Power distribution network voltage treatment closed-loop optimization method and system based on dynamic effect evaluation

Abstract

The invention discloses a power distribution network voltage governance closed-loop optimization method and system based on dynamic effect assessment, and relates to the technical field of power grid management.The method comprises the steps that working data of nodes of the whole network and state data and instruction values of all controllable devices are collected in a power distribution network, and time alignment is conducted on all the collected data; constructing a sub-time-scale core index evaluation mechanism, and respectively calculating core indexes of different time scales; the core indexes of different time scales are endowed with weights for weighted fusion to obtain a comprehensive effect value, and whether a voltage treatment strategy causes a target conflict or not is judged by calculating the collaboration degree of a target vector; quantizing the contribution of each controllable device using a sensitivity matrix; constructing an effect-cost Pareto leading edge sample point, and combining contributions of all controllable equipment into a vector to form an equipment layer energy efficiency label; and a worker adjusts a voltage treatment strategy according to the Pareto leading edge sample point and the energy efficiency label.

Description

Technical Field

[0001] This invention relates to the field of power grid management technology, specifically to a closed-loop optimization method and system for distribution network voltage management based on dynamic performance evaluation. Background Technology

[0002] Traditional power distribution networks operate in a passive mode, with voltage regulation primarily addressing changes in fixed loads. However, the large-scale integration of distributed renewable energy sources has ushered in an "active" era for power distribution networks. The intermittent and fluctuating characteristics of these power sources lead to significant temporal voltage fluctuations in the distribution network, resulting in frequent voltage spikes, drops, and limit violations. Statistics show that when the penetration rate of distributed power sources exceeds 30%, voltage fluctuation is 4-6 times higher than that of traditional power grids, making traditional management methods inadequate for handling these dynamic changes. Furthermore, the widespread adoption of new loads such as electric vehicle charging stations and precision manufacturing equipment further exacerbates voltage fluctuations, placing higher demands on the real-time and precise nature of voltage management. Breakthroughs in multiple technological fields have provided fundamental support for dynamic closed-loop optimization: First, data acquisition capabilities have been upgraded, with the voltage monitoring rate of the entire distribution network exceeding 95%. Smart meters, synchronous phasor measurement devices, and other equipment enable the acquisition of parameters such as voltage and power at the second level, providing a massive data foundation for dynamic evaluation. Second, breakthroughs in computing power and algorithms have enabled edge computing devices to process field data in real time, and deep learning algorithms can accurately predict voltage change trends with prediction errors controlled within 3%. Third, control equipment has been innovated, with power electronic equipment such as static synchronous compensators and static var generators achieving millisecond-level adjustment of reactive power, providing hardware guarantees for strategy execution. In addition, the widespread adoption of 5G communication technology has solved the problem of real-time transmission of multi-source data, making the closed-loop linkage of "monitoring-evaluation-control" a reality.

[0003] However, current traditional closed-loop optimization often only focuses on the static, instantaneous indicator of "whether the voltage is qualified at the current moment," neglecting the comprehensive effectiveness of mitigation measures at different time scales, as well as the negative impact of the mitigation process on other key indicators such as network losses, equipment lifespan, and operating costs. Furthermore, adjusting the voltage of a certain node during mitigation may cause voltage fluctuations in adjacent nodes or lead to increased line losses, i.e., "pressing down one gourd causes another to float up." Traditional methods lack quantitative assessment of the synergistic effects between control measures and between voltage and reactive or active power. Summary of the Invention

[0004] The purpose of this invention is to provide a closed-loop optimization method and system for distribution network voltage management based on dynamic effectiveness evaluation, so as to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation, the method comprising the following steps: S100. Collect working data of all nodes in the distribution network, status data and command values ​​of all controllable devices, and time-align all collected data; associate all collected data with the topology of the distribution network to form a "data-device-topology" mapping relationship. Furthermore, the specific steps to establish the mapping relationship between "data-device-topology" are as follows: S101. Collect operational data from all network nodes in the distribution network's SCADA, AMI, PMU, and distributed power monitoring systems. This operational data includes voltage, active power, and reactive power. Controllable devices include capacitor banks, static var generators, voltage regulating transformers, and photovoltaic inverters. All collected data is unified to the same timestamp. Integrating multi-source data from SCADA, AMI, PMU, and distributed power monitoring systems covers core parameters such as voltage and active or reactive power, avoiding information gaps from single data sources and providing a complete data foundation for subsequent analysis. Unifying all collected data to the same timestamp solves the time asynchrony problem of data from multiple devices and systems, ensuring the accuracy of subsequent index calculations and sensitivity analyses.

