A load prediction-based coordinated on-load voltage regulation system and method

CN122969004APending Publication Date: 2026-10-23FUJIAN HESHENG ZHIXIN AMORPHOUS ALLOY TRANSFORMER CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611328135.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-31
Publication Date
2026-10-23

AI Technical Summary

Technical Problem

[0004]本发明的目的在于克服现有技术的不足,适应现实需要,提供一种基于负荷预测的协同有载调压系统及方法,以解决当前无末端监测台区因无法实时获取末端电压实际值、难以判断越限风险,导致主站优化指令与本地电压控制需求冲突形成调压盲区的技术问题

Benefits of technology

1、本发明通过设计虚拟末端电压估算与置信度仲裁机制,实现无末端监测台区中末端电压越限风险的超前感知和主站指令的本地适应性决策,利用台区主干线路电阻、电抗以及首端电压、有功功率和无功功率推算虚拟末端电压估算值,预测末端电压达到电压上限所需的越限时间并生成风险等级信号,同时根据虚拟末端电压估算值与末端电压实际测量值之间的历史误差确定置信度等级,当主站下发普通优化指令且风险等级为中或高时,结合置信度等级执行本地超前调压或执行主站指令,实现主站全局调度与台区本地电压安全的协调统一,解决当前无末端监测台区因无法实时获取末端电压实际值、难以判断越限风险,导致主站优化指令与本地电压控制需求冲突形成调压盲区问题。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122969004A_ABST
    Figure CN122969004A_ABST
Patent Text Reader

Abstract

The application discloses a kind of based on load prediction's coordinated on-load voltage regulating system and method, it is related to distribution network voltage control technical field, to solve the current end monitoring area without because unable real-time acquisition end voltage actual value, difficult to determine risk of overrun, leading to master station optimization instruction and local voltage control demand conflict forms voltage regulating blind area technical problem, including according to virtual end voltage estimated value and the trend of change of first end active power, predict the time required for end voltage to reach voltage upper limit overrun, and based on the comparison result of overrun time and preset length generates risk level signal.The application is by designing virtual end voltage estimation and confidence arbitration mechanism, when master station issues ordinary optimization instruction and risk level is medium or high, combined with confidence level executes local advanced voltage regulation or executes master station instruction, realizes the coordination and unification of master station global scheduling and substation local voltage safety.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of distribution network voltage control technology, and more specifically, to a cooperative on-load voltage regulation system and method based on load forecasting. Background Technology

[0002] In the construction of smart grids, rural and urban-rural fringe distribution substations, as the end of the distribution network, often have long low-voltage power supply lines with small conductor cross-sections and high line resistance. Furthermore, with the large-scale integration of distributed photovoltaic power, the total installed capacity of photovoltaic power in the substation is close to or exceeds the transformer capacity, resulting in obvious power backflow during the day.

[0003] However, these types of distribution areas typically only have voltage and current transformers installed on the low-voltage side of the transformer, and no voltage monitoring device is installed at the furthest user at the end of the area, making it impossible to obtain the actual voltage value at the end in real time. Existing on-load tap changer systems mainly rely on instructions issued by the distribution network master station or only on the measurement data at the head end for adjustment. When the master station issues ordinary optimization instructions based on the overall network voltage level, the local terminal lacks accurate perception of the voltage change trend at the end, making it difficult to determine whether there is a risk of voltage exceeding the limit at the end. This information asymmetry can easily cause conflicts between the master station's optimization instructions and local voltage control requirements. For example, when the master station requests a voltage reduction, the voltage at the end may already be too high and about to exceed the limit, or a voltage reduction may be mistakenly executed when the voltage at the end is normal, causing the voltage at the head end to exceed the lower limit, thus creating a voltage regulation blind zone and seriously affecting the power supply voltage quality. In view of this, we propose a collaborative on-load tap changer system and method based on load forecasting. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art, adapt to practical needs, and provide a collaborative on-load voltage regulation system and method based on load forecasting. This addresses the technical problem that current areas without end-monitoring cannot obtain the actual end voltage value in real time, making it difficult to judge the risk of exceeding limits, resulting in a conflict between the master station's optimization instructions and local voltage control requirements, thus creating a voltage regulation blind zone.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a coordinated on-load tap changer method based on load forecasting, comprising: Real-time electrical quantity data of the transformer head end of the distribution area are collected synchronously, and the impedance parameters of the main line of the distribution area are obtained. The real-time electrical quantity data includes the head end voltage, head end active power and head end reactive power. Based on real-time electrical quantity data and impedance parameters, the voltage at the end of the transformer area is estimated through the line voltage drop relationship, and the virtual end voltage estimate is obtained. Based on the estimated virtual end voltage and the changing trend of the active power at the beginning, the time required for the end voltage to reach the upper limit is predicted, and a risk level signal is generated based on the comparison between the over-limit time and the preset duration. The confidence level of the current prediction result is determined based on the historical error statistics between the virtual terminal voltage estimate and the actual terminal voltage measurement. It receives voltage regulation instructions from the distribution network master station, combines risk level signals and confidence levels to execute arbitration logic, generates and outputs voltage regulation control signals to drive the on-load tap changer to perform voltage regulation actions.

[0006] Preferably, based on real-time electrical quantity data and impedance parameters, the voltage at the end of the transformer substation is estimated through the line voltage drop relationship to obtain a virtual end voltage estimate, including: Read the voltage, active power, and reactive power at the transformer's start-up terminals; The virtual terminal voltage estimation value is calculated using the virtual terminal voltage estimation formula, which is as follows: ; in, This is an estimated value for the virtual terminal voltage; This refers to the voltage at the beginning of the transformer. This refers to the active power at the transformer's head end. This refers to the reactive power at the transformer's head end. The resistance of the main line in the transformer substation area; Reactors of the main power line in the transformer substation area; When there are multiple branch lines in a transformer area, the voltage drop is first calculated based on the main line, and then corrected using the branch load ratio. The branch load ratio is determined based on the ratio of the rated capacity of each branch line to the sum of the rated capacities of all branch lines in the transformer area.

