An ac-dc power supply system for a substation

Abstract

The application discloses an AC / DC power supply system for a transformer substation, relates to the power supply regulation and control technical field, and aims to solve the problems of one-sided load awareness, solidified power supply distribution and passive regulation and control mechanism in the prior art. The application comprises a load situation awareness module, a power supply distribution setting module and a deviation closed-loop regulation and control module. A plurality of load parameters are synchronously collected, time sequence evolution curves are fused and generated, branch correlation coupling and fluctuation synchronization characteristics are analyzed, load importance grades are divided, and a dynamically updated full-station load situation dataset is constructed. Branch carrying capacity is calculated, a weight assignment library is constructed, power supply weights are differentially adjusted, and output parameters are dynamically set. Operation deviations are monitored, abnormal causes are traced, load trends are predicted, weights are hierarchically corrected, and power supply ratios are cyclically closed-loop optimized. The application realizes comprehensive load awareness, flexible power supply distribution and closed-loop adaptive regulation and control, effectively improves system operation stability and balance, and adapts to complex working conditions of the transformer substation.

Description

Technical Field

[0001] This invention belongs to the field of power system power regulation technology, specifically relating to an AC / DC power supply system for substations. Background Technology

[0002] The AC / DC power supply system of a substation is crucial for the safe operation of power. Currently, the industry generally adopts traditional monitoring and control schemes. Existing technologies suffer from drawbacks such as incomplete load sensing, unreasonable power distribution, and imperfect control mechanisms, making them difficult to adapt to complex operating conditions and limiting system stability and balance. Therefore, this invention urgently needs to solve the following technical problems: How to more comprehensively and dynamically perceive the overall load status, sort out the internal relationships between power supply units, form reliable global data support, and alleviate the problems of one-sided load perception and insufficient data support; How to formulate a flexible power supply allocation strategy by taking into account multiple factors, taking into account the differences between different units, reducing power supply imbalance, and improving the problems of rigid allocation and unreasonable resource allocation; To address the issues of passive error correction and poor system stability, we propose an AC / DC power supply system for substations. This system aims to improve load forecasting and troubleshooting, optimize adjustment methods, and establish closed-loop control. Summary of the Invention

[0003] The purpose of this invention is to provide an AC / DC power supply system for substations to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an AC / DC power supply system for a substation, comprising: Load Situation Awareness Module: Acquires AC / DC feeder branches for the entire station and synchronously collects load operating status; fuses and generates load time-series evolution curves, calculates correlation coupling and fluctuation synchronization levels; calculates the proportion of total load and branch contribution for the entire station and classifies them into importance levels; integrates multi-dimensional information to construct a load situation dataset for the entire station and updates it in real time; Power distribution setting module: calls the full station load status dataset, calculates the real-time remaining carrying capacity margin of branches and classifies the carrying capacity margin level; builds a weight assignment library, generates power supply priority weights, and normalizes them to obtain the global initial power supply allocation weights; adjusts the weights differently according to the collaborative, local and independent allocation candidate groups, normalizes them to determine the final power supply allocation weights; sets and distributes the active power setting parameters, and cyclically updates the power supply ratio parameters. Deviation closed-loop control module: continuously collects actual operating data of AC / DC feeder branches, calculates the relative deviation of the distribution ratio and determines the deviation level; conducts deviation root cause tracing and load trend prediction for branches with significant deviations, and gradually corrects the final power supply allocation weight in stages; re-sets the output active power setting parameters and sends them down for storage, forming a closed-loop adaptive control link in a loop, and dynamically optimizes the power supply ratio of each branch.

[0005] Preferably, the specific process for fusing and generating load time-series evolution curves and calculating the correlation coupling level is as follows: Acquire all AC and DC feeder branches of the entire station, set synchronous acquisition intervals, and synchronously acquire real-time active power, operating current and operating status indicators of each branch. Select the continuous time series analysis window, normalize the active power and operating current, and then weight and fuse them to generate the load time series evolution curves of each branch. Calculate the load time-series correlation coefficient between any two branches and classify them into three levels of coupling: strong coupling, weak coupling, and no coupling. Based on different levels of association and coupling, the corresponding branch combinations are marked as collaborative allocation candidate groups, local adaptation candidate groups, and independent allocation groups, respectively.

[0006] Preferably, the specific process for calculating the fluctuation synchronization level is as follows: For any two AC / DC feeder branches, the phase offset is calculated using the cross-correlation method within the same time series analysis window, based on the load time series evolution curves of the two branches. Three levels of fluctuation synchronization are set: synchronous, anti-synchronous, and asynchronous. Each level corresponds to a phase offset range that is non-overlapping and covers the entire range of values. The phase offset and level range of each matched branch are matched one by one, and the fluctuation synchronization level judgment result is output. According to the level, the corresponding branch combinations are marked as synchronous coordination group, complementary coordination group, and independent timing group respectively.

[0007] The preferred process for classifying importance levels, integrating multi-dimensional information to construct a full-site load status dataset, and updating it in real time is as follows: The system acquires the real-time active power and operating status indicators of each AC / DC feeder branch, calculates the overall real-time load of the entire station, and calculates the contribution ratio of each AC / DC feeder branch to the overall real-time load of the entire station. Two importance levels are set for the core load branch and the auxiliary load branch. Each level corresponds to a contribution percentage range that is non-overlapping and covers the entire range of values. Match the contribution percentage and level range for each branch, and output the branch importance level determination result; By integrating real-time load operation parameters, correlation coupling level, fluctuation synchronization level, and branch importance level, a full-station load status dataset is constructed. The dataset is stored in the system storage area and updated in real time according to the synchronous acquisition interval.

[0008] Preferably, the specific process for calculating the real-time remaining carrying capacity margin of branch lines and classifying the carrying capacity margin levels is as follows: After each update of the full station load status dataset, the full station load status dataset in the system storage area is called to extract the real-time active power, branch importance level determination results, correlation coupling level markers and fluctuation synchronization level markers of each AC / DC feeder branch. Obtain the rated carrying capacity of each AC / DC feeder branch and calculate the real-time remaining carrying capacity margin of the branch; Three carrying capacity margin levels are set: sufficient, moderate, and insufficient. Each level corresponds to a non-overlapping relative range of remaining carrying capacity margin that covers the entire range of values. The system matches the real-time remaining load margin and grade range for each branch, and outputs the load margin grade determination result for each AC / DC feeder branch.

