherwise, joint maintenance is not performed.
[0105] It should be noted that the maintenance execution time point is directly read from the specified data fields of the generated maintenance work orders. For Type I maintenance work orders, the maintenance execution time point is taken from the predictive maintenance execution time field in the corresponding maintenance work order attributes; for Type II maintenance work orders, the maintenance execution time point is taken from the planned maintenance execution time field in the corresponding maintenance work order attributes; and for Type III maintenance work orders, the maintenance execution time point is taken from the planned execution time field in the corresponding maintenance work order attributes.
[0106] It should be noted that the preset time window is a fixed duration directly set based on statistical analysis of historical operation and maintenance data in this region. The specific process is as follows: query the database for all successfully executed candidate collaborative task sets in the past year, calculate the total time spent from the start of the first maintenance work order to the end of the last maintenance work order, and then take a typical time (for example, 80% of the maintenance tasks in the candidate collaborative task set are completed within 4 hours) and set this time length directly as the corresponding time window.
[0107] It should be noted that the distance requirement refers to the geographical location of the equipment corresponding to each work order in the candidate collaborative task set being within a preset effective radius. The effective radius is calculated based on the speed of the vehicle and the working time, and is usually set at 5 kilometers to meet the requirement that the transfer time does not exceed 30% of the total working time. Resource requirements include: human resource requirements, which means that the types of work required for the collaborative task must be within the skill range of the executing team; and equipment resource requirements, which means that the key equipment required for the task (such as insulated bucket trucks and testing instruments) meets the sharing requirements and meets the needs of all work orders. For example, when the distance between the equipment to be managed in the three types of maintenance work orders is within 3 kilometers and all require the same type of professional testing equipment, it is determined that the distance and resource requirements are met.
[0108] Please see Figure 2 As shown, an intelligent management method for equipment information in a power distribution network includes the following steps:
[0109] S1. Obtain the management equipment in the distribution network corresponding to the management area, evaluate the static impact coefficient of each management equipment on the distribution network during a fault, and classify the dynamic perception level of each management equipment.
[0110] S2. Based on the static impact coefficient and dynamic perception level of each device to be managed, classify the devices to be managed.
[0111] S3. Based on the management category classification results of each device to be managed, perform health management on each device to be managed in each management category.
[0112] S4. Based on the health management results of each device to be managed, conduct collaborative health maintenance on each device to be managed.
[0113] In the collaborative maintenance decision-making process, this invention integrates various maintenance work orders according to their execution time points and evaluates whether they meet the requirements for sharing geographical distance, human resources, and equipment resources within a preset time window. This helps to optimize originally isolated maintenance tasks into efficient collaborative work groups, minimizes repeated trips by the maintenance team and repeated start-ups and shutdowns of equipment, thereby improving overall maintenance efficiency and reducing the scheduling costs of manpower and materials.
[0114] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.
Claims
1. An intelligent management system for equipment information in a power distribution network, characterized in that, include: The equipment attribute analysis module is used to obtain the management equipment in the distribution network corresponding to the management area, evaluate the static impact coefficient of each management equipment on the distribution network when it fails, and classify the dynamic perception level of each management equipment. The specific process for evaluating the static impact coefficient of each managed device on the distribution network during a fault is as follows: Obtain the quantity, type, and topology of each device to be managed, as well as the total number of users in the distribution network of the area to be managed; Set the importance level coefficient of the load supplied by each device to be managed, and obtain the number of users affected in the corresponding distribution network when each device to be managed fails and stops; Based on the total number of users in the distribution network under management, the importance level coefficient of the load supplied by each device under management, and the number of users affected by the corresponding distribution network when each device under management stops due to a fault, calculate the static impact coefficient of each device under management on the distribution network during a fault. The specific process for dynamically classifying the sensing levels of each device to be managed is as follows: Dynamic perception levels are divided into full perception level, basic perception level and no perception level. The data update frequency of each data monitoring point of each device to be managed is obtained, and the dynamic perception level of each device to be managed is divided by combining the preset fast data update frequency threshold and slow data update frequency threshold. Based on the divided devices to be managed, corresponding level thresholds are set for the full perception level, basic perception level and no perception level of each device to be managed, which are respectively denoted as the full perception level threshold, basic perception level threshold and no perception level threshold. The process for setting the data update fast frequency threshold, data update slow frequency threshold, full perception level threshold, basic perception level threshold, and no perception level threshold is as follows: Construct a threshold dataset and construct threshold datasets for each group. Based on the threshold datasets, dynamically classify the perception levels of each device to be managed. Obtain the historical over-maintenance costs, historical false alarm handling costs, and historical failure loss costs for each device to be managed; By combining the average historical over-maintenance cost, the average historical false alarm handling cost, and the average historical failure loss cost, as well as the set over-maintenance weight, false alarm weight, and failure loss weight, the comprehensive loss value corresponding to each group of threshold datasets is calculated. Based on the comprehensive loss value corresponding to each group of threshold datasets, the values of fast data update frequency threshold, slow data update frequency threshold, full perception level threshold, basic perception level threshold and no perception level threshold are obtained by filtering. The management category delineation module is used to classify each device to be managed based on its static impact coefficient and dynamic perception level. The equipment health management module is used to perform health management on each piece of equipment in each management category according to the classification results of the management categories corresponding to each piece of equipment to be managed; The collaborative maintenance module is used to perform collaborative health maintenance on each device under management based on the health management results of each device.
