Power grid operation analysis method and device, computer equipment, medium and program product

Abstract

The invention relates to a power grid operation analysis method and device, computer equipment, a medium and a program product, and relates to the field of power grids. Comprising the following steps: acquiring respective power fluctuation data of each power device in a target power grid and resource state information of the target power grid; based on the power fluctuation data, performing power fluctuation analysis on the power equipment to obtain a power fluctuation level of the power equipment; determining a data acquisition strategy matched with the power fluctuation level; based on the data acquisition strategy, acquiring operation data of the power equipment; performing resource allocation on the operation data according to the resource state information to obtain data analysis resources corresponding to the operation data; and according to each data analysis resource and each operation data, performing operation state analysis on the target power grid to obtain an operation state analysis result of the target power grid. The method can improve the accuracy of power grid operation analysis.

Description

Technical Field

[0001] This application relates to the field of power grids, and in particular to a power grid operation analysis method, apparatus, computer equipment, computer-readable storage medium, and computer program product. Background Technology

[0002] In modern energy systems, power system stability analysis is a crucial research area. With the widespread application of distributed energy resources and smart devices, how to quickly respond to demand and ensure power supply reliability in complex and ever-changing operating environments has become a major challenge that the power industry urgently needs to overcome.

[0003] However, current research and practice are difficult to fully adapt to the operating characteristics of different types of power equipment and lack refined consideration of resource allocation. As a result, in practical applications, the analysis results often fail to accurately reflect the true state of the power system, thus affecting the scientific nature of decision-making. Summary of the Invention

[0004] Therefore, it is necessary to provide a power grid operation analysis method, device, computer equipment, computer-readable storage medium, and computer program product that can improve the accuracy of power grid operation analysis in response to the above-mentioned technical problems.

[0005] Firstly, this application provides a power grid operation analysis method, including:

[0006] Acquire power fluctuation data for each power device in the target power grid, as well as resource status information of the target power grid;

[0007] Based on power fluctuation data, power fluctuation analysis is performed on power equipment to obtain the power fluctuation level of the power equipment;

[0008] Determine a data acquisition strategy that matches the level of power fluctuations;

[0009] Based on the data acquisition strategy, collect operational data of power equipment;

[0010] Based on the resource status information, resources are allocated to the running data to obtain the data analysis resources corresponding to the running data;

[0011] Based on various data analysis resources and operational data, the operational status of the target power grid is analyzed, and the operational status analysis results of the target power grid are obtained.

[0012] In one embodiment, power fluctuation analysis is performed on power equipment based on power fluctuation data to obtain the power fluctuation level of the power equipment, including:

[0013] The power fluctuation data is segmented to obtain multiple data fragments;

[0014] Analyze the power fluctuation characteristics of each data segment;

[0015] Based on the characteristics of each power fluctuation, the data segments are clustered to obtain multiple power fluctuation categories;

[0016] The power fluctuation level of power equipment is determined based on each power fluctuation category.

[0017] In one embodiment, the operational data includes multiple types of sub-operational data; based on resource status information, resources are allocated to the operational data to obtain the data analysis resources corresponding to the operational data, including:

[0018] When the resource status information characterizes the target power grid resources as limited, we analyze the importance of each sub-operational data to the power grid operation status analysis.

[0019] The data from each sub-operation are prioritized according to their importance.

[0020] Based on priority, resources are allocated to each sub-run data in turn to obtain the data analysis resources corresponding to each sub-run data.

[0021] In one embodiment, based on various data analysis resources and operational data, an operational status analysis of the target power grid is performed to obtain the operational status analysis results of the target power grid, including:

[0022] Based on various data analysis resources, the various operational data are integrated to obtain the power grid operation dataset;

[0023] The power grid operation dataset is segmented to obtain multiple power grid operation segments;

[0024] The operational status of each power grid segment is analyzed to obtain the operational stability of each segment.

[0025] Based on various operational stability parameters, a state distribution analysis is performed on the target power grid to obtain the operational state distribution of the target power grid.

[0026] Based on the distribution of operating states, the results of the operating state analysis of the target power grid are determined.

[0027] In one embodiment, the operational status analysis results include the flexibility score of the target power grid and the operational strategy; based on the operational status distribution, the operational status analysis results of the target power grid are determined, including:

[0028] Based on the distribution of operating states, a flexibility analysis is performed on the target power grid to obtain a flexibility score for the target power grid.

[0029] When the flexibility score is lower than the score threshold, the power grid operating parameters that affect the power grid flexibility are adjusted to obtain the adjusted power grid operating parameters.

[0030] Based on the adjusted power grid operating parameters, the operating strategy of the target power grid is determined.

[0031] In one embodiment, based on a data acquisition strategy, operational data of the power equipment is collected, including:

[0032] Based on the data acquisition strategy, raw operating data of power equipment is collected;

[0033] If the original operating data contains abnormal operating data, the abnormal operating data is isolated to obtain the normal operating data.

[0034] Extract power operation characteristic data to characterize the equipment's operating status from normal operation data;

[0035] The power operation characteristic data is compressed to obtain the operation data of the power equipment.

[0036] Secondly, this application also provides a power grid operation analysis device, comprising:

[0037] The information acquisition module is used to acquire power fluctuation data of each power device in the target power grid, as well as resource status information of the target power grid.

[0038] The power fluctuation analysis module is used to perform power fluctuation analysis on power equipment based on power fluctuation data, and to obtain the power fluctuation level of the power equipment.

[0039] The data acquisition strategy determination module is used to determine a data acquisition strategy that matches the power fluctuation level.

[0040] The data acquisition module is used to collect operating data of power equipment based on data acquisition strategies.

[0041] The resource allocation module is used to allocate resources to the running data based on the resource status information, so as to obtain the data analysis resources corresponding to the running data;

[0042] The power grid operation analysis module is used to analyze the operation status of the target power grid based on various data analysis resources and operation data, and obtain the operation status analysis results of the target power grid.

[0043] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0044] Acquire power fluctuation data for each power device in the target power grid, as well as resource status information of the target power grid;

[0045] Based on power fluctuation data, power fluctuation analysis is performed on power equipment to obtain the power fluctuation level of the power equipment;

[0046] Determine a data acquisition strategy that matches the level of power fluctuations;

[0047] Based on the data acquisition strategy, collect operational data of power equipment;

[0048] Based on the resource status information, resources are allocated to the running data to obtain the data analysis resources corresponding to the running data;

[0049] Based on various data analysis resources and operational data, the operational status of the target power grid is analyzed, and the operational status analysis results of the target power grid are obtained.

[0050] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0051] Acquire power fluctuation data for each power device in the target power grid, as well as resource status information of the target power grid;

[0052] Based on power fluctuation data, power fluctuation analysis is performed on power equipment to obtain the power fluctuation level of the power equipment;

[0053] Determine a data acquisition strategy that matches the level of power fluctuations;

[0054] Based on the data acquisition strategy, collect operational data of power equipment;

[0055] Based on the resource status information, resources are allocated to the running data to obtain the data analysis resources corresponding to the running data;

[0056] Based on various data analysis resources and operational data, the operational status of the target power grid is analyzed, and the operational status analysis results of the target power grid are obtained.

[0057] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0058] Acquire power fluctuation data for each power device in the target power grid, as well as resource status information of the target power grid;

[0059] Based on power fluctuation data, power fluctuation analysis is performed on power equipment to obtain the power fluctuation level of the power equipment;

[0060] Determine a data acquisition strategy that matches the level of power fluctuations;

[0061] Based on the data acquisition strategy, collect operational data of power equipment;

[0062] Based on the resource status information, resources are allocated to the running data to obtain the data analysis resources corresponding to the running data;

[0063] Based on various data analysis resources and operational data, the operational status of the target power grid is analyzed, and the operational status analysis results of the target power grid are obtained.