[0006] S102. Associate all collected data with the topology of the distribution network, mapping each collected data point to the physical devices and nodes of the distribution network, forming a "data-device-topology" mapping relationship. Through the "data-device-topology" mapping, abstract data is bound to the physical devices and node locations of the distribution network, giving subsequent analysis a clear physical orientation and preventing data from becoming disconnected from the actual system.

[0007] S200. After the distribution network staff implement the voltage management strategy output by the voltage management system, a core indicator evaluation mechanism with different time scales is constructed to calculate the core indicators at different time scales. Furthermore, the specific steps for calculating the core indicators at different time scales are as follows: S201. After the distribution network voltage management strategy is implemented, a time-scale core indicator evaluation mechanism is constructed. The time-scale core indicators include instantaneous scale indicators, short-time scale indicators and long-time scale indicators. For instantaneous indicators including voltage over-limit elimination rate and node voltage deviation, the number of all voltage over-limit nodes in the distribution network after the implementation of the voltage management strategy is collected, with the collection time limited to 3 seconds before and after the implementation of the voltage management strategy. The voltage over-limit elimination rate is then calculated using the following formula: ; In the formula, R elim (t) represents the voltage over-limit elimination rate, N viol(t0-) represents the number of all voltage-over-limit nodes in the distribution network before the voltage management strategy is implemented, N viol (t0+) represents the number of all voltage-over-limit nodes in the distribution network after the voltage management strategy is implemented; To calculate the node voltage deviation in the distribution network after the implementation of the voltage management strategy, the rated reference voltage of the distribution network is extracted, and the node voltage deviation is calculated using the following formula: ; In the formula, D key (t) represents the node voltage deviation of the distribution network, U i (t) represents the voltage deviation at the i-th node in the distribution network after the voltage management strategy is implemented, U ref Ω represents the rated reference voltage, and Ω represents the total number of nodes in the distribution network. S202. For short-time scale indicators including voltage fluctuation suppression rate and average voltage compliance rate, a short-time window ΔT is set, with the short-time window being on the minute level. The average standard deviation of the entire network voltage within the short-time window before and after the implementation of the voltage management strategy is collected; the voltage fluctuation suppression rate is calculated using the following formula: ; In the formula, R supp (△T) represents the voltage fluctuation suppression rate. This represents the average standard deviation of the entire network voltage within a short time window after the implementation of the voltage management strategy. This represents the average standard deviation of the entire network voltage within a short time window before the implementation of the voltage management strategy; Preset voltage acceptable range for distribution network [U] min U max ], U min U represents the lower limit of the acceptable voltage range. max This represents the upper limit of the voltage acceptable range. Within a short time window, the time the voltage at each node remains within the acceptable range is collected, and the average voltage acceptance rate is calculated using the following formula: ; In the formula, R qualofoed (△T) represents the average voltage qualification rate, T q，i (△T) represents the time during which the voltage of the i-th node is within the acceptable range when collected within a short time window; S203. For long-term indicators including changes in overall network losses and equipment operation costs, a long-term evaluation period T is preset, which is on the order of hours. The total active power loss of the distribution network at time t before and after the implementation of the voltage management strategy is collected, and the change in overall network losses is calculated using the following formula: ; In the formula, △P loss(T) represents the change in overall network loss, P loss，after (t) represents the total active power loss of the distribution network at time t after the voltage management strategy is implemented, P loss，before (t) represents the total active power loss of the distribution network at time t before the voltage management strategy is implemented, and T represents the long-term evaluation period and is used as the integration interval; Based on expert experience, the wear cost and operating loss cost per unit of reactive power regulation of each controllable device after a single action are extracted and quantified to calculate the cost of device action. The formula is as follows: ; In the formula, C action (T) represents the cost of equipment operation, c wear，k N represents the wear and tear cost of the k-th controllable device after a single action. act，k (T) represents the number of actions of the k-th controllable device during the long-term evaluation cycle, and c energy，k |△Q represents the operating loss cost per unit of reactive power regulation of the k-th controllable device. k (T) represents the cumulative absolute value of reactive power regulation of the k-th controllable device during the long-term evaluation period.