[0007] Preferably, the process includes predicting the time required for the terminal voltage to reach the upper voltage limit, and generating a risk level signal based on a comparison between the over-limit time and a preset duration, including: The over-limit time is calculated using the over-limit time prediction formula for terminal voltage. The over-limit time prediction formula for terminal voltage is as follows: ; in, To predict the time of exceeding the limit; This is the upper limit of the terminal voltage; This is the estimated virtual terminal voltage value at the current moment; This represents the change in active power at the beginning of the current sampling period. This represents the change in reactive power at the beginning of the current sampling period. This is the current voltage at the beginning of the transformer; The sampling period; The resistance of the main line in the transformer substation area; Reactors of the main power line in the transformer substation area; When the over-limit time is less than the first preset duration, a high-risk level signal is generated; when the over-limit time is greater than the first preset duration but less than the second preset duration, a medium-risk level signal is generated; when the over-limit time is greater than the second preset duration, a low-risk level signal is generated; the first preset duration is the longest operating time of the on-load tap changer, and the second preset duration is the cumulative duration of voltage over-limit allowed within the statistical period.

[0008] Preferably, the confidence level of the current prediction result is determined based on the historical error statistics between the estimated virtual terminal voltage and the actual measured terminal voltage, including: Calculate the average, maximum, and standard deviation of the absolute error between the estimated virtual terminal voltage and the actual measured terminal voltage within a preset time period; The comprehensive confidence index is calculated using the formula: ; in, This is a comprehensive confidence index; The sampling sequence number; This is the average value of the absolute error over a preset time period; This represents the maximum absolute error within a preset time period. The standard deviation of the absolute error within the preset time period; , , These are the benchmark thresholds corresponding to the average error, maximum error, and standard deviation of the error, respectively. , ,and The weighting coefficients are satisfied. ; When the overall confidence index is greater than the first confidence threshold, the confidence level is determined to be high; when the overall confidence index is greater than the second confidence threshold but not greater than the first confidence threshold, the confidence level is determined to be medium; when the overall confidence index is not greater than the second confidence threshold, the confidence level is determined to be low.

[0009] Preferably, the system receives voltage regulation commands from the distribution network master station, combines risk level signals and confidence levels to execute arbitration logic, and generates and outputs voltage regulation control signals, including: When the voltage regulation command is an emergency command, the voltage regulation control signal is output unconditionally. When the voltage regulation command is a normal optimization command and the risk level signal is low risk, the voltage regulation command is executed and the voltage regulation control signal is output. When the voltage regulation command is a normal optimization command and the risk level signal is medium or high risk: if the confidence level is high, local advance voltage regulation is executed first, the voltage regulation command is temporarily suspended, and the local risk level and confidence level are sent to the distribution network master station; if the confidence level is low, local forecasting is abandoned and the voltage regulation command is executed; if the confidence level is medium, the voltage regulation command is temporarily suspended and the local forecast information and confidence level are sent to the distribution network master station, waiting for confirmation from the distribution network master station.

[0010] Preferably, the execution of the voltage regulation command is temporarily suspended, and local forecast information and confidence level are sent to the distribution network master station, pending confirmation from the distribution network master station, including: Send a request confirmation message containing the risk level and confidence level to the distribution network master station; If the distribution network master station does not update the instruction within the preset communication cycle or still insists on the original instruction, the voltage regulation instruction will be executed. If the distribution network master station modifies or cancels the original instruction based on the confirmation information sent by the request, the voltage regulation action will be executed or canceled according to the modified new instruction; the preset communication cycle is adaptively adjusted according to the communication channel quality of the current transformer area, and the waiting time is extended when the channel quality is lower than the preset quality threshold.

[0011] Preferably, local proactive voltage regulation is executed first, including: When the risk level signal is high risk or medium risk, if it is determined that the end voltage has an over-limit trend, the on-load tap changer is controlled to be reduced by one level. Continue monitoring of the estimated values ​​of the first-end voltage and the virtual end voltage after adjustment; If the voltage at the first end is not lower than the lower limit of the voltage at the first end and the estimated value of the virtual voltage at the end returns to the qualified range, then this adjustment is completed. If the difference between the first-end voltage and the lower limit of the first-end voltage is less than one level of voltage difference, and the estimated value of the virtual end voltage still exceeds the limit, then stop downgrading and switch to switching reactive power compensation equipment for adjustment. If one adjustment is insufficient, and the difference between the first-end voltage and the lower limit of the first-end voltage is not less than the voltage difference of one adjustment level, then after meeting the minimum action interval, the voltage will be reduced by one level again until the voltage returns to the qualified range or reaches the preset adjustment upper limit.

[0012] Preferably, the preset adjustment limit is the maximum number of consecutive downshifts, and the maximum number of consecutive downshifts is determined as follows: Starting from the current position of the on-load tap changer, the taps are gradually reduced. With each reduction, the voltage at the beginning of the tap decreases by one tap voltage difference. Calculate the maximum number of steps that can be reduced while the voltage at the beginning is still not lower than the lower limit of the voltage at the beginning, and use the maximum number of steps that can be reduced as the upper limit of the number of consecutive steps. When the number of consecutive downgrades reaches the upper limit, downgrades will no longer continue even if the estimated virtual terminal voltage is still higher than the terminal voltage upper limit. The lower limit of the first terminal voltage is the rated voltage minus the allowable negative deviation value, and the upper limit of the terminal voltage is the rated voltage plus the allowable positive deviation value.

[0013] A collaborative on-load tap changer system based on load forecasting includes a high-precision synchronous acquisition module, a load trend analysis and advance adjustment module, an adaptive intelligent control module, an on-load tap changer drive execution module, and a collaborative communication module. The high-precision synchronous acquisition module is used to synchronously acquire the voltage, current, active power and reactive power at the beginning of the transformer in the distribution area. The load trend analysis and advance adjustment module is used to estimate the voltage at the end of the transformer area based on the transformer area line parameters and real-time electrical quantity data at the head end, predict the risk of voltage exceeding the limit at the end and generate a risk level signal, and determine the confidence level of the current prediction result based on historical errors. The adaptive intelligent control module is used to receive voltage regulation commands issued by the distribution network master station, combine risk level signals and confidence levels to execute arbitration logic, and output voltage regulation control signals. The on-load tap changer drive execution module is used to drive the on-load tap changer to perform tap switching according to the tap control signal, and to send back the gear position feedback signal. The collaborative communication module is used to communicate with the distribution network master station and reactive power compensation equipment in the transformer area, upload local operating data and forecast information, and receive instructions from the master station.