[0009] Preferably, the specific process of constructing a weight assignment library, generating power supply priority weights, and normalizing them to obtain the global initial power supply allocation weights is as follows: Construct a basic weight assignment library for the importance level of branch roads and a revised weight assignment library for the carrying capacity margin level; In the basic weight assignment library, core load branches and auxiliary load branches each correspond to a unique basic weight coefficient; In the weight assignment library, ample margin, moderate margin, and insufficient margin each correspond to a unique weight adjustment coefficient. Match the importance level and basic weight assignment library and the carrying capacity margin level and modified weight assignment library of each branch, and output the corresponding basic weight coefficient and modified weight coefficient. By combining the two types of weighting coefficients, the power supply priority weight of each AC / DC feeder branch is calculated; The power supply priority weights of all normally operating branches are selected and then processed globally to obtain the global initial power supply allocation weights of each AC / DC feeder branch.

[0010] Preferably, the weights of candidate groups are adjusted differently according to coordination, localization, and independent allocation, and the final power supply allocation weights are determined by normalization; the output active power setting parameters are set and issued, and the power supply ratio parameters are updated cyclically. The specific process is as follows: For branch pairs marked as candidate groups for coordinated allocation, the global initial power supply allocation weight of AC / DC feeder branches is adjusted differently according to the fluctuation synchronization level. For branch pairs marked as local adaptation candidate groups, calculate the load change rate of AC / DC feeder branches. When in a local fluctuation section, slightly adjust the global initial power supply allocation weight according to the local cooperative fluctuation correlation coefficient level, and keep it unchanged in a fluctuation-free section. For branch pairs marked as independent allocation groups, the global initial power supply allocation weight is directly applied. After completing the differentiated adjustment, the adjusted weights of all normally operating branches are globally normalized to obtain the final power supply allocation weights. Based on the overall real-time load of the entire station and the rated carrying capacity of the AC / DC feeder branches, the active power setting parameters are set and limited, and then sent to the corresponding branch control execution unit. The system stores the final power supply allocation weight and output active power setting parameters, executes the corresponding process cyclically according to the synchronous acquisition interval, and updates the power supply ratio parameters of each AC / DC feeder branch in real time.

[0011] Preferably, the specific process of continuously collecting actual operating data of AC / DC feeder branches, calculating the relative deviation of the distribution ratio, and determining the deviation level is as follows: According to the set synchronous acquisition interval, continuously collect the actual active power and actual operating status indicators of each AC / DC feeder branch after the power supply parameters are executed; The output active power setting parameters and actual active power are obtained for each branch, and the relative deviation of the AC / DC feeder branch ratio is calculated. Two deviation levels are set: normal deviation and significant deviation. Each level corresponds to a non-overlapping range of absolute values ​​of relative deviation of the proportion, which covers the entire range of values. The absolute value of the relative deviation of the matching ratio and the deviation level range are used to output the deviation level determination result of the AC / DC feeder branch.

[0012] Preferably, the specific process of tracing the root causes of deviations and predicting load trends for branches with significant deviations, and then progressively adjusting the final power supply allocation weights in stages, is as follows: Once the AC / DC feeder branch is determined to have a significant deviation, the deviation root cause tracing and adaptive fine-tuning process is initiated: Traverse the branch pairs formed by the branch with significant deviation and the other normally operating branches, retrieve the correlation coupling level, fluctuation synchronization level and real-time remaining carrying capacity margin, and determine the root cause of the deviation as the change of branch linkage characteristics, the change of the branch's own carrying capacity conditions or the disturbance of the external power grid conditions. Extract the historical active power time series of branches with significant deviations, construct a linear regression load trend model, predict load trends, and identify potential overload risks or capacity idle risks; Based on the root cause of the deviation and the load trend, the deviation correction, root cause expansion correction, and trend advance correction are performed in sequence, and the final power supply allocation weight of the relevant branches is gradually corrected in stages. After fine-tuning the weights of branches with significant deviations, the weights are globally normalized to obtain the updated final power supply allocation weights.

[0013] Preferably, the specific process of resetting the output active power setting parameters and sending them down for storage, forming a closed-loop adaptive control link, and dynamically optimizing the power supply ratio of each branch is as follows: For branches with significant deviations, the output active power setting parameters are readjusted and limited according to the branch's rated carrying capacity, taking into account the updated final power distribution weight and the overall real-time load of the entire station. The updated output active power setting parameters are sent to the control execution unit of the corresponding AC / DC feeder branch, and the updated final power supply allocation weight, output active power setting parameters, deviation tracing results and trend prediction results are stored in the system storage area. The entire process is executed cyclically according to the set synchronous acquisition interval, forming a closed-loop adaptive control link to dynamically optimize the power supply ratio of each AC / DC feeder branch and ensure the stable and balanced operation of the AC / DC power supply system.

[0014] Compared with the prior art, the beneficial effects of the present invention are: (1) This AC / DC power supply system for substations synchronously collects various parameters such as active power and operating current of branches and integrates them to generate load time-series evolution curves, sorts out the correlation coupling and fluctuation synchronization characteristics between branches, calculates the load contribution ratio to classify importance levels, and constructs a multi-dimensional dynamically updated full-station load status dataset to make up for the shortcomings of traditional load perception being one-sided and lacking global cognition, and provides comprehensive and reliable data support for subsequent power supply regulation.

[0015] (2) This AC / DC power supply system for substations constructs a weighted assignment library that takes into account both the importance of the load and the carrying capacity, and adjusts the power supply weights differently according to the candidate groups of coordinated, local and independent allocation. It combines the load of the whole station and the rated carrying capacity of the branch to dynamically set and limit the output parameters, thereby changing the problem of the traditional fixed power supply allocation and ignoring the differences of the branches, optimizing the rationality of power supply resource allocation, and adapting to different branch operating conditions.