[0008] The indicators are broken down into instantaneous, short-term, and long-term scales, corresponding to different scenario requirements such as "rapid elimination of voltage exceedances," "suppression of voltage fluctuations," and "control of long-term network and equipment losses," avoiding the one-sidedness of assessments based on a single time scale. The calculation logic of each scale indicator is clearly defined, transforming the "voltage governance effect" from a qualitative description into quantitative data, which facilitates subsequent performance comparison and strategy adjustment.

[0009] The instantaneous scale (within 3 seconds) matches the motion effects of fast-response devices such as SVG, while the long-term scale adapts to the cumulative effects of slow-response devices such as OLTC, allowing the indicator evaluation to be accurately matched with the device's motion characteristics, avoiding the irrationality of "judging instantaneous devices with long-term indicators".

[0010] S300: Normalize the core indicators at different time scales, assign weights to the core indicators at different time scales and perform weighted fusion to obtain a comprehensive performance value, and dynamically adjust the weights according to the current distribution network status. Furthermore, the specific steps for assigning weights to core indicators at different time scales and then weighting and merging them to obtain a comprehensive performance value are as follows: S301. Normalize the core indicators of different time scales to the [0,1] interval to eliminate the difference in dimensions. For the three core indicators of node voltage deviation, comprehensive network loss change and equipment operation cost, when normalizing, subtract the initial normalized value from 1 to obtain the final normalized value. Weights are assigned to core indicators at different time scales. First, time scale weights are allocated. Then, the weights of the two core indicators at the same time scale are redistributed. Finally, the six core indicators across the three time scales are weighted and fused using the weights, as shown in the formula: ; In the formula, P(t) represents the overall effectiveness value, and w inst w short w long These represent the weights of the instantaneous scale index, the short-term scale index, and the long-term scale index, respectively. , , These represent the normalized vectors of instantaneous, short-term, and long-term scale indicators, respectively. Through weighted fusion, a single "comprehensive effectiveness value" is obtained, condensing the multi-dimensional and multi-time-scale evaluation results into easily understandable and comparable core indicators, thus reducing the decision-making complexity for staff.

[0011] S302. Set a weighted constraint mechanism, using the proportion of voltage-limit-exceeding nodes in the distribution network to the total number of nodes as the severity of voltage limits in the distribution network. Preset an over-limit threshold. When the severity exceeds the threshold, the weight of the instantaneous scale indicator should be the maximum; when the severity is less than the threshold, the weight of the long-term scale indicator should be the maximum. inst +w short +w long =1. Adjust the weights of indicators at each time scale according to the real-time status of the distribution network to avoid a "one-size-fits-all" assessment caused by fixed weights.

[0012] S400 quantifies the three objectives of voltage quality, network loss and control cost of the distribution network into vector form, and judges whether the voltage management strategy causes objective conflict by calculating the synergy of the objective vector. Furthermore, the specific steps for determining whether the voltage management strategy causes target conflict by calculating the degree of cooperation of the target vector are as follows: S401. Quantify the three objectives of distribution network voltage quality, network loss, and control cost into vector form. The voltage quality is the reciprocal of the average voltage qualification rate, and the network loss is the total active power loss P. loss The control cost is the equipment operation cost; the target vectors before and after the voltage management strategy execution are obtained by extracting the three target quantizations before and after the voltage management strategy execution. and ; S402. Calculate the degree of synergy between the target vector before and after the implementation of the voltage control strategy. The formula is: ; In the formula, r represents the degree of synergy between the target vectors before and after the voltage management strategy is implemented, where r ∈ [-1, 1]. When r ≈ 1, it indicates strong synergy, indicating that the voltage management strategy is effective and conflict-free; when r ≈ -1, it indicates strong conflict, indicating that the voltage management strategy harms other targets; and when r ≈ 0, it indicates no correlation. By quantifying the three core targets—voltage quality, network loss, and control cost—as vectors, the synergy analysis determines whether the strategy "sacrifices one for another," avoiding the passive situation of discovering conflicts only after the strategy is implemented.