[0014] An electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the program to implement a coordinated on-load voltage regulation method based on load forecasting.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention, through the design of a virtual end-point voltage estimation and confidence level arbitration mechanism, achieves advanced perception of end-point voltage exceedance risks in transformer substations without end-point monitoring and local adaptive decision-making for master station commands. It uses the resistance and reactance of the main line of the substation, as well as the first-end voltage, active power, and reactive power to calculate the virtual end-point voltage estimate, predict the time required for the end-point voltage to reach the upper limit, and generate a risk level signal. Simultaneously, it determines the confidence level based on the historical error between the virtual end-point voltage estimate and the actual measured end-point voltage. When the master station issues a general optimization command with a medium or high risk level, it combines the confidence level to execute local advanced voltage regulation or the master station command, achieving coordinated and unified master station global scheduling and local voltage safety in the transformer substation. This solves the problem of voltage regulation blind spots caused by the inability to obtain real-time actual end-point voltage values ​​and the difficulty in judging exceedance risks in current transformer substations without end-point monitoring, leading to conflicts between master station optimization commands and local voltage control requirements.

[0016] 2. This invention further addresses the issue of adjusting amplitude control during local advance voltage regulation by designing a step-by-step local advance voltage regulation mechanism. During local advance voltage regulation, when the risk level signal is high or medium risk, the on-load tap changer is first controlled to decrease by one level. After adjustment, the estimated values ​​of the first-end voltage and the virtual end voltage are continuously monitored. If one level adjustment is insufficient, and the difference between the first-end voltage and its lower limit is not less than one level voltage difference, then the adjustment is repeated by one level after meeting the minimum action interval. This provides reliable operational assurance for subsequent information coordination with the main station during local advance voltage regulation.

[0017] 2. This invention also enhances the master station's perception of the voltage status of unmonitored distribution transformer areas and the reliability of local predictions by designing a collaborative uploading mechanism for risk level signals and confidence levels. This further addresses the problem of insufficient information coordination between local decision-making and the master station's global optimization objectives. Based on the risk level signals and confidence levels generated by the virtual end-point voltage estimation and confidence level arbitration mechanism, local voltage, winding load, switch position, reactive power data, risk level signals, and confidence levels are uploaded to the distribution network master station through a regional collaborative channel. This allows the master station to understand the risk of voltage exceeding limits in distribution transformer areas and the reliability of local predictions. It also enables the master station to verify its own decisions from a global perspective, thereby improving the overall coordination and reliability of distribution transformer voltage control. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating the overall method of the present invention; Figure 2 This is a flowchart of the virtual terminal voltage estimation process of the present invention; Figure 3 This is a flowchart illustrating the calculation of the risk level and confidence level for exceeding the limits in this invention. Figure 4 This is a flowchart of the main station instruction and local prediction arbitration logic of the present invention; Figure 5 This is a flowchart illustrating the step-by-step local advance voltage regulation execution process of the present invention. Detailed Implementation

[0019] Example: Figures 1 to 5 As shown, the present invention relates to a coordinated on-load tap changer system and method based on load forecasting, comprising: S100: Synchronously collects the voltage, current, active power and reactive power at the beginning of the transformer in the distribution area, and outputs real-time electrical quantity data with a unified timestamp. Specifically, voltage transformers and current transformers are installed in the low-voltage side busbar and outgoing circuit of the transformer to obtain the starting voltage and three-phase current; the voltage and current signals are synchronously converted into digital signals by a synchronous sampling chip to ensure that the voltage and current data at the same time have a consistent time reference; the sampled signals are subjected to multi-level digital filtering to eliminate harmonics and electromagnetic noise interference and improve data quality.

[0020] In this embodiment, a fault self-check is performed on the secondary circuit of the instrument transformer: when an abnormality is detected in the secondary circuit of the instrument transformer, an alarm signal is output and subsequent voltage regulation decisions are blocked to avoid erroneous regulation due to sampling distortion. The filtered voltage, current, active power, and reactive power data are assigned a unified timestamp for subsequent virtual end voltage estimation and trend prediction.

[0021] As a specific implementation method, the above-mentioned synchronous acquisition and digital filtering can be performed by the existing intelligent terminal equipment in the transformer area, without the need to add an additional independent acquisition device, thus reducing the workload of the transformation.

[0022] S200. Based on the transformer area line parameters and real-time electrical quantity data, estimate the voltage at the end of the transformer area to obtain the virtual end voltage estimate. The steps for estimating the voltage at the end of the transformer substation based on the substation line parameters and the real-time electrical quantity data specifically include: Obtain the resistance of the main line of the transformer area and reactance The parameters can be sourced from data in the power distribution GIS system ledger or from actual on-site measurement data. Read the voltage at the beginning of the transformer Active power at the head end and the reactive power at the head end ; Calculate the terminal voltage based on the voltage drop relationship. That is, the terminal voltage is equal to the starting voltage minus the voltage drop caused by the combined effect of active power and reactive power with line resistance and reactance, respectively. When a distribution area has multiple branch lines, the voltage drop is first calculated based on the main line, and then corrected using the branch load ratio. The branch load ratio equals the rated capacity of each branch line divided by the sum of the rated capacities of all branch lines in the distribution area. The rated capacity of each branch line is obtained from the line nameplate parameters or data in the distribution GIS system ledger. In the absence of active power measurement devices on the branch lines, this ratio is used as an approximate allocation of the load on each branch and remains constant during operation. Specifically, the terminal voltage is calculated based on the voltage drop relationship, using the following virtual terminal voltage estimation formula: ; in, This is a virtual end voltage estimate, representing the theoretical voltage calculation result of the line end of the transformer area under conditions without actual measurement. This is the voltage at the beginning of the transformer, representing the effective voltage value collected in real time by the voltage transformer at the low-voltage side busbar of the transformer; This represents the active power transmitted from the low-voltage side of the transformer. This represents the reactive power transmitted from the low-voltage side of the transformer. The resistance of the main line of the transformer substation represents the equivalent resistance from the low-voltage side outlet of the transformer to the main line at the end of the substation. This is the reactance of the main line of the transformer substation, representing the equivalent reactance from the low-voltage side outlet of the transformer to the main line at the end of the substation.