2. The system of claim 1, wherein, The specific process for dividing the devices to be managed is as follows: By combining the static impact coefficient and dynamic perception level of each device to be managed, the comprehensive criticality index corresponding to each device to be managed is calculated. Based on the comprehensive criticality index corresponding to each device to be managed, the devices to be managed are divided into categories. 3.The power distribution network equipment information intelligent management system of claim 2, wherein, The specific process for performing health management on each device under each management category is as follows: Health management of all equipment under management includes health management of Class I key maintenance equipment, health management of Class II general maintenance equipment, and health management of Class III routine maintenance equipment; Based on the set data collection time points of Category I, the status parameters of each managed device belonging to Category I are collected, and the normalized values of each status parameter of each managed device are obtained through function analysis. By combining the normalized values of the status parameters of each device under management and the corresponding weighting factors, the health assessment index of each device under management is calculated. By combining the health assessment index corresponding to each device to be managed, and the health assessment index of each device to be managed at the current collection time and the historical collection time, the predicted health assessment index of each device to be managed at the future collection time is calculated. Health management is implemented for Class I key maintenance equipment by combining health assessment index thresholds and predicted health assessment index thresholds.
4. The system of claim 3, wherein, The specific process for health management of the Class II general maintenance equipment is as follows: Based on the set data collection time points for Category II, key parameters are collected for each managed device belonging to Category II. Based on the weighting factors corresponding to the key parameters of each managed device belonging to Category II, calculate the equipment status score of each managed device belonging to Category II. Obtain the conservative basic maintenance cycle of each managed device belonging to Category II, combine the device status score of each managed device belonging to Category II, and the set adjustment coefficient to calculate the maintenance time interval between each managed device and the next maintenance. Based on the runtime of each device to be managed, the maintenance interval before the next maintenance, the conservative basic maintenance cycle, and the Class II device status scoring threshold, health management is performed on Class II ordinary maintenance devices.
5. The intelligent management system for equipment information in a power distribution network according to claim 4, characterized in that, The specific process for health management of the Class III routine maintenance equipment is as follows: Based on the set Class III data collection time points, the communication status of each managed device belonging to Class III is collected; When the communication status of a managed device is interrupted and the interruption duration exceeds the set time threshold for distinguishing between instantaneous communication fluctuations and actual device failures, the fault determination condition is met, and a fault warning is triggered. After triggering a fault warning, the number of affected users in the distribution network corresponding to the area to be evaluated is obtained. Based on the threshold range of the number of affected users corresponding to each response level, health management of Class III routine maintenance equipment is performed.
6. The system of claim 5, wherein, The specific process for performing collaborative health maintenance on each device to be managed is as follows: Based on maintenance work orders of type I, type II and type III, the corresponding maintenance execution time points are obtained. When the maintenance execution time points corresponding to two or three types of maintenance work orders are on the same working day or within a pre-set time window, the various types of maintenance work orders are grouped into a candidate collaborative task set, and then collaborative health maintenance is performed on the candidate collaborative task set.
7. A device information intelligent management method of a device information intelligent management system of a power distribution network according to any one of claims 1 to 6, characterized by, Includes the following steps: S1. Obtain the management equipment in the distribution network corresponding to the management area, evaluate the static impact coefficient of each management equipment on the distribution network during a fault, and classify the dynamic sensing level of each management equipment. S2. Divide the devices to be managed according to their static impact coefficient and dynamic perception level; S3. Based on the management category classification results of each device to be managed, perform health management on each device to be managed in each management category; S4. Based on the health management results of each device to be managed, conduct collaborative health maintenance on each device to be managed.