[0064] The aforementioned power grid operation analysis method, apparatus, computer equipment, computer-readable storage medium, and computer program product first acquire the power fluctuation data of each power device in the target power grid, as well as the resource status information of the target power grid. Based on the power fluctuation data, power fluctuation analysis is performed on the power devices to obtain their power fluctuation levels, and a data acquisition strategy matching these levels is determined. Based on this strategy, operational data of the power devices is collected. Resource allocation is performed on the operational data according to the resource status information to obtain the corresponding data analysis resources. Finally, based on each data analysis resource and each operational data point, the operational status of the target power grid is analyzed to obtain the operational status analysis results. Thus, this solution adaptively determines and executes differentiated data acquisition strategies by analyzing the real-time fluctuation characteristics of power devices, ensuring data validity and economic efficiency from the source. Furthermore, by intelligently allocating and scheduling the collected operational data based on the global resource status, efficient execution of the analysis task is ensured under limited resource constraints. It effectively resolves the contradiction between the blindness of grid data collection and the scarcity of computing resources in the context of a high proportion of distributed energy access to the grid, thereby improving the accuracy and timeliness of power system operation status analysis, providing more precise decision support for grid dispatch and operation, and significantly enhancing the flexibility of the power system in responding to fluctuations and the overall operational stability. Attached Figure Description

[0065] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0066] Figure 1 This is a diagram illustrating the application environment of a power grid operation analysis method in one embodiment.

[0067] Figure 2This is a flowchart illustrating a power grid operation analysis method in one embodiment;

[0068] Figure 3 This is a schematic diagram of the process for determining the power fluctuation level in one embodiment;

[0069] Figure 4 This is a schematic diagram of the resource allocation process in one embodiment;

[0070] Figure 5 This is a flowchart illustrating the power grid operation status analysis in one embodiment;

[0071] Figure 6 This is a flowchart illustrating the power grid system flexibility analysis in one embodiment;

[0072] Figure 7 This is a schematic diagram of the data processing flow in one embodiment;

[0073] Figure 8 This is a structural block diagram of a power grid operation analysis device in one embodiment;

[0074] Figure 9 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0075] In modern energy systems, power system operation analysis, such as flexibility analysis, is a crucial research area, directly impacting the safe and stable operation of the power grid and the efficient utilization of new energy sources. With the widespread application of distributed energy resources and smart power equipment, how to quickly respond to demand and ensure power supply reliability in complex and ever-changing operating environments has become a major challenge that the power industry urgently needs to overcome. Power grid operation analysis is not only key to optimizing resource allocation but also an important support for promoting energy transformation.

[0076] However, current research and practice struggle to fully adapt to the operating characteristics of different types of power equipment, particularly in the dynamic balancing of data acquisition and processing, where a lack of refined consideration of resource allocation is evident. This results in analytical results that often fail to accurately reflect the true state of the system in practical applications, thus impacting the scientific validity of decision-making.

[0077] A deeper technical challenge lies in determining the appropriate data acquisition frequency and the efficiency of data processing at the edge. Different devices and scenarios have vastly different data acquisition needs. For example, some new energy devices experience extremely rapid changes in their operating status, requiring very high-frequency data recording to capture their fluctuations. Other devices operate relatively smoothly, and excessively high acquisition frequencies would waste resources. This difference directly leads to an imbalance in data processing pressure, resulting in bottlenecks in data transmission and computation allocation between the edge and central sides of the system. Taking new energy power generation equipment as an example, in situations with frequent fluctuations within a short period, if sufficiently high-frequency data cannot be acquired in a timely manner, its actual contribution to grid flexibility cannot be accurately assessed. Excessive data, on the other hand, increases the burden on transmission and storage, creating a dilemma.

[0078] Therefore, this application proposes a power grid operation analysis method to reasonably determine the data acquisition strategy of different power equipment and reasonably allocate the corresponding analysis resources, which effectively improves the accuracy and efficiency of power grid operation analysis, thereby ensuring the stable operation of the power system in complex environments.

[0079] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0080] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0081] The power grid operation analysis method provided in this application embodiment can be applied to, for example... Figure 1The power system shown includes at least one edge side 102, a central side 104, and a resource scheduling side 106. The edge side 102 is deployed at each power device in the target power grid and is used to perform data sensing and preprocessing for each power device. Specifically, it includes: collecting real-time operating data of the power devices; analyzing the real-time operating data to obtain the power fluctuation characteristics of the power devices and determining the power fluctuation level of the power devices through cluster analysis; determining an appropriate data acquisition strategy based on the power fluctuation level and executing subsequent data acquisition according to the strategy; filtering, detecting and isolating anomalies, extracting key features, and compressing and encoding the collected operating data to obtain processed operating data; and transmitting the processed operating data to the central side 104 based on the resources allocated by the resource scheduling side 106. The resource scheduling side 106 is communicatively connected to both the edge side 102 and the central side 104, and is specifically used to monitor the resource status information of the power system and dynamically allocate resources to the operating data based on the resource status information. The central side 104 and the edge side 102 are communicatively connected to perform system-level power grid operation status analysis and decision-making. Specifically, this includes receiving operation data transmitted from each edge side 102, performing operation status analysis on the target power grid based on the data, and obtaining the operation status analysis results. Through the collaboration of the edge side 102, the central side 104, and the resource scheduling side 106, this system achieves a closed loop from local data acquisition to global situation analysis, improving data processing efficiency while ensuring the accuracy and real-time performance of power grid operation status analysis.

[0082] In this embodiment, the edge side 102 can be, but is not limited to, at least one of the following: a smart controller built into the power equipment, a sensor, a locally installed edge gateway / server, or a regional edge computing node. The central side 104 can be a cloud data center deployed far from the site, typically a server cluster with data analysis capabilities. The resource scheduling side 106 can be the main dispatch center of the power system, used for resource scheduling.

[0083] In one exemplary embodiment, such as Figure 2 As shown, a power grid operation analysis method is provided, which can be applied to... Figure 1 Taking the power system in China as an example, the following steps are included:

[0084] Step S202: Obtain the power fluctuation data of each power device in the target power grid and the resource status information of the target power grid.

[0085] The target power grid refers to a complete power network entity comprising various power equipment, transmission networks, distribution networks, and loads. Power equipment refers to the power generation equipment in the target power grid that requires monitoring and analysis, such as at least one of thermal power generation equipment, hydropower generation equipment, wind power generation equipment, and photovoltaic power generation equipment. In this embodiment, particular attention is paid to new energy equipment with intermittent and random output, such as wind power generation equipment. Power fluctuation data refers to data reflecting the continuous fluctuations in the output of power equipment over time, and may include, but is not limited to, at least one of output power, voltage, and current. Resource status information refers to relevant information reflecting the resource status of the target power grid, and may include, but is not limited to, at least one of communication resource status and computing resource status. Communication resource status may include, for example, at least one of the available network bandwidth, current network latency, and packet loss rate at the central data receiving end, which can be obtained in real time through network monitoring tools such as iperf. Computing resource status may include, for example, at least one of the CPU (Central Processing Unit) utilization, memory usage, and storage I / O performance of the central analysis server, which can be obtained through system monitoring tools such as Prometheus.

[0086] For example, this embodiment serves as the starting point for data awareness and state awareness. The edge side acquires power fluctuation data from each power device at a basic frequency, such as 10 times per second, and packages it into timestamped data packets. Simultaneously, the central side continuously collects its own communication and computing resource indicators. For instance, taking wind power generation equipment as an example, the edge-side device sensors collect the wind turbine's output power data in real time, forming a raw record dataset containing timestamps and power values.

[0087] Step S204: Based on power fluctuation data, perform power fluctuation analysis on the power equipment to obtain the power fluctuation level of the power equipment.

[0088] Among them, the power fluctuation level is an indicator reflecting the degree of fluctuation in the output of power equipment. For example, there are a limited number of levels such as Level 1, Level 2, and Level 3. The higher the level, the more severe the fluctuation and the greater the impact on the operation of the power grid. Of course, in practical applications, it is not limited to using numerical form to represent the power fluctuation level. Other representation forms such as letters or a combination of letters and numbers can also be used. This embodiment does not impose any restrictions on this.

[0089] For example, after acquiring power fluctuation data, the edge computing platform can first standardize and segment the data, dividing it into multiple continuous data segments. Then, for each data segment, the core fluctuation features are extracted to form a feature vector describing the fluctuation pattern of that segment. Based on this, an unsupervised clustering algorithm is used to automatically analyze the feature vector set of all segments, classifying the segments into different fluctuation categories according to feature similarity. Finally, according to preset business rules, each fluctuation category obtained from the clustering analysis is mapped to a fluctuation level with clear physical meaning and ordered relationships. By comprehensively considering the level distribution of the equipment across all time periods, the overall power fluctuation level of the power equipment in the current monitoring period is ultimately determined.

[0090] In some embodiments, the edge side can also preset corresponding fluctuation threshold ranges for power fluctuation data, thereby determining the corresponding fluctuation level based on the fluctuation threshold range. For example, fluctuation amplitudes less than 100 kW are set as Level 1, 100 kW to 300 kW as Level 2, and greater than 300 kW as Level 3. This method is simple to implement, transparent in decision-making, and has low computational overhead. Alternatively, at least one classification model such as support vector machine or random forest can be used to distinguish power fluctuation levels.