[0023] For example, suppose the main line resistance of a certain transformer area is... The resistance is 0.1Ω. The current transformer head-end voltage is 0.02Ω. It is 230V, and the active power at the first end is... 50kW (negative value for reverse feed), reactive power at the first end. The value is 5 kvar. Calculated using the virtual terminal voltage estimation formula: ; If the photovoltaic power is reversed at this time... For a power output of -50kW, the voltage drop becomes negative, meaning the voltage rises by approximately 22.17V. The estimated terminal voltage is approximately 252.17V, higher than the starting voltage. This example demonstrates that during power backfeed, the terminal voltage will be higher than the starting voltage, and a virtual terminal voltage estimate is necessary for timely detection.

[0024] In this embodiment, the obtained resistance and reactance of the main line of the transformer substation are combined with the active power and reactive power of the transformer head end read in real time. By utilizing the physical relationship between line voltage drop and transmission power, the estimated value of virtual end voltage is calculated. The estimated value of end voltage can be obtained without the need to deploy additional voltage monitoring devices at the end of the transformer substation, which reduces the transformation cost and realizes real-time estimation of end voltage.

[0025] S300: Based on the estimated virtual end voltage and the trend of change in active power at the beginning, predict whether there is a risk of exceeding the limit at the end voltage, generate a risk level signal, and determine the confidence level of the current prediction result based on the historical error between the estimated virtual end voltage and the actual measured end voltage. The specific steps for predicting whether there is a risk of exceeding the limit at the end voltage and generating a risk level signal include: Continuously record the estimated virtual terminal voltage within a preset time window. and active power at the head end The preset time window consists of multiple sampling periods, and the current sampling period is the latest completed sampling period within the preset time window; When a continuous increase in photovoltaic backfeed power and a decrease in the first-end voltage are detected... When the voltage is within the acceptable range, a linear extrapolation method is used to predict the trend of the terminal voltage. The continuous increase in photovoltaic backfeed power is manifested in: the active power at the head end... It changes monotonically along the reverse direction over multiple consecutive sampling periods. At this time, the rate of change of the terminal voltage can be determined by the change in the active power at the beginning of the current sampling period. and reactive power change It is characterized by the rate of change of line voltage drop caused by the voltage drop.

[0026] Specifically, the predicted terminal voltage reaches the terminal voltage upper limit. The required time is predicted using the following formula for predicting the end voltage over-limit time: ; in, To predict the over-limit time, it represents the estimated time required for the terminal voltage to reach the upper voltage limit under the current power change trend; This represents the upper limit of the terminal voltage, signifying the maximum acceptable value that the terminal voltage is allowed to reach. This is the estimated virtual terminal voltage at the current moment; This represents the change in active power at the beginning of the current sampling period, indicating the difference in active power at the beginning between adjacent sampling periods. This represents the change in reactive power at the beginning of the current sampling period, indicating the difference in reactive power at the beginning between adjacent sampling periods. This is the current voltage at the beginning of the transformer; The sampling period represents the time interval between two adjacent synchronous samplings; The resistance of the main line in the transformer substation area; It is the reactor of the main line of the transformer area.

[0027] The physical meaning of this formula is as follows: the numerator represents the distance between the current virtual terminal voltage and the upper voltage limit; the denominator represents the rate of change of the virtual terminal voltage under the current rate of change of active and reactive power at the head end; dividing the two gives the predicted time required for the virtual terminal voltage to reach the upper voltage limit under the current trend. For example, suppose the current virtual end voltage estimate is... 236V, upper limit of terminal voltage It is 242V, the voltage at the first end. The voltage is 230V, and the resistance of the main line in the transformer area is... The resistance is 0.1Ω. Ω is 0.02, sampling period The change in active power at the beginning of the current sampling period is 1 second. For 10kW, the change in reactive power If the value is 1 kvar, then: ; If the first preset duration is 2 seconds, and the predicted over-limit time is less than the first preset duration, a high-risk level signal should be generated, and local advance voltage regulation should be executed immediately.

[0028] Obtain the predicted over-limit time Then, a risk level signal is generated based on the length of this time: When the predicted time of exceeding the limit is less than the first preset duration, a high-risk level signal is generated; When the predicted time of exceeding the limit is greater than the first preset duration but less than the second preset duration, a medium-risk level signal is generated. When the predicted time exceeds the second preset duration, a low-risk level signal is generated.

[0029] It should be noted that the first preset duration is determined as follows: the longest operating time required for the on-load tap changer to complete tap switching from receiving the voltage regulation control signal is calculated, and this longest operating time is taken as the first preset duration. Thus, if the predicted over-limit time is less than the first preset duration, it means that even if voltage regulation is executed immediately, the tap changer cannot complete switching before the voltage exceeds the limit, and is therefore judged as high-risk. The second preset duration is determined as follows: based on the voltage qualification rate requirements of the transformer area, the allowable cumulative duration of voltage over-limit within the statistical period is determined, and this allowable cumulative duration of over-limit is used as the second preset duration. Thus, when the predicted over-limit time is greater than the second preset duration, it indicates that there is still sufficient adjustment margin while meeting the voltage qualification rate requirements, and therefore it is judged as low risk; the situation between the first and second preset durations is judged as medium risk. The first preset duration is shorter than the second preset duration; Among them, the longest operating time of the on-load tap changer can be obtained by recording the maximum time spent in the historical operation process of the switch; the voltage qualification rate requirement and statistical period of the distribution area are given by the distribution network operation management regulations; the allowable cumulative duration of voltage over-limit is calculated by multiplying the statistical period by the allowable over-limit time ratio, and the allowable over-limit time ratio is equal to 1 minus the voltage qualification rate requirement.

[0030] In this embodiment, the time margin for voltage overshoot is quantified by continuously tracking the power change rate and using linear extrapolation.