[0016] (3) This AC / DC power supply system for substations detects power supply ratio deviation in real time, investigates abnormal causes, and predicts load change trends. It adopts a graded and gradual approach to correct power supply weight and cyclically executes the entire process to form a closed-loop control link. This improves the shortcomings of traditional control, which is passive correction and prone to system fluctuations. It continuously optimizes the power supply ratio and effectively enhances the stability and balance of the AC / DC power supply system in long-term operation. Attached Figure Description

[0017] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1; Please see Figure 1This invention provides an AC / DC power supply system for a substation, comprising: Load Situation Awareness Module: Acquires AC / DC feeder branches for the entire station, synchronously collects load operating status; fuses and generates load time-series evolution curves, calculates correlation coupling and fluctuation synchronization levels; calculates the proportion of total station load and branch contribution, classifies importance levels; integrates multi-dimensional information to construct a station-wide load situation dataset and updates it in real time. The specific process is as follows: Obtain all AC / DC feeder branches in the entire station, and set the feeder branch index as i, satisfying i∈[1,n] where n represents the total scale of AC / DC feeder branches in the entire station; Set a synchronous acquisition interval to synchronously acquire the load operating status of all AC / DC feeder branches. The load operating status includes: the branch's real-time active power, operating current, and operating condition status indicator (the operating condition status indicator is represented in a normalized binary form, with a value of 1 corresponding to normal load operation and a value of 0 corresponding to abnormal branch shutdown, among which abnormal branches are not included in the subsequent correlation analysis, weight calculation, and power supply dispatch process).

[0020] Select a continuous time series analysis window and extract the load time series evolution curves of each branch within the window. The load time series evolution curves are multi-parameter fused load time series sequences of the branches, and their construction method is as follows: First, the real-time active power and operating current of the branch are normalized to eliminate dimensional differences, thus obtaining the normalized time series. , Then, the normalized values ​​at the same time point are weighted and fused to construct the load time-series evolution curve of the branch. Its time-by-time expression is: ; in, and These are preset weighting coefficients, and the sum of their values ​​is one.

[0021] For any two different branches, perform time-series correlation characteristic calculation to construct the load time-series correlation coefficient. The calculation formula is as follows: , in, Let be the load time-series correlation coefficient between branch i and branch j, used to characterize the degree of correlation between the load fluctuations of the two branches. Let i be the load time sequence of branch i within the time window. Let j be the load time sequence of branch j within the time window. This is a temporal covariance calculation used to characterize the coordinated changing trend of two time series curves. This is a time series variance calculation used to characterize the degree of fluctuation and dispersion of a single time series curve; Three levels of coupling are set: strong coupling, weak coupling and no coupling. Each level of coupling corresponds to a range of absolute values ​​of the load time series correlation coefficient. The ranges corresponding to each level of coupling do not overlap and cover the complete range of values ​​of the load time series correlation coefficient. For any two different branches among all AC and DC feeder branches in the substation, the load time-series correlation coefficients corresponding to the two branches are extracted one by one. The absolute value of the extracted load time-series correlation coefficients is matched with the absolute value range of the load time-series correlation coefficients corresponding to the three preset association coupling levels. Based on the matching results, the association coupling level determination result corresponding to the two branches is output.

[0022] Branch combinations that are determined to be strongly coupled are marked as collaborative allocation candidate groups, branch combinations that are determined to be weakly coupled are marked as local adaptation candidate groups, and branch combinations that are determined to be uncoupled are marked as independent allocation groups.

[0023] For any two different branch load time-series evolution curves, the phase offset is calculated using the cross-correlation method within the same time-series analysis window; Three levels of fluctuation synchronization are set: synchronous, anti-synchronous, and asynchronous. Each wave synchronization level is pre-defined to correspond to a phase offset range. The ranges corresponding to each wave synchronization level do not overlap and cover the complete value range of the phase offset. For any two different branches among all AC / DC feeder branches in the substation, extract the phase offset corresponding to the two branches one by one. The extracted phase offset is matched one by one with the phase offset intervals corresponding to the three preset wave synchronization levels. Based on the matching results, the wave synchronization level judgment results corresponding to the two branches are output. The branch combinations that are judged to be synchronized are marked as synchronized coordination groups, the branch combinations that are judged to be antisynchronized are marked as complementary coordination groups, and the branch combinations that are judged to be asynchronous are marked as independent timing groups.

[0024] Among them, the synchronous dispatch group is where the phase offset of the load time sequence evolution curve of the two branches is within the preset synchronous range, the load fluctuation direction and rate are basically consistent, and they enter the peak or valley period at the same time, which has the basis for synchronous load increase and decrease. The complementary dispatch group has two branches whose load time evolution curves have phase offsets within a preset anti-synchronization range, and the load fluctuations are completely opposite. When one branch enters the peak load period, the other branch just enters the valley load period, thus creating a natural load complementarity space. The independent timing group consists of two branches whose load timing evolution curves have phase offsets within a preset asynchronous range. The load fluctuations have no fixed phase correlation, and the fluctuation directions and rates are independent of each other, lacking the basis for coordinated or complementary allocation.

[0025] Real-time active power and operating status indicators of each AC / DC feeder branch of the substation are obtained, and the formula is used: To obtain the overall real-time load of the entire station ,in, Let be the real-time active power of the i-th AC / DC feeder branch at time t. The operating status indicator for the i-th AC / DC feeder branch at time t.

[0026] For each AC / DC feeder branch, use the formula: The contribution percentage of this branch to the total station load at time t is obtained. ; Two importance levels are set for branch lines: core load branch lines and auxiliary load branch lines. Each branch importance level is pre-defined to correspond to a contribution percentage range, and the ranges corresponding to each level do not overlap and cover the complete value range of contribution percentage. For all AC / DC feeder branches of the substation, extract the contribution percentage corresponding to each branch; match the extracted contribution percentage with the contribution percentage intervals corresponding to the two preset importance levels of the branches one by one; and output the importance level determination result corresponding to the branch based on the matching result.

[0027] The real-time load operation parameters of each branch, the correlation and coupling level determination results of any two branches, the fluctuation synchronization level determination results, and the importance level determination results of each branch are integrated to form a full-station load status dataset that carries the linkage characteristics of the branches. The full-station load status dataset is stored in the system storage area, and the above process is repeated according to the set synchronous acquisition interval to update the full-station load status dataset in real time.