[0091] Step S206: Determine a data acquisition strategy that matches the power fluctuation level.

[0092] Among them, the data acquisition strategy is the set of execution instructions or parameter configuration that guides the data acquisition behavior, which may include, but is not limited to, at least one of the following: data acquisition frequency, sampling accuracy, triggering conditions, and data encapsulation format.

[0093] For example, the edge side pre-defines a mapping table between power fluctuation levels and data acquisition strategies. This table defines the data acquisition strategy parameters corresponding to different fluctuation levels. For instance, fluctuation level 3 corresponds to continuous high-frequency data acquisition, fluctuation level 2 corresponds to normal-frequency data acquisition, and fluctuation level 1 corresponds to low-frequency data acquisition. Once the edge side obtains the power fluctuation level of the power equipment, it can query this mapping table to retrieve and lock the corresponding data acquisition strategy. It can be understood that a higher fluctuation level means that the equipment state is undergoing rapid and drastic changes, containing a large amount of critical information and potentially indicating risks; therefore, a high-frequency acquisition strategy should be used. Conversely, a low fluctuation level means that the equipment is operating smoothly, its state changes slowly, and the information increment is limited. In this case, a low-frequency acquisition strategy is sufficient to track its trend. If high-frequency acquisition is still used, a large amount of redundant data will be generated, resulting in a waste of storage, computing, and communication resources. Through this dynamic matching mechanism, adaptive adjustment of data acquisition is achieved.

[0094] For example, when the fluctuation level is high, such as level 3, the frequency adjustment mechanism will be triggered, and the current data acquisition frequency will be switched to a higher range according to the preset mapping table, such as from 10Hz to 20Hz.

[0095] In some embodiments, by acquiring data at high frequency, the edge side can capture more refined power fluctuation data, which can be used to provide feedback to optimize the fluctuation level classification model itself, such as by retraining using a support vector machine algorithm to further refine the levels.

[0096] Step S208: Collect operating data of power equipment based on the data acquisition strategy.

[0097] The operational data refers to the structured dataset reflecting the electrical state and behavior of power grid equipment, collected according to the data acquisition strategy. It should be noted that acquiring power fluctuation data in step S202 is an initial monitoring action, the purpose of which is to collect raw state signals for analyzing the fluctuation characteristics of the equipment itself. This data has a basic frequency, is not optimized by the strategy, and is mainly used to determine the fluctuation level. In contrast, acquiring operational data in step S208 is a goal-oriented adaptive acquisition action performed after the fluctuation level has been determined and a specific acquisition strategy has been matched. This data is a structured, lightweight dataset based on dynamically adjusted frequencies and preprocessed in real time at the edge, specifically designed to provide high-quality input for subsequent resource allocation and system-level analysis.

[0098] For example, after determining the data acquisition strategy, the edge device can dynamically configure the sampling parameters of its local data acquisition hardware or driver software based on this strategy, switching sampling parameters such as the data acquisition frequency from the original frequency to the target frequency. After configuration, synchronous sampling for a period of time is initiated according to the new parameters, continuously acquiring raw electrical signals such as voltage and current of the power grid equipment to form an initial sampling dataset. Subsequently, the initial sampling dataset is preprocessed, such as removing redundant data, repairing outliers using interpolation methods, suppressing noise interference using sliding window smoothing techniques, or removing obviously invalid values ​​according to simple rules. The specific preprocessing operations can be determined according to the actual situation, thereby obtaining higher quality and more compact operating data.

[0099] For example, lightweight data acquisition software deployed at the edge can be used to achieve real-time sampling of power data. Assuming an adjusted sampling frequency of 10 times per second in the power equipment of a substation, the software will continuously record data such as power, voltage, and current of the power equipment at this frequency, forming an initial sampling dataset. For the filtering mechanism of the initial sampling dataset, a statistical rule-based method can be used to identify redundant samples. For instance, if the voltage value is found to be almost constant for some time periods in a day's sampling data, with a variation of less than 0.5 volts, these data can be considered redundant and discarded, thereby reducing data storage pressure and improving the efficiency of subsequent processing. When detecting anomalies, time series analysis methods can be used to focus on abrupt changes in the data. For example, if the voltage value suddenly jumps from 220 volts to 260 volts in a certain segment of initial sampling data, exceeding the normal range, interpolation methods can be used to repair this abnormal data point based on the average value of the preceding and following time points, ensuring data continuity and providing a reliable foundation for subsequent analysis. To address noise interference, the sliding window smoothing technique can effectively reduce random fluctuations in the data. Assuming the window size is set to 5 sampling points, the data is smoothed by calculating the average value within each window. This can effectively weaken subtle fluctuations caused by device interference, making the data trend clearer and facilitating feature extraction.

[0100] In some embodiments, power fluctuation characteristics for key time periods can be further extracted from the preprocessed operational data. If the fluctuation amplitude exceeds a preset threshold, it is marked as a high-risk interval, thus identifying potential risk interval data. The potential risk interval data is logically compared with locally stored historical anomaly records to analyze whether their changing trends are similar to known fault modes, thereby determining whether the power equipment is operating abnormally. For abnormal operation situations, at least one feature characterizing the abnormal operation, such as fluctuation amplitude, fluctuation frequency, and duration, is combined into a power data feature set and input into a pre-trained classification model, such as a support vector machine, for analysis to obtain the classification results of the power equipment's operating status, such as bearing wear, instantaneous grid disturbances, or at least one other category.

[0101] For example, when extracting power fluctuation characteristics during key time periods, peak and trough data can be focused on. If a voltage fluctuation reaches 10 volts within a certain time period, exceeding a preset threshold of 5 volts, it is marked as a high-risk area. By logically comparing historical anomaly records, the periodicity of the fluctuation trend can be analyzed. If the comparison reveals that the fluctuation trend of a certain high-risk area is similar to the anomaly records of the past week, exhibiting a sudden voltage drop every 2 hours, it can be identified as a potential equipment aging problem. This analysis helps to identify equipment anomalies in advance.

[0102] Step S210: Based on the resource status information, allocate resources to the running data to obtain the data analysis resources corresponding to the running data.

[0103] Resource allocation refers to the process by which the central server dynamically allocates appropriate resources to the running data based on the current resource status and the characteristics of the running data. Data analysis resources refer to virtualized resources that are explicitly bound to the running data after allocation, and may include, but are not limited to, at least one of bandwidth resources, memory resources, and computing resources.

[0104] For example, this embodiment is executed on the resource scheduling side of the power system and is a key control link for achieving efficient and stable analysis. Its implementation is a dynamic, closed-loop scheduling process: the resource scheduling side continuously monitors the resource status information of the target power grid. When it detects that multiple edge-side operational data are waiting to be transmitted, it initiates real-time assessment and scheduling. When resources are sufficient, a default or balanced allocation strategy is adopted; when resource constraints are detected, such as CPU utilization exceeding 85% or network bandwidth usage exceeding 90%, a dynamic priority scheduling mechanism is triggered. This mechanism sorts the queued operational data according to preset rules, such as prioritizing data with high volatility or prioritizing critical operational data, and allocates dedicated computing queues, guaranteed memory areas, and high-bandwidth transmission channels to high-priority tasks. Simultaneously, low-priority tasks can be delayed or batch-processed. Finally, the resource scheduling side generates and binds a unique resource token to each operational data packet, which encodes the specific resource quota allocated to it to guide its transmission behavior.

[0105] During data transmission, the resource scheduling side can monitor resource status in real time, such as tracking bandwidth fluctuations. If bandwidth fluctuations exceed a preset range, the transmission order of the operational data is dynamically adjusted. When data transmission is complete and the central side receives the compressed operational data, the post-processing stage begins: the central side first decompresses the received compressed data packets to initially recover the data, and then uses verification tools to perform integrity checks. If data is found to be missing or corrupted, a retransmission mechanism is automatically triggered until the final complete device operational data is obtained. This complete dataset will be stored in the target database, and the operational details of the entire transmission process can be recorded by log tools for auditing and analysis.