[0031] In one embodiment, the step of determining the confidence level of the current prediction result based on the historical error between the estimated virtual terminal voltage and the actual measured terminal voltage in step S300 specifically includes: Periodically obtain the actual measured value of the terminal voltage at the same time. The actual measured values ​​can be obtained by maintenance personnel carrying a portable voltmeter to measure at the end user, or by collecting data after temporarily installing a voltage monitoring device at the end. The absolute error between the estimated virtual terminal voltage and the actual measured value is calculated using the following error calculation formula: ; Furthermore, the confidence level is determined based on the error statistics results, using the following formula for calculating the comprehensive confidence index: ; in, The comprehensive confidence index is a comprehensive quantitative evaluation index representing the credibility of the current virtual end voltage prediction results. This is the sampling sequence number, and its value range is the number of samples within a preset time period, for example... , This represents the total number of samples taken within a preset time period. For the first The absolute error at each sampling time represents the absolute value of the deviation between the estimated virtual terminal voltage and the actual measured terminal voltage at the corresponding time. For the first The estimated value of the second virtual terminal voltage, i.e., the first The estimated virtual terminal voltage value is obtained by calculating the virtual terminal voltage estimation formula. The actual measured value of the terminal voltage at the corresponding moment is obtained from portable voltmeter measurement, data collection by temporary voltage monitoring devices installed at the terminal, or data collection by low-voltage smart meters at the terminal user's meter box; It is the average value of the absolute error within a preset time period, representing the overall deviation level of the error, and is derived from the arithmetic mean of the absolute errors within the preset time period. The maximum absolute error within a preset time period represents the extreme deviation level of the error, and is derived from taking the maximum value of each absolute error within the preset time period. The standard deviation of the absolute error within a preset time period represents the degree of fluctuation and dispersion of the error, and is derived from the statistical calculation of the standard deviation of the absolute error for each time period within the preset time period. , , These are the baseline thresholds corresponding to the average error, maximum error, and standard error, respectively, representing the upper limit of the corresponding error statistics allowed for the transformer area under normal operating conditions; , ,and Let be the weighting coefficients, representing the weight percentages of the mean error, maximum error, and standard deviation of error in the overall confidence index, respectively, and satisfying the following conditions: .

[0032] The mean error, maximum error, and standard deviation are normalized by dividing each by its corresponding benchmark threshold. Then, a weighted sum is used to obtain the comprehensive error evaluation value. Finally, a negative exponent is applied to map the comprehensive error evaluation value to the range of 0 to 1. The smaller the comprehensive error evaluation value, the closer the comprehensive confidence index is to 1, indicating a more reliable prediction result; conversely, the larger the comprehensive error evaluation value, the closer the comprehensive confidence index is to 0, indicating a less reliable prediction result.

[0033] It should be noted that the reference threshold is determined based on historical statistical values ​​of the virtual terminal voltage estimation error of the transformer area under normal operating conditions: The maximum value of the average error during the historical normal period is taken as the upper limit of the average error allowed during normal operation; Take the maximum value of the maximum error within the historical normal period, which is the maximum allowable error limit during normal operation; The maximum value of the standard deviation of error during the historical normal period is taken as the upper limit of the allowable error fluctuation during normal operation; The basis for this determination is as follows: when the actual error statistics exceed the historical normal level, it indicates that the current virtual end voltage estimation is abnormal and the confidence level should decrease; when all error statistics are within the historical normal range, the normalized sub-items are no greater than 1, and the overall confidence index is relatively high. Weighting coefficient , ,and The determination method is as follows: In the historical operation data, record the average error, maximum error, standard deviation of error, and whether the prediction result is consistent with the actual voltage over-limit situation for each prediction; calculate the correlation coefficient between the average error, maximum error, standard deviation of error and the prediction error state respectively; normalize the absolute value of the three correlation coefficients to obtain the weight coefficient.

[0034] For example, assuming the average error over a preset time period The maximum error is 0.8V. 1.5V, standard deviation 0.3V; reference threshold 1V, 2V 0.5V; weighting coefficient , , Let's take values ​​of 0.5, 0.3, and 0.2 respectively. Then: ; If the first confidence threshold is 0.7 and the second confidence threshold is 0.4, then the index is between the two, and the confidence level is determined to be medium. During arbitration, the main station's instructions are suspended and local prediction information is sent up.

[0035] According to the comprehensive confidence index Confidence levels are determined by dividing the numerical range: When the overall confidence index is greater than the first confidence threshold, the confidence level is high. When the overall confidence index is greater than the second confidence threshold but not greater than the first confidence threshold, the confidence level is medium. When the overall confidence index is not greater than the second confidence threshold, the confidence level is low.

[0036] The first confidence threshold is used to distinguish between high and medium confidence levels, while the second confidence threshold is used to distinguish between medium and low confidence levels. When the overall confidence index is higher than the first confidence threshold, the local prediction result is considered reliable; when it is lower than the second confidence threshold, the local prediction result is considered unreliable; and when it falls between the two, the local prediction result is considered to have some reference value but is not yet sufficient.

[0037] It should be noted that the first and second confidence thresholds are determined as follows: In historical operational data, the comprehensive confidence index corresponding to each prediction and whether the prediction result matches the actual voltage exceedance are recorded. The minimum comprehensive confidence index among historical samples where the prediction result matches the actual voltage is used as the first confidence threshold; the maximum comprehensive confidence index among historical samples where the prediction result does not match the actual voltage is used as the second confidence threshold. The first confidence threshold is greater than the second confidence threshold. When historical data is insufficient to determine the first and second confidence thresholds, the current confidence level is directly set to medium. The reason for this setting is that when the confidence level is high, the arbitration logic will prioritize executing local proactive voltage regulation and postpone the master station's instructions; this behavior assumes the local prediction is highly reliable. When the confidence level is low, the arbitration logic will directly execute the master station's instructions; this behavior assumes the local prediction is unreliable. In situations where historical data is insufficient and the reliability of the prediction cannot be verified, it is neither appropriate to directly trust the local prediction nor to completely reject it. Setting the confidence level to medium involves postponing the master station's instructions and sending the local prediction information for master station confirmation. The master station then performs a secondary verification based on global information before making a decision. Therefore, setting the level to medium when the confidence level cannot be determined is the most prudent and conservative approach.