[0028] It should be noted that the real-time active power, operating current and operating status indicators of each AC / DC feeder branch are collected simultaneously. Normalized binary form is used to distinguish between normally operating branches and abnormal branches that are out of service. Abnormal branches are not included in subsequent correlation analysis, weight calculation and power supply dispatching process, which reduces interference from invalid data, lays a reliable data foundation for subsequent related work and reduces the possibility of control deviation caused by invalid data.

[0029] The real-time active power and operating current of the branch are normalized to eliminate the dimensional differences between different parameters. Then, a multi-parameter load time-series evolution curve is generated by weighted fusion. Compared with single-parameter monitoring, it can more completely present the dynamic change law of the branch load and closely match the actual load operation status.

[0030] By calculating the load time-series correlation coefficient, three levels of correlation coupling are classified: strong correlation coupling, weak correlation coupling, and no correlation coupling. The phase offset is solved by using the cross-correlation method to distinguish three levels of fluctuation synchronicity: synchronous, anti-synchronous, and asynchronous. Different dispatch candidate groups are marked accordingly. This effectively captures the load fluctuation correlation and time-series synchronization characteristics between branches, providing a reference for subsequent differentiated power supply dispatch and improving the situation where the branch linkage relationship is ignored in traditional control.

[0031] The system calculates the overall real-time load of the entire substation and the load contribution ratio of each branch. Based on the contribution ratio, it classifies the branches into two importance levels: core load branches and auxiliary load branches. This aligns with the actual load importance of the substation and provides an objective reference for subsequent power supply priority weighting, which is conducive to ensuring the stability of power supply to the core load.

[0032] By integrating multi-dimensional information such as real-time load operation parameters, correlation coupling level, fluctuation synchronization level, and branch importance level, a full-station load status dataset is constructed and updated in real time. The data dimensions are relatively complete and dynamically updated, which can provide real-time data support for the weight calculation and parameter tuning of the power supply distribution setting module, as well as the deviation source tracing and trend prediction of the deviation closed-loop control module. This helps the AC / DC power supply system control to conform to the actual operating conditions and maintain the stable and balanced operation of the system.

[0033] The power supply allocation setting module: calls the entire station load status dataset, calculates the real-time remaining carrying capacity margin of branches and classifies the carrying capacity margin levels; constructs a weight assignment library, generates power supply priority weights, and normalizes them to obtain the global initial power supply allocation weights; adjusts the weights differently according to the collaborative, local, and independent allocation candidate groups, and normalizes them to determine the final power supply allocation weights; sets and distributes the active power setting parameters, and cyclically updates the power supply ratio parameters. The specific process is as follows: After each update of the full station load status dataset, the full station load status dataset in the system storage area is called to extract the real-time active power of each branch, the branch importance level judgment result, the correlation coupling level mark and the fluctuation synchronization level mark. For each AC / DC feeder branch, obtain the rated carrying capacity of the branch and construct the expression for the branch's real-time remaining carrying capacity margin: The remaining load margin of the AC / DC feeder branch at time t is obtained. ; in, This refers to the rated load capacity of the branch circuit.

[0034] Three load margin levels are set: ample margin, moderate margin, and insufficient margin. Each load margin level is pre-defined to correspond to a relative range of remaining load margin. The ranges corresponding to each level do not overlap and cover the complete value range of the remaining load margin. For each AC / DC feeder branch, extract the real-time remaining capacity margin corresponding to that branch; The extracted real-time remaining load margin is matched one by one with the intervals corresponding to the three preset load margin levels. Based on the matching results, the load margin level determination result corresponding to the branch is output.

[0035] Construct a basic weight assignment library for the importance level of branch roads and a modified weight assignment library for the carrying capacity margin level; in the basic weight assignment library for the importance level of branch roads, preset a unique basic weight coefficient for each core load branch road and auxiliary load branch road; In the load margin level correction weight assignment library, the three levels of sufficient margin, moderate margin, and insufficient margin each correspond to a unique correction weight coefficient; For each AC / DC feeder branch, extract the branch importance level, match it with the branch importance level basic weight assignment library, and output the basic weight coefficient corresponding to the branch. Simultaneously, the load margin level of the branch is extracted, matched with the load margin level correction weight assignment library, and the correction weight coefficient corresponding to the branch is output.

[0036] Based on the basic weighting coefficient and the corrected weighting coefficient of the AC / DC feeder branch, using the formula: The power supply priority weight of the AC / DC feeder branch at time t is obtained. ; in, These are the basic weighting coefficients for the AC / DC feeder branches at time t. This refers to the corrected weighting coefficient for the AC / DC feeder branch at time t. Obtain the power supply priority weights of all normally operating branches with the status flag "normal operation", and perform global normalization calculations to obtain the global initial power supply allocation weights for each AC / DC feeder branch. The calculation formula is as follows: , in, Assign weights to the global initial power supply of AC / DC feeder branch i at time t. This is used to identify the real-time operating status of the j-th AC / DC feeder branch.

[0037] For each strongly coupled branch i and j marked as a candidate group for coordinated allocation, the fluctuation synchronization level label corresponding to the branch pair is obtained. Based on the fluctuation synchronization level label, the global initial power supply allocation weights of the two branches in the branch pair are adjusted differentially, specifically as follows: If it is a synchronous dispatch group: Extract the real-time remaining capacity margin of the two branches in the middle of the branch, and use the arithmetic mean calculation formula to obtain the average remaining capacity margin of the synchronous dispatch group. For AC / DC feeder branches within a synchronous dispatch group, the formula is used: The real-time power supply allocation weight adjustment amount of the AC / DC feeder branch at time t is obtained. ,in, The rated carrying capacity of the AC / DC feeder branch. The average remaining load margin of the synchronous allocation group, This is the preset synchronization adjustment coefficient; The real-time global initial power supply allocation weights of the two branches are adjusted synchronously to obtain the adjusted real-time power supply allocation weights. , ; The adjusted formula is as follows: ,in, Assign weights to the adjusted real-time power supply of the AC / DC feeder branch at time t.