[0106] For example, suppose the resource scheduling monitoring detects that the currently available transmission bandwidth is 10 Mbps (megabits per second), while the preset bandwidth requirement threshold is 15 Mbps. This situation is judged as resource-constrained. Based on preset dynamic scheduling rules, a priority resource allocation mechanism is triggered. The core logic of this mechanism is: under bandwidth-constrained conditions, priority is given to ensuring the transmission resources of core characteristic data that is crucial to power grid status analysis. For example, suppose a running dataset contains 1000 data packets, of which 200 are related to core parameters. During resource allocation, bandwidth resources will be allocated first to these 200 data packets, while the remaining data will be queued according to secondary priority. This ensures that critical data packets can be delivered to the central side preferentially and reliably, thereby maximizing the timeliness and completeness of the data required for subsequent analysis under the constraint of overall resource constraints.

[0107] When performing hierarchical data transmission tasks, the configuration of priority queues is particularly important. Assuming data is divided into high, medium, and low priority levels using priority queues, high-priority data will be transmitted first when bandwidth is limited. For example, in a transmission task, if high-priority data accounts for 30%, even with bandwidth fluctuations, this portion of data can be guaranteed to be transmitted in the shortest possible time.

[0108] Network status can be monitored in real time when tracking bandwidth fluctuations. For example, if the bandwidth suddenly drops from 10Mbps to 5Mbps during transmission, exceeding the preset fluctuation range of 2Mbps, the resource scheduling side will automatically adjust the transmission task order, transmitting high-priority data in advance.

[0109] When performing data integrity checks at the central side, if a data packet's checksum is found to be inconsistent with the original value, indicating potential data corruption, the central side will request data retransmission from the edge side to reacquire the data packet and ensure data integrity.

[0110] The final complete dataset can be stored in the target database in timestamp order for easy querying and retrieval later. Simultaneously, a transmission operation log is generated, recording, for example, the transmission start time as 9:00 AM, the end time as 9:30 AM, and the total transmitted data volume as 2GB (gigabytes). This facilitates subsequent tracing and analysis of problems during the transmission process.

[0111] Step S212: Based on the data analysis resources and the operational data, perform an operational status analysis on the target power grid to obtain the operational status analysis results of the target power grid.

[0112] Operational status analysis refers to the process of evaluating the overall performance and stability of the target power grid; in this embodiment, it specifically refers to the quantitative assessment of power grid flexibility. The operational status analysis results are the outputs of the above-mentioned operational status analysis process, and may include, but are not limited to, at least one of the following: power grid flexibility score, risk level identification, and operational strategy.

[0113] For example, the edge side efficiently transmits the processed operational data to the central side based on dynamically allocated data transmission resources, such as bandwidth resources. After fully receiving and aggregating the operational data from multiple edge sides, the central side initiates grid operation status analysis. Specifically, the central side first divides the integrated operational data into time windows, forming multiple grid operation segments representing different operating conditions. Subsequently, the central side utilizes allocated computing resources, such as memory resources, to perform in-depth analysis of each grid operation segment in parallel, calculating at least one key indicator such as operational stability and regulation capability. It then synthesizes the results of all segments to statistically obtain an operational status distribution reflecting the overall behavioral characteristics of the grid. Based on this operational status distribution, the central side invokes a built-in flexibility measurement model, which maps multi-dimensional status indicators to a comprehensive flexibility score, thereby providing a complete quantitative assessment of the overall flexibility of the power system.

[0114] In this embodiment, the power fluctuation data of each power device in the target power grid and the resource status information of the target power grid are first acquired. Power fluctuation analysis is performed on the power devices based on the power fluctuation data to obtain the power fluctuation level of the power devices, and a data acquisition strategy matching the power fluctuation level is determined. Based on this data acquisition strategy, the operating data of the power devices is collected. Resource allocation is performed on the operating data according to the resource status information to obtain the data analysis resources corresponding to the operating data. Finally, based on each data analysis resource and each operating data, the operating status analysis of the target power grid is performed to obtain the operating status analysis results of the target power grid. Thus, this embodiment adaptively determines and executes differentiated data acquisition strategies by analyzing the real-time fluctuation characteristics of power devices, ensuring the validity and economic efficiency of data acquisition from the source. Furthermore, by intelligently allocating and scheduling the collected operating data based on the global resource status, the efficient execution of the analysis task under limited resource constraints is ensured. It effectively resolves the contradiction between the blindness of grid data collection and the scarcity of computing resources in the context of a high proportion of distributed energy access to the grid, thereby improving the accuracy and timeliness of power system operation status analysis, providing more precise decision support for grid dispatch and operation, and significantly enhancing the flexibility of the power system in responding to fluctuations and the overall operational stability.

[0115] In one exemplary embodiment, such as Figure 3 As shown, based on power fluctuation data, power fluctuation analysis is performed on power equipment to obtain the power fluctuation level of the power equipment, including:

[0116] Step S302: Segment the power fluctuation data to obtain multiple data segments.

[0117] In this context, a data segment can also be understood as a segment of equipment operating status. It refers to a finite-length subsequence obtained by cutting continuous power fluctuation data into fixed or adaptive time windows. Each segment represents the operating status of the equipment within a specific short period of time.

[0118] For example, after acquiring the power fluctuation data of the equipment, the edge side can use a fixed-duration sliding window method to segment the power fluctuation data. That is, a fixed duration, such as 1 hour, is set as the window length, and the window slides continuously on the time axis with a certain step size, such as a sliding interval of 1 minute. Each time, all sampling points within the window are captured to form a data segment. It can be understood that the output of power equipment, especially new energy, is prone to non-stationarity due to changes in environment and operating conditions over a long time scale. By segmenting the data, the non-stationary long sequence can be transformed into a series of locally stationary short sequences, thereby significantly reducing the complexity of subsequent analysis. Each data segment, as an independent sample, can more accurately capture the fluctuation pattern of the equipment within a micro-period, such as sudden rise, sudden fall, or stability, providing standardized input for subsequent calculation of the fluctuation characteristics within the segment.

[0119] For example, after obtaining the raw record dataset of wind power generation equipment, the edge computing side can further segment the dataset according to the time series, such as dividing the output power data of a day into 24 segments per hour, so as to analyze the performance of the equipment in different time periods in more detail.

[0120] Step S304: Analyze the power fluctuation characteristics of each data segment.

[0121] Among them, power fluctuation characteristics refer to a set of quantitative characteristic indicators used to quantitatively describe the degree of change in the output of a device within a data segment, such as fluctuation amplitude, fluctuation standard deviation, rate of change, etc.

[0122] For example, for each data segment, the edge device can first extract its output, such as the time series value of power, and then calculate the fluctuation characteristics of the series, such as fluctuation amplitude, fluctuation standard deviation, rate of change, etc. Fluctuation amplitude is the difference between the maximum and minimum power values ​​within the segment. This physical quantity directly reflects the total range of output variation of the equipment during that period and is the most intuitive indicator of the severity of fluctuations. Simultaneously, to more comprehensively describe the fluctuations, the fluctuation standard deviation of the segment can be calculated in parallel to characterize the degree of dispersion of the power value around the average value. Subsequently, the edge device compares the calculated fluctuation amplitude with a preset fluctuation amplitude threshold, which can be comprehensively set based on equipment type, historical operating data, and grid safety operation requirements. If the fluctuation amplitude exceeds this threshold, the segment is marked as an abnormal fluctuation segment; otherwise, it is marked as a normal fluctuation segment. It can be understood that in power system operation, excessively large power fluctuations in a short period are usually manifestations of equipment failure, strong external interference, or extreme operating conditions, and are abnormal events that need to be identified and monitored. By performing this calculation and labeling on all segments, the original time-series data stream is transformed into a structured segment feature set consisting of fluctuation characteristic values ​​and state labels. This elevates continuous numerical data into a feature space representation containing semantic information, providing a clear and effective input feature vector for subsequent clustering analysis.

[0123] For example, if the power value of a device in a certain data segment frequently changes between 500 kW and 800 kW, with a change range of 300 kW, while the preset threshold is 200 kW, then the segment will be marked as an abnormal segment.

[0124] Step S306: Based on the characteristics of each power fluctuation, cluster each data segment to obtain multiple power fluctuation categories.

[0125] Among them, power fluctuation categories, also known as data clusters, refer to several groups automatically formed after analyzing the feature vectors of all data segments using clustering algorithms. Each category contains data segments with similar fluctuation characteristics.