[0038] After accumulating sufficient historical data, the first confidence threshold and the second confidence threshold will be determined in the manner described above.

[0039] In this embodiment, a multi-dimensional confidence evaluation model is constructed by comprehensively considering the average level of error, extreme peak value, and fluctuation dispersion. This model can effectively filter out the risk of false predictions caused by line parameter aging or measurement noise, and improve the reliability of arbitration logic decision-making.

[0040] As a specific implementation method, the above-mentioned actual measured value of the terminal voltage It can be obtained through low-voltage smart meters deployed at key branch nodes in the distribution area or at end-user meter boxes, and transmitted back to edge computing devices via power line carrier or low-power wireless communication for periodic error calibration.

[0041] S400 receives voltage regulation instructions from the distribution network master station and executes arbitration logic in conjunction with the risk level signal and confidence level.

[0042] Specifically, it receives voltage regulation commands issued by the distribution network master station. These commands are divided into two categories: ordinary optimization commands and emergency commands. Ordinary optimization commands are commands that require the transformers in the distribution area to adjust their output voltage according to the overall network voltage level. Emergency commands are commands that require rapid voltage adjustment after feeder overload or short-circuit faults, as well as commands that require remote forced voltage regulation from the master station.

[0043] Arbitration shall be conducted in accordance with the following rules, taking into account the risk level signal and confidence level generated by S300: When the master station issues an emergency command, it unconditionally outputs a voltage regulation control signal and does not arbitrate with the local prediction result; When the master station issues a normal optimization command and the risk level is low, it means that the local prediction has not detected that the terminal voltage is about to exceed the limit. At this time, the master station command is executed and the corresponding voltage regulation control signal is output. When the main site issues a general optimization command and the risk level is medium or high, further determine the confidence level: If the confidence level is high, it indicates that the local prediction results are reliable. In this case, the local advance voltage adjustment will be executed first, the main station command will be temporarily suspended, and the local risk level and confidence level will be sent to the main station. If the confidence level is low, it means that the local prediction result may be unreliable. In this case, the local prediction is abandoned and the main station command is executed. If the confidence level is medium, it means that the local prediction results have some reference value but are not sufficient. In this case, the execution of the main station's instructions will be temporarily suspended and the local prediction information and confidence level will be sent to the main station for confirmation. For example: Suppose a certain distribution area currently has a high risk level and a high confidence level. The distribution network master station issues a general optimization command requesting a reduction in the transformer's outlet voltage. According to arbitration logic, if the local forecast result is reliable and the risk is urgent, then local proactive voltage regulation is prioritized, and the master station command is temporarily suspended. Simultaneously, the risk level and confidence level are sent to the master station. If the risk level is high but the confidence level is low, it indicates that the local forecast is unreliable, and the local forecast is abandoned; the master station command is executed directly. If the risk level is medium and the confidence level is medium, the master station command is suspended, a request for confirmation is sent, and the master station confirms or modifies the command within a preset communication period.

[0044] The steps to suspend the execution of the master station's instructions and wait for confirmation from the master station include: sending a request for confirmation information containing the risk level and confidence level to the master station; if the master station does not update the instructions within the preset communication period or still insists on the original instructions, then the master station's instructions are executed; if the master station modifies or cancels the original instructions based on the sent information, then the new instructions are executed.

[0045] In this embodiment, a request confirmation mechanism is set up to introduce a global perspective from the main station for secondary verification when local prediction confidence is insufficient. Furthermore, based on the above principles, this approach can avoid erroneous actions caused by prediction deviations of a single local node, while also preventing control deadlocks caused by communication delays.

[0046] As a specific implementation method, the preset communication cycle can be adaptively adjusted according to the current communication channel quality of the transformer area (such as 4G / 5G network latency or fiber optic private network status), and the waiting time can be appropriately extended when the channel quality is poor.

[0047] When the arbitration result is to prioritize the implementation of local advance voltage regulation, the following process is executed: Based on the risk level signal generated by S300, when the risk level is high risk or medium risk, it is determined that the terminal voltage has an over-limit trend, triggering a step-down regulation, that is, controlling the on-load tap changer to reduce one level.

[0048] After adjustment, continue to monitor the estimated values ​​of the first-end voltage and the virtual last-end voltage. If the first-end voltage is not lower than the lower limit and the last-end voltage returns to the qualified range, the adjustment is complete. If the difference between the first-end voltage and the lower limit of the first-end voltage is less than one voltage level difference, and the last-end voltage still exceeds the limit, then stop downgrading and switch to switching the reactive power compensation equipment for adjustment. If one level adjustment is insufficient, and the difference between the first-end voltage and the lower limit of the first-end voltage is not less than one voltage level difference, then adjust one level again after meeting the minimum action interval, until the voltage returns to the qualified range or reaches the preset adjustment upper limit.

[0049] It should be noted that the acceptable range is that the first-end voltage is not lower than the lower limit of the first-end voltage obtained by subtracting the allowable negative deviation value from the rated voltage, and the estimated value of the virtual end voltage is not higher than the upper limit of the end voltage obtained by adding the allowable positive deviation value to the rated voltage. The preset adjustment upper limit is the maximum number of consecutive tap reductions. It is determined as follows: starting from the current tap position of the on-load tap changer, the tap is reduced step by step. For each reduction, the initial voltage decreases by one tap difference. The maximum number of tap reductions allowed, provided the initial voltage remains above the lower limit, is calculated. This number is the maximum number of consecutive tap reductions. When the maximum number of consecutive tap reductions is reached, even if the estimated virtual terminal voltage is still higher than the terminal voltage upper limit, further reductions will cease, and the reactive power compensation equipment will be switched on and off for adjustment.

[0050] In this embodiment, a step-by-step adjustment strategy is adopted, which involves reducing the voltage by one level each time, verifying the operation afterward, and reducing it again if necessary. This avoids the voltage at the first end from falling below the lower limit due to excessive adjustment at one time, and also prevents over-adjustment due to prediction deviation.

[0051] S500 outputs a voltage regulation control signal based on the arbitration result, driving the on-load tap changer to perform voltage regulation.