[0038] If it is a complementary allocation group: Extract the peak and valley time parameters of the load time evolution curves of the two branches in the middle of the branch, and compare them with the current time to determine the branch that is at peak load and the branch that is at valley load at the current time. Obtain the real-time remaining capacity margin of the valley load branch and record it as the adjustable remaining capacity of that branch pair. Simultaneously, obtain the power supply priority weights of the peak load branch and the valley load branch, and calculate the adjustment amount of the power supply allocation weights for the two branches. The calculation formula is as follows: , in, The adjustable remaining capacity of the branch pair at time t; The initial global power allocation weights of the two branches at time t are adjusted in reverse to obtain the adjusted power allocation weights. , The formula is adjusted as follows: .

[0039] If it is an independent timing group: Preset buffer margin ratio coefficient The initial global power allocation weights for the two branches are pre-defined with margin, resulting in the final power allocation weights for the two branches. , The calculation formula is: .

[0040] For weakly coupled branch pairs marked as local fit candidate groups: For each pair of weakly coupled branches i and j marked as local fit candidate groups, calculate the load change rate of the two branches respectively. , The calculation formula is: ,in, The synchronous acquisition interval; If the load change rate of any branch is within the preset local fluctuation judgment interval at the current moment, it is judged to be in the local fluctuation segment. Within the local fluctuation range, the correlation coefficient of local coordinated fluctuation between two branches is calculated by using time series correlation analysis to trace back a preset duration of local fluctuation from the current time. Several local coordinated fluctuation correlation coefficient levels are set, and each local coordinated fluctuation correlation coefficient level corresponds to a small adjustment range of power supply allocation weight. The calculated local co-fluctuation correlation coefficient is matched to the corresponding local co-fluctuation correlation coefficient level, the corresponding small adjustment range of power supply allocation weight is matched, and based on the matched small adjustment range of power supply allocation weight, the global initial power supply allocation weight of the two branches is slightly corrected and adjusted according to the preset adjustment amount, and the adjustment amount does not exceed the preset maximum adjustment threshold. If the current time is not in a local fluctuation zone, the global initial power supply allocation weights of the two branches remain unchanged.

[0041] For unassociated coupled branch pairs marked as independent allocation groups: For each pair of unrelated coupled branches i and j marked as independent allocation groups, without considering any linkage between the two branches, the global initial power supply allocation weight is directly used as the final power supply allocation weight.

[0042] After completing the differential adjustment of all branch pairs, the adjusted weights of all normally operating branches are globally normalized to obtain the final power supply allocation weights of each branch.

[0043] For each AC / DC feeder branch, combined with the overall real-time load of the entire station Branch rated load capacity The active power output parameters of each branch are dynamically adjusted, specifically as follows: Calculate the initial output active power setting parameters: ; Rated power carrying capacity limit: ; in, Let be the real-time output active power setting parameters for the i-th AC / DC feeder branch at time t. This is an operation to find the minimum value; The adjusted active power output parameters of each branch are sent to the control execution unit of the corresponding branch, driving the branch to output power according to the parameters; The final power supply allocation weight and corresponding output active power setting parameters of each branch are stored in the storage area for subsequent steps.

[0044] According to the set synchronous acquisition interval, the real-time remaining load margin calculation, load margin level determination, power supply priority weight generation, differentiated load allocation and power supply parameter setting process are executed cyclically to realize the real-time update of the power supply ratio parameters of each branch.

[0045] It should be noted that after each update of the full-station load status dataset, the data is retrieved to extract relevant information of the branches, and the real-time remaining load margin is calculated and classified according to the rated load capacity. This allows for a clear understanding of the actual load capacity of each branch, providing an objective basis for subsequent weight adjustments. It helps to avoid branch overload or capacity idleness and is consistent with the actual load conditions of the branches.

[0046] Two weighting libraries are constructed: one for branch importance and the other for carrying capacity. The weighting coefficients are determined by matching the corresponding levels. The two types of coefficients are then combined to generate power supply priority weights. This approach takes into account both the importance of branch loads and the current carrying capacity. The weighting no longer relies on a single dimension, reducing bias and better aligning with the actual power supply needs of the substation.

[0047] The power supply priority weights of normally operating branches are screened and globally normalized to obtain the global initial power supply allocation weights. This adapts to scenarios where multiple branches operate in coordination, making the initial weight allocation more balanced and laying a reasonable foundation for subsequent differentiated adjustments.

[0048] Differentiated adjustment weights are applied to branch pairs with different coupling types: For strongly coupled collaborative dispatch candidate groups, synchronous, reverse, or margin-reserved adjustments are made according to the fluctuation synchronicity level; for weakly coupled local adaptation candidate groups, slight adjustments are made based on the load change rate; for uncoupled independent dispatch groups, the initial weights are directly used to match the actual linkage characteristics between branches, avoid the irrationality caused by uniform adjustments, adapt to the power supply dispatch needs under different coupling scenarios, and improve the flexibility of weight adjustment.

[0049] After completing the differential adjustment, the final power supply allocation weight is obtained by normalization. Then, combined with the overall real-time load of the entire station and the rated carrying capacity of the branch, the output active power parameters are dynamically adjusted and limited to prevent the branch output from exceeding the rated carrying capacity and ensure the safe operation of the branch. The parameters are sent to the control execution unit to make the power supply output match the preset ratio requirements, and the relevant data is stored for subsequent use.

[0050] The entire process is executed cyclically according to the set synchronous acquisition interval, continuously updating the power supply ratio parameters of each branch, adapting to the characteristics of dynamic load changes, so that the power supply ratio always matches the real-time operating conditions, and helping the AC / DC power supply system maintain a stable operating state.

[0051] Deviation Closed-Loop Control Module: Continuously collects actual operating data of AC / DC feeder branches, calculates relative deviations in the power distribution ratio, and determines the deviation level; for branches with significant deviations, it traces the root causes of deviations and predicts load trends, and progressively corrects the final power supply allocation weights in stages; it readjusts the output active power setting parameters and sends them down for storage, forming a closed-loop adaptive control link to dynamically optimize the power supply ratio of each branch. The specific process is as follows: According to the set synchronous acquisition interval, the actual operating data of each AC / DC feeder branch after the power supply parameters are executed are continuously collected, including: actual active power. Actual operating condition status indicator ; For each AC / DC feeder branch, obtain its corresponding output active power setting parameters. With actual active power And using the formula: The relative deviation of the AC / DC feeder branch ratio is obtained. ; Two deviation levels are preset: normal deviation and significant deviation. Each deviation level corresponds to a range of absolute values ​​of relative deviation in the proportion, and these ranges do not overlap and are mutually exclusive. The entire range of values ​​for; The absolute value of the relative deviation of the AC / DC feeder branch ratio It matches the preset deviation level range and outputs the deviation level judgment result corresponding to the AC / DC feeder branch.