[0126] For example, after obtaining the power fluctuation characteristics of each data segment, these characteristics can be converted into feature vectors and used as input for a clustering algorithm. In this embodiment, the clustering algorithm can be K-means clustering. In practical applications, other clustering algorithms can be selected according to specific needs, and this embodiment does not impose any restrictions on this. During the clustering process, K points can be randomly selected as initial cluster centers, and the value of K can be set according to the actual situation. Then, the Euclidean distance from each feature vector to each cluster center is calculated, and the vector is assigned to the cluster containing the nearest center. Subsequently, the mean of all feature vectors in each cluster is recalculated as the new cluster center. The above steps are repeated until the cluster centers no longer change significantly or the maximum number of iterations is reached, at which point the algorithm converges. Finally, the algorithm can output at least one piece of information, such as all power fluctuation categories, the number of power fluctuation categories K, the category label of each data segment, and the coordinates of the cluster center of each category.

[0127] For example, when using the K-means algorithm to group data segments, power fluctuations can be categorized into three types based on their amplitude: low fluctuation, medium fluctuation, and high fluctuation. If the fluctuation amplitude of one type of data is concentrated below 100 kilowatts, it is classified as low fluctuation, while another type is concentrated above 300 kilowatts and can be classified as high fluctuation.

[0128] Step S308: Determine the power fluctuation level of the power equipment according to each power fluctuation category.

[0129] Among them, the power fluctuation level refers to the ordered classification label with clear physical meaning assigned to each power fluctuation category. For example, level 1 corresponds to low fluctuation, level 2 corresponds to medium fluctuation, and level 3 corresponds to high fluctuation.

[0130] For example, after obtaining multiple power fluctuation categories, the edge computing side can assign a corresponding fluctuation level to each category according to a preset level mapping rule. The level mapping rule refers to a pre-defined business logic rule that maps power fluctuation categories to specific fluctuation levels, such as low fluctuation corresponding to level 1, medium fluctuation corresponding to level 2, and high fluctuation corresponding to level 3. Next, the edge computing side can perform a comprehensive equipment level determination stage. This involves determining the overall power fluctuation level of the equipment based on the category of all data segments belonging to the equipment during the analysis period, using the limit principle. For example, if the equipment has any data segment at level 3 during a certain period, the equipment is determined to be at level 3 during that period. Other methods can also be used, such as using the level with the most occurrences as the final power fluctuation level of the equipment; this embodiment does not impose any limitations on this. Finally, the edge computing side compares and verifies the power fluctuation level with the preset level standard. If an inconsistency is found, such as high fluctuation being mapped to level 2, a level adjustment mechanism is triggered. This involves adjusting at least one parameter of the clustering algorithm, such as re-initializing cluster centers or increasing the number of categories K, and re-executing clustering and mapping until the power fluctuation level matches the preset level standard. By mapping power fluctuation levels, maintenance personnel can quickly locate problems and take measures, significantly reducing the risk of equipment failure and improving operational efficiency and safety.

[0131] In this embodiment, continuous power fluctuation data is segmented into fragments and features are extracted. Then, a clustering algorithm is used to automatically identify different fluctuation pattern categories, and finally, the fluctuation level of the equipment is determined based on the category features. This achieves accurate identification of the operating status of power equipment, transforming raw data into clear and business-meaning level labels. This provides accurate and reliable decision-making basis for subsequent adaptive adjustment of data collection frequency and optimization of resource allocation, thereby fundamentally improving the intelligence and efficiency of data collection and enhancing the system's ability to perceive and respond to power grid fluctuations.

[0132] In one exemplary embodiment, such as Figure 4 As shown, the operational data includes multiple types of sub-operational data; based on resource status information, resources are allocated to the operational data to obtain the corresponding data analysis resources, including:

[0133] Step S402: Under the condition that the target power grid resources are limited, the importance of each sub-operational data to the power grid operation status analysis is analyzed.

[0134] Sub-operational data refers to a subset of data further divided from the processed operational data according to data category or feature dimension. For example, the operational data of the same device can be divided into at least one subset based on parameter type, such as voltage data, current data, and power data. Importance is a quantitative indicator used to measure the information value or criticality level contributed by a specific sub-operational data to completing the final power grid operation status analysis task.

[0135] For example, the resource scheduling side first parses the metadata accompanying each sub-run data. This metadata may contain at least one piece of information such as the fluctuation level of its source device, parameter type, and timestamp. It then invokes a pre-defined importance assessment rule base, built based on domain knowledge, which defines the criticality of different attribute data for state analysis. For instance, the rule might specify that voltage data from a device with fluctuation level 3 has the highest importance, while power data has a relatively high importance. Based on these rules, the resource scheduling side can calculate or match an importance score or level label for each sub-run data. With sufficient resources, all data can be processed equally. However, under resource constraints, data must be differentiated based on its contribution to the final analysis objective to ensure that limited resources are prioritized for processing the data with the highest information value, thereby maximizing the accuracy of the final analysis results within the constraints.

[0136] Step S404: Prioritize each sub-run data according to its importance.

[0137] Priority sorting refers to the process of arranging each sub-run data according to its assessed importance, forming a task queue from highest to lowest priority.

[0138] For example, all sub-process data to be processed are arranged according to preset sorting rules, such as prioritizing data with higher importance. For data of equal importance, secondary sorting can be performed using auxiliary rules, such as data generation time or data packet size. For instance, data with timestamps more recent than the current time are given priority, or smaller data packets are given priority, to quickly free up resources. Finally, an ordered priority task list is generated.

[0139] Step S406: Based on priority, allocate resources to each sub-run data in sequence to obtain the data analysis resources corresponding to each sub-run data.

[0140] For example, the resource scheduler allocates resources sequentially to each sub-run data transmission task, starting from the top of the priority task list. For the highest priority task, sufficient or guaranteed resource quotas can be allocated from the global resource pool, such as allocating a dedicated CPU core, reserving a fixed amount of memory, and ensuring high-bandwidth I / O. Then, the allocated resources are deducted, and the remaining resources are allocated to the next priority task until the resource pool is exhausted or all tasks have received basic resources. For low-priority tasks at the end of the list that cannot be immediately allocated resources due to insufficient resources, the resource scheduler can mark them as pending, waiting for higher-priority tasks to complete and release resources before scheduling. Ultimately, each sub-run data task allocated resources receives a unique resource token, which encapsulates the specific resource information it can use; this is the data analysis resource corresponding to the data transmission task.

[0141] In this embodiment, when system resources are limited, the criticality of different operating data to state analysis is intelligently evaluated and distinguished, and priority is assigned and resources are allocated accordingly. This ensures that limited computing and communication resources are used preferentially and reliably to process the most valuable information.

[0142] In one exemplary embodiment, such as Figure 5 As shown, based on various data analysis resources and operational data, the operational status analysis of the target power grid is performed, and the operational status analysis results of the target power grid are obtained, including:

[0143] Step S502: Based on the various data analysis resources, integrate the various operational data to obtain the power grid operation dataset.

[0144] For example, the edge sides efficiently transmit the processed operational data to the central side based on dynamically allocated data transmission resources, such as bandwidth resources. After fully receiving and aggregating the operational data from multiple edge sides, the central side integrates this operational data to form a complete power grid operation dataset and initiates power grid operation status analysis.

[0145] Step S504: Segment the power grid operation dataset to obtain multiple power grid operation segments.

[0146] Among them, the power grid operation segment refers to the data subset obtained by dividing the power grid operation dataset into preset or adaptive time windows. Each segment can characterize the operation status of the power grid within a specific time period.

[0147] For example, after obtaining the power grid operation dataset, the central side can segment it according to a preset segmentation strategy. A common strategy is fixed-duration segmentation, such as hourly or daily segmentation. Another is adaptive segmentation based on events or state transitions, such as automatically defining the starting point of a new segment when a significant change is detected in the total load of the entire network or the penetration rate of new energy sources. The segmentation process needs to maintain the integrity of the data within each segment and the temporal continuity between segments. Segmentation can transform complex, non-stationary long sequences into a series of relatively stationary or state-consistent short sequences, which not only reduces the complexity of a single analysis but also facilitates subsequent comparative analysis of the power grid's performance under different states.

[0148] For example, by using timestamps, the power grid operation dataset can be segmented by hour or day to divide it into data subsets under different states such as peak period, off-peak period, and normal operation period, forming classified power operation segments.

[0149] Step S506: Perform operational status analysis on each power grid operation segment to obtain the operational stability of each power grid operation segment.

[0150] Operational stability is used to quantitatively describe the level of operational stability of the power grid within a specific segment of the power grid operation. It can be obtained based on a combination of multiple indicators, such as voltage fluctuation rate, power fluctuation intensity, and load rate of key equipment, among others.