[0052] Specifically, a voltage regulation control signal is generated based on the arbitration result of S400. The voltage regulation control signal is a control command that includes the target gear.

[0053] The voltage regulation control signal is sent to the drive motor of the on-load tap changer. The drive motor drives the tap changer to switch the taps via a gear transmission mechanism. The switching process follows a sequence of first turning on and then turning off to ensure that no power outage occurs during voltage regulation.

[0054] After the tap changer is switched, the position feedback sensor transmits the current position signal in real time to verify whether the voltage regulation action is executed correctly. When the limit position is reached, the power is automatically cut off and an alarm signal is sent to prevent damage to the equipment from over-range operation.

[0055] In this embodiment, the closed-loop verification of the tap position feedback signal and the voltage regulation control signal ensures that the tap changer is actually switched to the correct position, avoiding failure to operate or malfunction due to mechanical jamming or electrical faults, thereby reducing the risk of mechanical wear and contact burning of the on-load tap changer.

[0056] After the voltage regulation action is executed, the estimated values ​​of the first-end voltage and the virtual end voltage continue to be monitored to determine the voltage regulation effect. If the adjusted voltage has returned to the qualified range, the current voltage regulation process ends; if the voltage is still not qualified, the prediction, arbitration and voltage regulation are repeated according to the above steps to form a continuous closed-loop control.

[0057] Example 2: A coordinated on-load tap changer system based on load forecasting, comprising: The high-precision synchronous acquisition module is used to synchronously acquire the voltage, current, active power and reactive power at the beginning of the transformer in the distribution area; The load trend analysis and proactive adjustment module is used to estimate the voltage at the end of the transformer area, predict the risk of exceeding limits, and determine the confidence level. The adaptive intelligent control module receives instructions from the distribution network master station and executes arbitration logic based on risk level and confidence level, outputting voltage regulation control signals; The on-load tap changer drive execution module is used to drive the on-load tap changer to operate according to the tap control signal and to send back the tap position feedback signal in real time. The collaborative communication module is used to communicate with the distribution network master station and reactive power compensation equipment in the transformer area, upload local operating data and forecast information, and receive instructions from the master station.

[0058] Example 3: An electronic device includes a memory and a processor. The memory stores a computer program, and the processor executes the program to implement a coordinated on-load voltage regulation method based on load forecasting.

[0059] The following explains the meanings of several technical terms used in the embodiments of this invention: Virtual end voltage estimate: refers to the theoretical voltage value at the end of the transformer substation line calculated from the electrical quantities and line parameters at the beginning of the substation without directly installing end voltage measuring equipment.

[0060] Confidence level: An evaluation metric used to characterize the degree to which the estimated or predicted virtual terminal voltage matches the actual physical state.

[0061] Arbitration logic: When local control policies interact with control commands from the superior master station, a decision-making mechanism is established to determine the final execution authority and control path based on preset rules.

[0062] On-load tap changer: A mechanical or electronic switching device that can change the winding turns ratio of a transformer under load to achieve voltage regulation.

[0063] Photovoltaic backfeed power: refers to the power of excess electricity flowing back into the upstream power grid when the distributed photovoltaic power generation exceeds the local load consumption; Reactive power compensation equipment: refers to equipment used to inject or absorb reactive power into the power grid of the distribution area in order to regulate voltage and improve power factor, including but not limited to parallel capacitors, parallel reactors and static var generators.

[0064] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A coordinated on-load voltage regulation method based on load forecasting, characterized in that, include: Real-time electrical quantity data of the transformer head end of the distribution area are collected synchronously, and the impedance parameters of the main line of the distribution area are obtained. The real-time electrical quantity data includes the head end voltage, head end active power and head end reactive power. Based on real-time electrical quantity data and impedance parameters, the voltage at the end of the transformer area is estimated through the line voltage drop relationship, and the virtual end voltage estimate is obtained. Based on the estimated virtual end voltage and the changing trend of the active power at the beginning, the time required for the end voltage to reach the upper limit is predicted, and a risk level signal is generated based on the comparison between the over-limit time and the preset duration. The confidence level of the current prediction result is determined based on the historical error statistics between the virtual terminal voltage estimate and the actual terminal voltage measurement. It receives voltage regulation instructions from the distribution network master station, combines risk level signals and confidence levels to execute arbitration logic, generates and outputs voltage regulation control signals to drive the on-load tap changer to perform voltage regulation actions.

2. A coordinated on-load tap changer method based on load forecasting according to claim 1, characterized in that, Based on real-time electrical quantity data and impedance parameters, the voltage at the end of the transformer substation is estimated through the line voltage drop relationship, resulting in a virtual end-voltage estimate, including: Read the voltage, active power, and reactive power at the transformer's start-up terminals; The virtual terminal voltage estimation value is calculated using the virtual terminal voltage estimation formula, which is as follows: ; in, This is an estimated value for the virtual terminal voltage; This refers to the voltage at the beginning of the transformer. This refers to the active power at the transformer's head end. This refers to the reactive power at the transformer's head end. The resistance of the main line in the transformer substation area; Reactors of the main power line in the transformer substation area; When there are multiple branch lines in a transformer area, the voltage drop is first calculated based on the main line, and then corrected using the branch load ratio. The branch load ratio is determined based on the ratio of the rated capacity of each branch line to the sum of the rated capacities of all branch lines in the transformer area.

3. A coordinated on-load voltage regulation method based on load forecasting according to claim 1, characterized in that, Predict the time required for the terminal voltage to reach the upper voltage limit, and generate a risk level signal based on the comparison between the over-limit time and a preset duration, including: The over-limit time is calculated using the over-limit time prediction formula for terminal voltage. The over-limit time prediction formula for terminal voltage is as follows: ; in, To predict the time of exceeding the limit; This is the upper limit of the terminal voltage; This is the estimated virtual terminal voltage at the current moment; This represents the change in active power at the beginning of the current sampling period. This represents the change in reactive power at the beginning of the current sampling period. This is the current voltage at the beginning of the transformer; The sampling period; The resistance of the main line in the transformer substation area; Reactors of the main power line in the transformer substation area; When the over-limit time is less than the first preset duration, a high-risk level signal is generated; when the over-limit time is greater than the first preset duration but less than the second preset duration, a medium-risk level signal is generated; when the over-limit time is greater than the second preset duration, a low-risk level signal is generated; the first preset duration is the longest operating time of the on-load tap changer, and the second preset duration is the cumulative duration of voltage over-limit allowed within the statistical period.