[0052] If the AC / DC feeder branch is determined to have a significant deviation, the deviation root cause tracing and adaptive fine-tuning process will be initiated, specifically as follows: For branch i identified as having a significant deviation, iterate through the branch pairs (i, j) formed by i and all other normally operating branches j, retrieve the correlation coupling level and fluctuation synchronization level of each branch pair, and simultaneously retrieve the real-time remaining capacity margin of the significantly deviating branch i. Real-time remaining carrying capacity margin of the corresponding branch to the middle branch j To determine the root cause of the deviation, the following steps are taken: If there exists at least one branch j such that the branch pair (i, j) is strongly correlated and coupled, and branch j also shows a significant deviation in the same direction, and the correlation coupling level or fluctuation synchronization level of the branch pair (i, j) at the current moment is different from the corresponding level at the previous analysis moment and has changed dynamically, then the root cause of the deviation is determined to be the change of branch linkage characteristics. If none of the branches j that form a branch pair with the significantly deviated branch i show significant deviation in the same direction, and the difference between the actual remaining load margin of branch i and the real-time remaining load margin of the previous moment exceeds the preset load margin deviation threshold (and the difference continues to exceed two collection cycles), then the root cause of the deviation is determined to be the change in the load condition of the branch itself. If at least g branches that do not have a strong correlation or coupling relationship simultaneously show significant deviations in the same direction, the root cause of the deviation is determined to be external power grid operating condition disturbance, where g is a preset threshold for the number of external power grid disturbance determination branches.

[0053] Extract branches with significant deviations and backtrack their historical active power time series within a preset time period at the current moment to construct a linear regression load trend model: , in, For the i-th AC / DC feeder branch at future time... The load forecast value; The trend slope represents the direction and rate of load evolution; This is the regression intercept.

[0054] For each branch with a significant deviation, the average predicted load value within a preset time period is calculated based on the corresponding linear regression load trend model. A comprehensive assessment of load trends and potential risks is made by combining the current actual active power, the rated load capacity of the branch circuit, and the average predicted load value. Preset positive threshold, negative threshold, and steady-state range; If the branch trend slope with significant deviation If the load value is greater than the preset positive threshold and the average predicted load value is greater than or equal to the preset overload warning threshold, then the branch load with significant deviation is judged to be on an upward trend, and there is a potential overload risk. If the trend slope of the branch with significant deviation If the load is less than the preset negative threshold and the average predicted load is less than or equal to the preset idle capacity threshold, then the branch load with significant deviation is judged to be decreasing and there is potential idle capacity. If the trend slope of the branch with significant deviation If the load is within the preset steady-state range and the deviation between the average predicted load value and the current actual active power is within the preset range, then the load is determined to be stable and there is no significant risk.

[0055] For branches identified as having significant deviations, the final power supply allocation weights of the relevant branches are adjusted in a tiered and progressive manner based on the root causes of the deviations and load trends. This involves sequentially performing deviation correction, root cause expansion correction, and trend pre-correction to obtain the updated final power supply allocation weights. Specifically: For each branch with significant deviation, calculate the deviation correction fine-tuning amount. The calculation formula is: , in, This is the preset fine-tuning coefficient.

[0056] Based on the root cause of the deviation corresponding to the significantly deviated branch, perform extended correction on the relevant branch: When the root cause of the deviation is the change in the branch linkage characteristics, the power supply allocation weight of another branch that forms a strongly correlated coupled branch pair with the branch with the significant deviation is adjusted synchronously; the weight adjustment amount of this branch is taken as the preset ratio of the deviation correction adjustment amount of the branch with the significant deviation. When the root cause of the deviation is the change in the load-bearing conditions of the branch itself, the weight correction is only performed on the branch with significant deviation, and the calculated deviation correction fine-tuning amount is directly used as the weight correction value for this time, without involving other branches; When the root cause of the deviation is the disturbance of the external power grid operating conditions, a global synchronous correction is performed on all normally operating branches. The sum of the absolute values ​​of the deviation correction fine-tuning of all branches with significant deviations is taken as the total deviation correction amount. The correction amount is distributed proportionally according to the proportion of the current final power supply allocation weight of each branch. The weight correction amount of each branch is proportional to its own weight proportion.

[0057] Based on the load trend prediction results corresponding to branches with significant deviations, a trend correction fine-tuning amount is added to branches with potential overload or idle risks: For branches where the load is trending upward, calculate the trend correction fine-tuning amount. : ; For branches where the load is decreasing, calculate the trend correction fine-tuning amount. : ; in, To pre-set the trend adjustment coefficient, for branches where the load tends to be stable and there is no significant risk, the trend correction fine-tuning amount is set to zero; (the parameters have been normalized and dimensionless before calculation). The deviation correction fine-tuning amount and the trend correction fine-tuning amount are added to the original final power supply allocation weight to obtain the fine-tuned final power supply allocation weight. The calculation formula is: ; After all corrections are completed, the fine-tuned weights of all branches with significant deviations are globally normalized to obtain the updated final power supply allocation weights.

[0058] For each branch with significant deviation, based on the updated final power allocation weights and combined with the overall real-time load of the entire station, the output active power parameters are readjusted, specifically as follows: Combined with the overall real-time load of the entire site With the updated final power allocation weight Calculate the initial output active power setting parameters for branches with significant deviations. The calculation formula is: ; By applying a limiting effect to the initial output active power setting parameters based on the branch's rated carrying capacity, ensuring that the branch's output does not exceed its rated carrying capacity, the updated output active power setting parameters are obtained, as shown in the formula: ; in, The rated carrying capacity of the branch with significant deviation. This is an operation to find the minimum value; The updated output active power parameters are sent to the control execution unit of the corresponding branch, driving the branch to perform power output according to the new parameters; At the same time, the updated final power supply allocation weight, output active power parameters, deviation tracing results, and trend prediction results will be stored in the system storage area.