[0151] For example, the central side can analyze at least one indicator within each power grid operation segment, such as voltage fluctuation rate, power fluctuation intensity, and critical equipment load rate. Each indicator reflects the stability of the power grid during that period. Finally, a stability index vector containing the values ​​of the above multiple indicators is output for each segment, which characterizes the operational stability of the power grid under the corresponding power grid operation segment.

[0152] For example, suppose that during peak periods, the voltage fluctuation rate reaches 5%, higher than the 2% during normal operation, indicating that the system stability decreases during peak periods. In practical applications, an operating status distribution diagram can also be plotted to visually demonstrate the differences in system stability at different time periods, providing data support for subsequent flexibility assessments.

[0153] Step S508: Based on the operational stability, perform state distribution analysis on the target power grid to obtain the operational state distribution of the target power grid.

[0154] State distribution analysis refers to the overall analysis of stability indices across all segments to obtain the statistical regularities, central tendency, and dispersion of the power grid's operating state. The operating state distribution is a comprehensive statistical description, including at least one of the following: the mean and variance of each stability index vector, the proportion of segments with different stability levels, and the correlation between stability indices.

[0155] For example, after obtaining the stability index vectors for all segments, descriptive statistics can be performed, i.e., calculating at least one of the mean, median, standard deviation, and extreme values ​​for each index across all segments. Subsequently, cluster analysis can be performed to automatically classify the segments according to their stability characteristics, thereby obtaining the distribution proportions of categories such as high stability, medium stability, and low stability, and finally generating a comprehensive operational status distribution report.

[0156] Step S510: Based on the operating status distribution, determine the operating status analysis results of the target power grid.

[0157] For example, after obtaining the operational status distribution, the central side can input it into a preset flexibility measurement model. The core of this model is a comprehensive flexibility evaluation function. It takes at least one key parameter from the distribution information, such as the proportion of high-fluctuation segments or the variance of stability indicators, as input. Through weighted calculation, fuzzy logic, or other algorithms, it calculates a single, easily understandable overall grid flexibility score, such as a percentage score. Simultaneously, the model combines information on weak links in the distribution, such as time periods or regions where low-stability segments frequently occur, to generate risk identification and grid operation strategy recommendations. Finally, the central side packages the flexibility score, risk level identification, and operation strategies into a structured operational status analysis result and outputs it.

[0158] In this embodiment, by integrating the operating data of multiple devices, a unified power grid operation dataset is constructed, and further segmented for refined status assessment. Finally, based on the statistical distribution of all segments, an overall power grid flexibility score is obtained, making the assessment results of the power grid operation status more systematic and accurate.

[0159] In one exemplary embodiment, such as Figure 6 As shown, the operational status analysis results include the flexibility score and operational strategy of the target power grid; based on the operational status distribution, the operational status analysis results of the target power grid are determined, including:

[0160] Step S602: Based on the operating state distribution, perform a flexibility analysis on the target power grid to obtain a flexibility score for the target power grid.

[0161] For example, after obtaining the operating status distribution, the central side can input it into a pre-trained flexibility measurement model. This model is typically a multi-index decision analysis model, which internally defines at least one core dimension constituting grid flexibility, such as ramp-up capability, regulation capacity, and response speed, as well as corresponding quantitative sub-indicators for each dimension. These sub-indicators can be directly extracted from the operating status distribution; for example, the proportion of high-fluctuation segments reflects the intensity of ramp-up demand. The model first normalizes each extracted sub-indicator to eliminate the influence of dimensions. Then, it performs weighted aggregation of the normalized sub-indicators according to the built-in weight system, and finally calculates a comprehensive flexibility score.

[0162] Step S604: If the flexibility score is lower than the score threshold, adjust the grid operation parameters that affect grid flexibility to obtain the adjusted grid operation parameters.

[0163] The scoring threshold refers to a preset critical score used to distinguish grid flexibility. If the flexibility score is below this threshold, the grid is considered to have low flexibility; conversely, if the flexibility score is not below this threshold, the grid is considered to have high flexibility. Grid operating parameters refer to dispatchable parameters in the power system that directly affect its operating state, such as at least one of the following: peak load, response time, generator output plan and reserve capacity, energy storage charging and discharging power, interruptible load reduction, and reactive power compensation device operation.

[0164] For example, when the flexibility score falls below a threshold, the central system will trace the source of the insufficient flexibility through the operational status distribution, identify the key operational parameters causing the inadequacy, and determine the associated set of schedulable parameters. Subsequently, a mathematical optimization problem is constructed with the objective of maximizing flexibility or making the flexibility score exceed the threshold. The decision variables for this problem are the aforementioned set of schedulable parameters, and the constraints include, but are not limited to, at least one of the following: power grid safety constraints such as power flow constraints or equipment operating limits, physical constraints, and market rules. An optimization algorithm, such as linear programming, mixed-integer programming, or a heuristic algorithm, is then used to solve the problem. Under the premise of satisfying all constraints, the algorithm iteratively searches for the optimal values ​​of the decision variables that optimize the objective function, and the resulting solution is the adjusted power grid operating parameters.

[0165] For example, assuming a flexibility score ranges from 0 to 100, a power system with a flexibility score of 65, below the preset threshold of 80, indicates low flexibility. In this case, key operating parameters such as peak load and response time are extracted to form a set of dispatchable parameters. When adjusting this set of dispatchable parameters, the correlation between parameters can be analyzed using linear regression. Assuming a positive correlation between peak load and response time, the peak load setting can be appropriately reduced, such as from 800 MW to 750 MW, resulting in adjusted grid operating parameters.

[0166] Step S606: Determine the operating strategy of the target power grid based on the adjusted power grid operating parameters.

[0167] Among them, the operating strategy refers to the operable instructions or plans set for executing the adjusted power grid operating parameters.

[0168] For example, after obtaining the adjusted grid operating parameters, the central side can convert them into specific strategy forms adapted to different execution objects and interface requirements. For instance, for generator sets, day-ahead or real-time generation plan curves are generated; for energy storage systems, charging and discharging power plans are generated. Before the strategy is officially issued, fast power flow calculations or simulations can be used to verify the safety of grid operation under the strategy, performing final verification. After the verification process, the generated strategy instructions can be encapsulated according to the specified communication protocol and data format, and metadata such as timestamps and version numbers can be added. The final output is a structured target grid operation strategy that can be issued immediately or at scheduled intervals.

[0169] In some embodiments, to verify the effectiveness and safety of the generated operating strategy, the central side can conduct simulation tests based on complete historical or real-time datasets. Specifically, the adjusted set of power grid operating parameters is used as input to drive the power grid model in a simulation environment, simulating a complete scheduling cycle, such as the system dynamics over the next 24 hours. By comparing key performance indicators before and after the strategy implementation, the optimization effect can be quantitatively evaluated. For example, simulation results show that during peak electricity consumption periods, the average voltage fluctuation rate of the system decreases from the original 5% to 3%, indicating that the new parameter configuration effectively enhances voltage support capability and the optimization effect is significant. By comprehensively analyzing the optimized system operating status of the simulation output, the system can scientifically confirm whether the final response parameter optimization result has achieved the expected goal, and provide an objective and reliable decision-making basis for whether to put this strategy into actual operation or store it as empirical knowledge in the strategy library.

[0170] In some embodiments, the central side can also synchronously update the power system operation strategy database, record the adjusted grid operation parameters and corresponding flexibility scores, provide reference data for future operation strategy adjustments, and ensure continuous system improvement.

[0171] In this embodiment, by establishing a complete decision-making closed loop that includes flexibility quantitative assessment, intelligent parameter optimization, simulation verification and confirmation, and generation of executable strategies, not only is accurate perception and quantitative evaluation of grid flexibility achieved, but also effective optimization schemes can be automatically generated and verified when deficiencies are found. Finally, safe and reliable strategies that can directly guide dispatching operations are output, thereby significantly enhancing the grid's adaptive capability and overall operational safety level in the face of new energy fluctuations and complex operating conditions.

[0172] In one exemplary embodiment, such as Figure 7 As shown, based on the data acquisition strategy, the operating data of the power equipment is collected, including:

[0173] Step S702: Based on the data acquisition strategy, collect the raw operating data of the power equipment.

[0174] The raw running data refers to the initial sampled dataset that has not been processed.

[0175] For example, after determining the data acquisition strategy, the edge device can dynamically configure the sampling parameters of the local data acquisition hardware or driver software based on the strategy, switching sampling parameters such as the data acquisition frequency from the original frequency to the target frequency. After configuration, synchronous sampling for a period of time is started according to the new parameters to continuously acquire the original electrical signals such as voltage and current of the power grid equipment. The resulting initial sampling dataset is the original operating data.