4. A coordinated on-load tap changer method based on load forecasting according to claim 1, characterized in that, Based on the historical error statistics between the estimated virtual terminal voltage and the actual measured terminal voltage, the confidence level of the current prediction result is determined, including: Calculate the average, maximum, and standard deviation of the absolute error between the estimated virtual terminal voltage and the actual measured terminal voltage within a preset time period; The comprehensive confidence index is calculated using the formula: ; in, This is a comprehensive confidence index; The sampling sequence number; This is the average value of the absolute error over a preset time period; This represents the maximum absolute error within a preset time period. The standard deviation of the absolute error within the preset time period; , , These are the benchmark thresholds corresponding to the average error, maximum error, and standard deviation of the error, respectively. , ,and The weighting coefficients are and satisfy the following conditions: ; When the overall confidence index is greater than the first confidence threshold, the confidence level is determined to be high; when the overall confidence index is greater than the second confidence threshold but not greater than the first confidence threshold, the confidence level is determined to be medium; when the overall confidence index is not greater than the second confidence threshold, the confidence level is determined to be low.

5. A coordinated on-load tap changer method based on load forecasting according to claim 1, characterized in that, Receive voltage regulation commands from the distribution network master station, combine risk level signals and confidence levels to execute arbitration logic, generate and output voltage regulation control signals, including: When the voltage regulation command is an emergency command, the voltage regulation control signal is output unconditionally. When the voltage regulation command is a normal optimization command and the risk level signal is low risk, the voltage regulation command is executed and the voltage regulation control signal is output. When the voltage regulation command is a normal optimization command and the risk level signal is medium or high risk: if the confidence level is high, local advance voltage regulation is executed first, the voltage regulation command is temporarily suspended, and the local risk level and confidence level are sent to the distribution network master station; if the confidence level is low, local forecasting is abandoned and the voltage regulation command is executed; if the confidence level is medium, the voltage regulation command is temporarily suspended and the local forecast information and confidence level are sent to the distribution network master station, waiting for confirmation from the distribution network master station.

6. A coordinated on-load tap changer method based on load forecasting according to claim 5, characterized in that, Temporarily suspend the execution of voltage regulation commands and send local forecast information and confidence levels to the distribution network master station, awaiting confirmation from the distribution network master station, including: Send a request confirmation message containing the risk level and confidence level to the distribution network master station; If the distribution network master station does not update the instruction within the preset communication cycle or still insists on the original instruction, the voltage regulation instruction will be executed. If the distribution network master station modifies or cancels the original instruction based on the confirmation information sent by the request, the voltage regulation action will be executed or canceled according to the modified new instruction; the preset communication cycle is adaptively adjusted according to the communication channel quality of the current transformer area, and the waiting time is extended when the channel quality is lower than the preset quality threshold.

7. A coordinated on-load tap changer method based on load forecasting according to claim 5, characterized in that, Prioritize local proactive voltage regulation, including: When the risk level signal is high risk or medium risk, if it is determined that the end voltage has an over-limit trend, the on-load tap changer is controlled to be reduced by one level. Continue monitoring of the estimated values ​​of the first-end voltage and the virtual end voltage after adjustment; If the voltage at the first end is not lower than the lower limit of the voltage at the first end and the estimated value of the virtual voltage at the end returns to the qualified range, then this adjustment is completed. If the difference between the first-end voltage and the lower limit of the first-end voltage is less than the voltage difference of one level, then stop downgrading and switch to switching the reactive power compensation equipment for adjustment. If one adjustment is insufficient, and the difference between the first-end voltage and the lower limit of the first-end voltage is not less than the voltage difference of one adjustment level, then after meeting the minimum action interval, the voltage will be reduced by one level again until the voltage returns to the qualified range or reaches the preset adjustment upper limit.

8. A coordinated on-load tap changer method based on load forecasting according to claim 7, characterized in that, The preset adjustment limit is the maximum number of consecutive downshifts. The maximum number of consecutive downshifts is determined as follows: Starting from the current position of the on-load tap changer, the taps are gradually reduced. Each time the tap is reduced, the voltage at the beginning of the switch decreases by one tap voltage difference. Calculate the maximum number of steps that can be reduced while the voltage at the beginning is still not lower than the lower limit of the voltage at the beginning, and use the maximum number of steps that can be reduced as the upper limit of the number of consecutive steps. When the number of consecutive downgrades reaches the upper limit, downgrades will no longer continue even if the estimated virtual terminal voltage is still higher than the terminal voltage upper limit. The lower limit of the first terminal voltage is the rated voltage minus the allowable negative deviation value, and the upper limit of the terminal voltage is the rated voltage plus the allowable positive deviation value.

9. A coordinated on-load tap changer system based on load forecasting, characterized in that, It includes a high-precision synchronous acquisition module, a load trend analysis and advance adjustment module, an adaptive intelligent control module, an on-load tap changer drive execution module, and a collaborative communication module; The high-precision synchronous acquisition module is used to synchronously acquire the voltage, current, active power and reactive power at the beginning of the transformer in the distribution area. The load trend analysis and advance adjustment module is used to estimate the voltage at the end of the transformer area based on the transformer area line parameters and real-time electrical quantity data at the head end, predict the risk of voltage exceeding the limit at the end and generate a risk level signal, and determine the confidence level of the current prediction result based on historical errors. The adaptive intelligent control module is used to receive voltage regulation commands issued by the distribution network master station, combine risk level signals and confidence levels to execute arbitration logic, and output voltage regulation control signals. The on-load tap changer drive execution module is used to drive the on-load tap changer to perform tap switching according to the tap control signal, and to send back the gear position feedback signal. The collaborative communication module is used to communicate with the distribution network master station and reactive power compensation equipment in the transformer area, upload local operating data and forecast information, and receive instructions from the master station.

10. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the program to implement the coordinated on-load voltage regulation method based on load forecasting as described in any one of claims 1 to 8.