[0059] According to the set synchronous acquisition interval, the actual operation data acquisition, ratio deviation calculation, deviation level judgment, deviation root cause tracing, load trend modeling, graded gradual fine adjustment, power supply parameter setting and distribution and data storage process are executed in a loop to form a closed-loop adaptive control link, continuously and dynamically optimize the power supply ratio of each branch, and maintain the long-term stable and balanced operation of the AC / DC power supply system.

[0060] It should be noted that by continuously collecting the actual active power and operating status indicators of each AC / DC feeder branch after the power supply parameters are executed, calculating the relative deviation of the distribution ratio and classifying the deviation level, the deviation between the power supply output and the set parameters can be captured in a timely manner, distinguishing between normal fluctuations and significant deviations, which facilitates targeted subsequent control, reduces ineffective adjustment operations, and improves the accuracy of control.

[0061] By conducting multi-dimensional source tracing for branches with significant deviations, and combining information on branch correlation and coupling, fluctuation synchronization, and load margin, we can distinguish three root causes: changes in branch linkage characteristics, changes in their own load conditions, and disturbances in the external power grid conditions. This can clearly locate the core cause of the deviation, avoid general judgments, provide a clear direction for subsequent corrections, and adapt to the deviation handling needs under different inducements.

[0062] By extracting historical active power time series to construct load trend models, we can predict load change trends and identify potential overload and capacity idle risks. This allows us to perceive dynamic load change trends in advance, moving beyond post-event correction and facilitating early intervention and adjustment. This reduces the impact of sudden load changes on system operation and enhances operational stability.

[0063] Based on the root causes of deviations and load trends, a graded and gradual approach is adopted to correct the power supply allocation weights. The process of deviation correction, root cause expansion correction, and trend pre-correction is completed in sequence. The adjustment process is gradual to avoid operational fluctuations caused by a one-time large change. It adapts to different deviation levels and risk levels, and the weight optimization is more in line with actual operating conditions, reducing system oscillations.

[0064] Based on the updated final power supply allocation weight, the output active power parameters are readjusted and limited according to the total station load and the rated carrying capacity of the branch. This ensures that the adjusted output parameters meet the carrying capacity of the branch and avoids the risk of overload operation. At the same time as the parameters are issued and executed, the relevant data is stored, providing a reference for subsequent control, traceability and optimization, forming a data support closed loop.

[0065] The entire control process is executed cyclically according to the synchronous acquisition interval, and a closed-loop adaptive control link is constructed. It can dynamically adapt to real-time changes in load and operating conditions, continuously optimize the power supply ratio of each branch, gradually improve the rationality of power supply distribution, and help the AC / DC power supply system maintain a long-term stable and balanced operating state.

[0066] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An AC / DC power supply system for a substation, characterized in that, include: Load Situation Awareness Module: Acquires AC / DC feeder branches for the entire station and synchronously collects load operating status; fuses and generates load time-series evolution curves, and calculates correlation coupling and fluctuation synchronization levels; Calculate the total load of the entire station and the contribution ratio of branch lines, and classify them into importance levels; Integrate multi-dimensional information to construct a full-site load status dataset and update it in real time; Power distribution setting module: calls the entire station load status dataset, calculates the real-time remaining capacity margin of branches and classifies the capacity margin level; Build a weight assignment library, generate power supply priority weights, and normalize them to obtain the global initial power supply allocation weights; The weights of candidate groups are adjusted differently based on coordination, local allocation, and independent allocation, and the final power supply allocation weights are determined by normalization. Set the output active power setting parameters and issue them, and cyclically update the power supply ratio parameters; Deviation closed-loop control module: continuously collects actual operating data of AC / DC feeder branches, calculates the relative deviation of the distribution ratio and determines the deviation level; For branches with significant deviations, the root causes of deviations and load trend predictions are carried out, and the final power supply allocation weights are gradually adjusted in stages. The output active power setting parameters are readjusted and sent to storage, forming a closed-loop adaptive control link to dynamically optimize the power supply ratio of each branch.

2. The AC / DC power supply system for a substation according to claim 1, characterized in that: The specific process for merging and generating load time-series evolution curves and calculating the correlation coupling level is as follows: Acquire all AC and DC feeder branches of the entire station, set synchronous acquisition intervals, and synchronously acquire real-time active power, operating current and operating status indicators of each branch. Select the continuous time series analysis window, normalize the active power and operating current, and then weight and fuse them to generate the load time series evolution curves of each branch. Calculate the load time-series correlation coefficient between any two branches and classify them into three levels of coupling: strong coupling, weak coupling, and no coupling. Based on different levels of association and coupling, the corresponding branch combinations are marked as collaborative allocation candidate groups, local adaptation candidate groups, and independent allocation groups, respectively.

3. The AC / DC power supply system for a substation according to claim 2, characterized in that: The specific process for calculating the fluctuation synchronization level is as follows: For any two AC / DC feeder branches, the phase offset is calculated using the cross-correlation method within the same time series analysis window, based on the load time series evolution curves of the two branches. Three levels of fluctuation synchronization are set: synchronous, anti-synchronous, and asynchronous. Each level corresponds to a phase offset range that is non-overlapping and covers the entire range of values. The phase offset and level range of each matched branch are matched one by one, and the fluctuation synchronization level judgment result is output. According to the level, the corresponding branch combinations are marked as synchronous coordination group, complementary coordination group, and independent timing group respectively.

4. The AC / DC power supply system for a substation according to claim 3, characterized in that: The specific process of classifying importance levels, integrating multi-dimensional information to construct a full-site load status dataset, and updating it in real time is as follows: The system acquires the real-time active power and operating status indicators of each AC / DC feeder branch, calculates the overall real-time load of the entire station, and calculates the contribution ratio of each AC / DC feeder branch to the overall real-time load of the entire station. Two importance levels are set for the core load branch and the auxiliary load branch. Each level corresponds to a contribution percentage range that is non-overlapping and covers the entire range of values. Match the contribution percentage and level range for each branch, and output the branch importance level determination result; By integrating real-time load operation parameters, correlation coupling level, fluctuation synchronization level, and branch importance level, a full-station load status dataset is constructed. The dataset is stored in the system storage area and updated in real time according to the synchronous acquisition interval.