[0176] Step S704: If the original operating data contains abnormal operating data, isolate the abnormal operating data to obtain normal operating data.

[0177] Abnormal operating data refers to outlier data points or anomalous data segments in the original operating data that significantly deviate from the normal range of variation due to at least one reason, such as measurement noise, sensor failure, transient interference, or abnormal equipment operating conditions. Isolation refers to the process of identifying, marking, and physically separating abnormal operating data from the original operating data to ensure that abnormal operating data does not contaminate subsequent analysis. Normal operating data refers to data that conforms to the expected statistical characteristics and physical laws after isolating abnormal operating data.

[0178] For example, after obtaining the raw operational data, the edge computing side can first format and standardize it to unify timestamps and data formats, resulting in a pre-cleaned dataset. Then, the core anomaly detection and judgment stage begins, extracting key features such as values ​​and rates of change from the dataset and applying anomaly detection algorithms for judgment. A typical implementation uses the IsolationForest algorithm. After initializing the model, the dataset is fitted, and an anomaly score is calculated for each data point. This anomaly score is compared to a preset threshold; points not lower than the threshold are marked as abnormal operational data. Next, a data separation operation is performed, removing the marked anomaly data from the original operational data to generate an independent abnormal operational dataset, while the remaining portion constitutes the initial normal operational dataset.

[0179] In some embodiments, after obtaining the initial normal operation dataset, a data integrity check can be performed. This involves checking whether there are any missing or formatting issues in the normal data set due to the removal of abnormal operation data. If so, interpolation methods such as linear interpolation are applied to complete the data, ensuring data continuity and thus obtaining a complete normal operation dataset. To facilitate subsequent analysis, the edge side can also group this complete normal operation dataset by time period and store it in a database.

[0180] For example, in the processing of raw operational data, preliminary formatting and standardization are fundamental to ensuring data consistency. Suppose voltage data for various electrical devices in a substation are collected over a day. This data may contain different units or inconsistent timestamp formats. Through formatting and standardization, all timestamps can be standardized to a format such as "year-month-day; hour:minute:second," and voltage values ​​can be standardized to volts for easier subsequent analysis.

[0181] For anomaly detection and judgment, the IsolationForest algorithm can be used to effectively identify outliers in the original running data. The algorithm works by the fact that outlier data points often have significantly different characteristics from normal data points, making them easier to isolate quickly within a randomly partitioned tree structure. Taking a voltage data set containing 1000 consecutive sampling points as an example, most data points are within the normal range (e.g., 220V to 230V), but there may be a few outliers caused by interference or faults, whose values ​​deviate significantly from the normal range (e.g., above 300V or below 100V). The IsolationForest algorithm fits the data by constructing multiple randomly partitioned isolated trees. The shorter the path length required for each data point to be isolated by the tree structure, the easier it is to distinguish, and the lower the outlier score for that point. In typical output, the score for normal data points is usually close to 0, such as 0.1, while obviously outlier points receive lower scores, such as -0.5. The system presets an anomaly judgment threshold, such as -0.3; all data points with anomaly scores below this threshold are judged as abnormal running data and marked. The marked abnormal operation data will be extracted separately. For example, 20 abnormal points will be separated from 1,000 points, and the remaining 980 points will form a normal data set.

[0182] In the data integrity verification and interpolation completion stages, it is necessary to handle missing values ​​that may arise in the time series due to the removal of abnormal data. Linear interpolation is a commonly used and efficient data repair method. Its principle is based on the assumption that the change of physical quantities is approximately linear over short time intervals. For example, in a voltage data segment, data loss at five consecutive time points is caused by abnormal data removal. Given that the voltage value before the missing point is 220 volts and the voltage value after the missing point is 225 volts, the system can automatically calculate the voltage value that transitions smoothly between the two endpoints using the linear interpolation formula. Specifically, the five missing points will be filled with 221 volts, 222 volts, 223 volts, 224 volts, and 225 volts respectively. This ensures data continuity.

[0183] In the data grouping stage, grouping the complete normal data set based on time period attributes is a key step in achieving structured analysis of power load characteristics. This operation divides continuous time series data into multiple logical units according to fixed time windows, such as hourly intervals, enabling the system to identify and analyze periodic patterns and typical characteristics in power operation by time period. For example, after grouping the 24-hour operating data by hour, the system can clearly present the load differences in different time periods: data from 1:00 AM to 2:00 AM usually corresponds to lower power load, reflecting the characteristics of nighttime power consumption troughs; while data from 12:00 PM to 1:00 PM often shows higher load levels, reflecting the peak of daytime production and residential power consumption. In addition, each group of data is independently classified and stored, with each time period corresponding to a table or field. This not only facilitates querying but also provides data support for subsequent power operation status assessment.

[0184] Step S706: Extract power operation characteristic data from the normal operation data to characterize the operating status of the equipment.

[0185] Among them, power operation characteristic data refers to the core analytical parameters extracted from normal operation data that can characterize the operating status of equipment, such as at least one type of data such as voltage and current.

[0186] For example, after obtaining normal operation data, the edge computing side can further filter it, that is, extract initial operational feature data to characterize the equipment's operating status from the normal operation data, forming an initial dataset. This initial dataset is then classified to determine the range and category of core analytical parameters. Subsequently, the classified core analytical parameters are standardized to obtain a standardized parameter set containing the power operation feature data.

[0187] For example, when processing the initial dataset, core analytical parameters such as voltage and current can be divided into different ranges and categories according to preset rules. For instance, voltage parameters can be divided into low voltage, medium voltage, and high voltage based on their numerical range, while current parameters can be divided into light load and heavy load based on their load level. This classification provides a clearer understanding of the data distribution, laying the foundation for subsequent analysis and processing.

[0188] When standardizing the core analytical parameters after classification, parameters with different dimensions can be unified to the same scale. For example, assuming that the voltage value ranges from 200 volts to 400 volts and the current value ranges from 5 amps to 50 amps, standardization can map these values ​​to the interval between 0 and 1, resulting in a standardized set of parameters, which facilitates the comparison and comprehensive evaluation of different parameters in subsequent analyses.

[0189] Step S708: Compress the power operation characteristic data to obtain the operation data of the power equipment.

[0190] For example, after obtaining the power operation characteristic data, principal component analysis can be applied to reduce the dimensionality of the data, generating compressed intermediate data. Then, one-hot encoding is used to perform a structured transformation on the intermediate data to obtain the encoded feature matrix, which is the compressed power equipment operation data.

[0191] In some embodiments, after obtaining the encoded feature matrix, the matrix can be used for classification verification, such as distinguishing between the normal operating state and the abnormal operating state of a device. If the compressed feature matrix can distinguish the different states well, it indicates that the feature data extraction is successful and the key core parameters are preserved. If the classification effect is poor, it is necessary to go back and readjust the parameters for feature data extraction or compression.

[0192] For example, when applying principal component analysis (PCA) for dimensionality reduction, multiple related parameters can be compressed into a few principal components. Assuming the original data contains three dimensions—voltage, current, and power—dimensionality reduction can extract two principal components, retaining most of the information while reducing data complexity and generating an intermediate dataset. This helps improve the efficiency of subsequent processing.

[0193] When performing one-hot encoding on intermediate datasets, categorical variables can be converted into binary form. Assuming the data contains device types, such as device type 1 and device type 2, one-hot encoding converts them into two independent feature columns with values ​​of 0 or 1, resulting in the encoded feature matrix. This structured transformation facilitates subsequent algorithmic recognition and processing.

[0194] When applying the Support Vector Machine (SVM) algorithm for classification verification, the feature matrix can be input into a trained SVM model. The model analyzes the implicit patterns in the features and attempts to classify data records into predefined categories such as normal power generation and potentially abnormal power generation. If the classification results have high accuracy, it proves that the extracted features can effectively capture the essential differences between different operating states, thus indirectly verifying the success of the feature extraction and compression process. Furthermore, to ensure the reliability of the classification conclusions, a rationality check based on historical benchmarks can be introduced. That is, the potentially abnormal data groups classified by the above algorithm are compared with historically accumulated and verified normal power data sets using multi-dimensional parameters. For example, simultaneously checking whether the voltage and current continuously deviate from the historical normal range. If the comparison results are consistent, the confidence that the record is indeed abnormal is enhanced; if inconsistent, it suggests that the parameters of the feature extraction or classification model may need to be re-examined. This dual verification mechanism, which combines algorithm classification with rule validation, forms a closed-loop quality control process. It can not only screen out highly reliable analytical data, but its feedback results can also be used to continuously optimize feature extraction methods and model performance, thereby progressively improving the reliability and intelligence level of the entire data processing chain and providing a solid data foundation for the stable operation and accurate analysis of the power system.