5. The AC / DC power supply system for a substation according to claim 4, characterized in that: The specific process for calculating the real-time remaining carrying capacity margin of branch lines and classifying the carrying capacity margin levels is as follows: After each update of the full station load status dataset, the full station load status dataset in the system storage area is called to extract the real-time active power, branch importance level determination results, correlation coupling level markers and fluctuation synchronization level markers of each AC / DC feeder branch. Obtain the rated carrying capacity of each AC / DC feeder branch and calculate the real-time remaining carrying capacity margin of the branch; Three carrying capacity margin levels are set: sufficient, moderate, and insufficient. Each level corresponds to a non-overlapping relative range of remaining carrying capacity margin that covers the entire range of values. The system matches the real-time remaining load margin and grade range for each branch, and outputs the load margin grade determination result for each AC / DC feeder branch.

6. The AC / DC power supply system for a substation according to claim 5, characterized in that: The specific process of constructing a weight assignment library, generating power supply priority weights, and normalizing them to obtain the global initial power supply allocation weights is as follows: Construct a basic weight assignment library for the importance level of branch roads and a revised weight assignment library for the carrying capacity margin level; In the basic weight assignment library, core load branches and auxiliary load branches each correspond to a unique basic weight coefficient; In the weight assignment library, ample margin, moderate margin, and insufficient margin each correspond to a unique weight adjustment coefficient. Match the importance level and basic weight assignment library and the carrying capacity margin level and modified weight assignment library of each branch, and output the corresponding basic weight coefficient and modified weight coefficient. By combining the two types of weighting coefficients, the power supply priority weight of each AC / DC feeder branch is calculated; The power supply priority weights of all normally operating branches are selected and then processed globally to obtain the global initial power supply allocation weights of each AC / DC feeder branch.

7. The AC / DC power supply system for a substation according to claim 6, characterized in that: The weights of candidate groups are adjusted differently based on coordination, local allocation, and independent allocation. The final power supply allocation weights are then normalized to determine the final power supply allocation weights. The specific process of setting and issuing output active power setting parameters and cyclically updating power supply ratio parameters is as follows: For branch pairs marked as candidate groups for coordinated allocation, the global initial power supply allocation weight of AC / DC feeder branches is adjusted differently according to the fluctuation synchronization level. For branch pairs marked as local adaptation candidate groups, calculate the load change rate of AC / DC feeder branches. When in a local fluctuation section, slightly adjust the global initial power supply allocation weight according to the local cooperative fluctuation correlation coefficient level, and keep it unchanged in a fluctuation-free section. For branch pairs marked as independent allocation groups, the global initial power supply allocation weight is directly applied. After completing the differentiated adjustment, the adjusted weights of all normally operating branches are globally normalized to obtain the final power supply allocation weights. Based on the overall real-time load of the entire station and the rated carrying capacity of the AC / DC feeder branches, the active power setting parameters are set and limited, and then sent to the corresponding branch control execution unit. The system stores the final power supply allocation weight and output active power setting parameters, executes the corresponding process cyclically according to the synchronous acquisition interval, and updates the power supply ratio parameters of each AC / DC feeder branch in real time.

8. The AC / DC power supply system for a substation according to claim 7, characterized in that: The specific process of continuously collecting actual operating data of AC / DC feeder branches, calculating the relative deviation of the distribution ratio, and determining the deviation level is as follows: According to the set synchronous acquisition interval, continuously collect the actual active power and actual operating status indicators of each AC / DC feeder branch after the power supply parameters are executed; The output active power setting parameters and actual active power are obtained for each branch, and the relative deviation of the AC / DC feeder branch ratio is calculated. Two deviation levels are set: normal deviation and significant deviation. Each level corresponds to a non-overlapping range of absolute values ​​of relative deviation of the proportion, which covers the entire range of values. The absolute value of the relative deviation of the matching ratio and the deviation level range are used to output the deviation level determination result of the AC / DC feeder branch.

9. The AC / DC power supply system for a substation according to claim 8, characterized in that: The specific process of tracing the root causes of deviations and predicting load trends for branches with significant deviations, and then progressively adjusting the final power supply allocation weights in stages, is as follows: Once the AC / DC feeder branch is determined to have a significant deviation, the deviation root cause tracing and adaptive fine-tuning process is initiated: Traverse the branch pairs formed by the branch with significant deviation and the other normally operating branches, retrieve the correlation coupling level, fluctuation synchronization level and real-time remaining carrying capacity margin, and determine the root cause of the deviation as the change of branch linkage characteristics, the change of the branch's own carrying capacity conditions or the disturbance of the external power grid conditions. Extract the historical active power time series of branches with significant deviations, construct a linear regression load trend model, predict load trends, and identify potential overload risks or capacity idle risks; Based on the root cause of the deviation and the load trend, the deviation correction, root cause expansion correction, and trend advance correction are performed in sequence, and the final power supply allocation weight of the relevant branches is gradually corrected in stages. After fine-tuning the weights of branches with significant deviations, the weights are globally normalized to obtain the updated final power supply allocation weights.

10. An AC / DC power supply system for a substation according to claim 9, characterized in that: The specific process of resetting the output active power setting parameters and sending them for storage, forming a closed-loop adaptive control link, and dynamically optimizing the power supply ratio of each branch is as follows: For branches with significant deviations, the output active power setting parameters are readjusted and limited according to the branch's rated carrying capacity, taking into account the updated final power distribution weight and the overall real-time load of the entire station. The updated output active power setting parameters are sent to the control execution unit of the corresponding AC / DC feeder branch, and the updated final power supply allocation weight, output active power setting parameters, deviation tracing results and trend prediction results are stored in the system storage area. The entire process is executed cyclically according to the set synchronous acquisition interval, forming a closed-loop adaptive control link to dynamically optimize the power supply ratio of each AC / DC feeder branch and ensure the stable and balanced operation of the AC / DC power supply system.

Images