[0195] In this embodiment, the reliability of input features is ensured by isolating abnormal data. Effective information condensation of the data is achieved by extracting core features characterizing the device status. Finally, compression processing significantly reduces the data volume while retaining key status information. This process not only improves data reliability but also significantly reduces network transmission load and central processing pressure, laying a high-quality data foundation for subsequent accurate analysis and efficient decision-making.

[0196] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0197] Based on the same inventive concept, this application also provides a power grid operation analysis device for implementing the power grid operation analysis method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more power grid operation analysis device embodiments provided below can be found in the limitations of the power grid operation analysis method described above, and will not be repeated here.

[0198] In one exemplary embodiment, such as Figure 8 As shown, a power grid operation analysis device is provided, comprising:

[0199] The information acquisition module 802 is used to acquire the power fluctuation data of each power device in the target power grid and the resource status information of the target power grid.

[0200] The power fluctuation analysis module 804 is used to perform power fluctuation analysis on power equipment based on power fluctuation data to obtain the power fluctuation level of the power equipment.

[0201] The data acquisition strategy determination module 806 is used to determine a data acquisition strategy that matches the power fluctuation level.

[0202] The data acquisition module 808 is used to collect operating data of power equipment based on a data acquisition strategy.

[0203] The resource allocation module 810 is used to allocate resources to the running data according to the resource status information, so as to obtain the data analysis resources corresponding to the running data;

[0204] The power grid operation analysis module 812 is used to perform operation status analysis on the target power grid based on various data analysis resources and operation data, and obtain the operation status analysis results of the target power grid.

[0205] In one embodiment, the power fluctuation analysis module 804 is further configured to:

[0206] The power fluctuation data is segmented to obtain multiple data fragments;

[0207] Analyze the power fluctuation characteristics of each data segment;

[0208] Based on the characteristics of each power fluctuation, the data segments are clustered to obtain multiple power fluctuation categories;

[0209] The power fluctuation level of power equipment is determined based on each power fluctuation category.

[0210] In one embodiment, the runtime data includes multiple types of sub-running data; the resource allocation module 810 is further used for:

[0211] When the resource status information characterizes the target power grid resources as limited, we analyze the importance of each sub-operational data to the power grid operation status analysis.

[0212] The data from each sub-operation are prioritized according to their importance.

[0213] Based on priority, resources are allocated to each sub-run data in turn to obtain the data analysis resources corresponding to each sub-run data.

[0214] In one embodiment, the power grid operation analysis module 812 is further configured to:

[0215] Based on various data analysis resources, the various operational data are integrated to obtain the power grid operation dataset;

[0216] The power grid operation dataset is segmented to obtain multiple power grid operation segments;

[0217] The operational status of each power grid segment is analyzed to obtain the operational stability of each segment.

[0218] Based on various operational stability parameters, a state distribution analysis is performed on the target power grid to obtain the operational state distribution of the target power grid.

[0219] Based on the distribution of operating states, the results of the operating state analysis of the target power grid are determined.

[0220] In one embodiment, the operational status analysis results include a flexibility score for the target power grid and an operational strategy; the power grid operation analysis module 812 is further used for:

[0221] Based on the distribution of operating states, a flexibility analysis is performed on the target power grid to obtain a flexibility score for the target power grid.

[0222] When the flexibility score is lower than the score threshold, the power grid operating parameters that affect the power grid flexibility are adjusted to obtain the adjusted power grid operating parameters.

[0223] Based on the adjusted power grid operating parameters, the operating strategy of the target power grid is determined.

[0224] In one embodiment, the data acquisition module 808 is further configured to:

[0225] Based on the data acquisition strategy, raw operating data of power equipment is collected;

[0226] If the original operating data contains abnormal operating data, the abnormal operating data is isolated to obtain the normal operating data.

[0227] Extract power operation characteristic data to characterize the operating status of equipment from normal operation data;

[0228] The power operation characteristic data is compressed to obtain the operation data of the power equipment.

[0229] Each module in the aforementioned power grid operation analysis device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0230] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores power grid operation analysis data. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a power grid operation analysis method.

[0231] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0232] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0233] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0234] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0235] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0236] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0237] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0238] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A power grid operation analysis method, characterized in that, The method includes: Acquire power fluctuation data for each power device in the target power grid, as well as resource status information of the target power grid; Based on the power fluctuation data, power fluctuation analysis is performed on the power equipment to obtain the power fluctuation level of the power equipment; Determine a data acquisition strategy that matches the power fluctuation level; Based on the data acquisition strategy, the operating data of the power equipment are collected; Based on the resource status information, resources are allocated to the running data to obtain the data analysis resources corresponding to the running data; Based on the data analysis resources and the operational data, the target power grid is subjected to operational status analysis to obtain the operational status analysis results of the target power grid.

2. The method according to claim 1, characterized in that, The step of performing power fluctuation analysis on the power equipment based on the power fluctuation data to obtain the power fluctuation level of the power equipment includes: The power fluctuation data is segmented to obtain multiple data fragments; Analyze the power fluctuation characteristics of each of the data segments; Based on the power fluctuation characteristics described above, the data segments are clustered to obtain multiple power fluctuation categories. The power fluctuation level of the power equipment is determined according to each of the power fluctuation categories.

3. The method according to claim 1, characterized in that, The operational data includes multiple types of sub-operational data; the process of allocating resources to the operational data based on the resource status information to obtain the data analysis resources corresponding to the operational data includes: When the resource status information indicates that the target power grid resources are limited, the importance of each of the sub-operational data to the power grid operation status analysis is analyzed. The sub-operation data are prioritized according to their respective importance. Based on the priority order, resources are allocated to each of the sub-running data in sequence to obtain the data analysis resources corresponding to each of the sub-running data.

4. The method according to claim 1, characterized in that, The step of performing an operational status analysis on the target power grid based on the data analysis resources and the operational data to obtain the operational status analysis results of the target power grid includes: Based on the data analysis resources described above, the operational data described above are integrated to obtain a power grid operation dataset; The power grid operation dataset is segmented to obtain multiple power grid operation segments; The operational status of each power grid operation segment is analyzed to obtain the operational stability of each power grid operation segment. Based on the aforementioned operational stability, a state distribution analysis is performed on the target power grid to obtain the operational state distribution of the target power grid; Based on the operational status distribution, the operational status analysis results of the target power grid are determined.

5. The method according to claim 4, characterized in that, The operational status analysis results include the flexibility score of the target power grid and the operational strategy; The determination of the operational status analysis results of the target power grid based on the operational status distribution includes: Based on the operating state distribution, a flexibility analysis is performed on the target power grid to obtain a flexibility score for the target power grid; If the flexibility score is lower than the score threshold, the power grid operating parameters that affect the power grid flexibility are adjusted to obtain the adjusted power grid operating parameters. Based on the adjusted power grid operating parameters, the operating strategy of the target power grid is determined.

6. The method according to claim 1, characterized in that, The process of collecting operational data of the power equipment based on the data acquisition strategy includes: Based on the data acquisition strategy, the raw operating data of the power equipment are collected; If the original operating data contains abnormal operating data, the abnormal operating data is isolated to obtain normal operating data; Extract power operation characteristic data to characterize the equipment's operating status from the normal operation data; The power operation characteristic data is compressed to obtain the operation data of the power equipment.

7. A power grid operation analysis device, characterized in that, The device includes: The information acquisition module is used to acquire the power fluctuation data of each power device in the target power grid, as well as the resource status information of the target power grid. The power fluctuation analysis module is used to perform power fluctuation analysis on the power equipment based on the power fluctuation data, and to obtain the power fluctuation level of the power equipment; A data acquisition strategy determination module is used to determine a data acquisition strategy that matches the power fluctuation level. The data acquisition module is used to collect the operating data of the power equipment based on the data acquisition strategy. The resource allocation module is used to allocate resources to the running data according to the resource status information, so as to obtain the data analysis resources corresponding to the running data; The power grid operation analysis module is used to perform operation status analysis on the target power grid based on the data analysis resources and the operation data, and obtain the operation status analysis results of the target power grid.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 6.

Images