[0013] S500: When the voltage management strategy is executed, calculate the sensitivity matrix of the action of each controllable device to the voltage of each node in the distribution network and the total active power loss, and use the sensitivity matrix to quantify the contribution of each controllable device. Furthermore, the specific steps for quantifying the contribution of each controllable device using the sensitivity matrix are as follows: S501. In the power distribution network system, based on the power flow equations, the sensitivity matrix of the action of each controllable device to the voltage of each node and the total active power loss of the distribution network is calculated using numerical methods when the voltage management strategy is implemented. The contribution of each controllable device is quantified using the sensitivity matrix, and the formula is as follows: ; In the formula, C dev，k This represents the contribution of the k-th controllable device. This represents the sensitivity matrix of the action of the k-th controllable device to the voltage of each node in the distribution network. This represents the sensitivity matrix Δu to the action of the k-th controllable device and the total active power loss of the distribution network. k This represents the action quantity of the k-th controllable device. This represents the voltage management adjustment amount at node i of the voltage management strategy; the action amount is extracted from the voltage management strategy; the sensitivity matrix clarifies the specific impact of the action of a single controllable device on the voltage of each node and the total network loss, solving the problem of "not knowing which device contributes the most to the management effect" and avoiding blind adjustment.

[0014] When the contribution is positive, it is determined that the corresponding controllable equipment is moving in the correct direction, and the voltage control effect is good; when the contribution is negative, it is determined that the corresponding controllable equipment is moving in the wrong direction, and the voltage control effect is poor. Converting equipment contribution into quantifiable data makes it easier to identify high-quality equipment with "high contribution and low cost" and inefficient equipment with "low contribution and high cost," providing "precise control targets" for subsequent strategy optimization.

[0015] S600: Construct Pareto frontier sample points for "efficiency-cost" and combine the contributions of all controllable equipment into vectors to form equipment-level energy efficiency labels; staff adjust voltage management strategies based on Pareto frontier sample points and energy efficiency labels.

[0016] Furthermore, the specific steps taken by staff to adjust the voltage management strategy based on Pareto front sample points and energy efficiency labels are as follows: S601. After each voltage management strategy is implemented, the evaluated comprehensive effectiveness value and equipment action cost are stored as a two-dimensional sample point in the historical sample database. The power distribution network system performs Pareto non-dominated sorting on the historical sample database daily to construct the Pareto front. When staff adjust the voltage management strategy, the two-dimensional sample point of the adjusted voltage management strategy is restricted to be on the Pareto front. The Pareto front sample point clarifies the optimal trade-off between effectiveness and cost, providing staff with an intuitive decision boundary and avoiding "blindly pursuing high effectiveness while ignoring costs" or "over-controlling costs while sacrificing effectiveness".

[0017] S602. After the voltage management strategy is implemented, a contribution vector is generated for all controllable devices in the distribution network. This contribution vector is stored, and when staff adjust the voltage management strategy, controllable devices are selected and prioritized based on their contribution vectors, from largest to smallest. Equipment-level energy efficiency labels integrate the contribution and cost of individual devices, allowing staff to quickly identify the "optimal energy efficiency device combination" and reduce the trial-and-error costs of strategy adjustments. The distribution network voltage governance closed-loop optimization system based on dynamic effectiveness evaluation includes a data acquisition module, a time-scale evaluation module, a comprehensive effectiveness calculation module, a strategy conflict judgment module, a contribution analysis module, and a feedback optimization module. The data acquisition module is used to collect working data of all nodes in the distribution network, status data of all controllable devices, and command values, and to perform time alignment on all collected data. The time-scale evaluation module is used to construct a time-scale core indicator evaluation mechanism after the distribution network staff executes the voltage management strategy output by the voltage management system, and to calculate the core indicators at different time scales respectively. The comprehensive performance calculation module is used to normalize the core indicators at different time scales, assign weights to the core indicators at different time scales and perform weighted fusion to obtain the comprehensive performance value, and dynamically adjust the weights according to the current distribution network status. The strategy conflict judgment module is used to quantify the three objectives of voltage quality, network loss and control cost of the distribution network into vector form, and to judge whether the voltage management strategy causes target conflict by calculating the synergy of the target vector. The contribution analysis module is used to calculate the sensitivity matrix of the action of each controllable device to the voltage of each node in the distribution network and the total active power loss when the voltage management strategy is executed, and to quantify the contribution of each controllable device using the sensitivity matrix. The feedback optimization module is used to construct Pareto frontier sample points of "efficiency-cost", and combine the contributions of all controllable equipment into vectors to form equipment-level energy efficiency labels; staff adjust the voltage management strategy based on the Pareto frontier sample points and energy efficiency labels.

[0018] The time-scale evaluation module includes instantaneous scale units, short-time scale units, and long-time scale units; The instantaneous scale unit is used to limit the acquisition time to 3 seconds before and after the voltage management strategy is executed, and to calculate the voltage over-limit elimination rate and node voltage deviation. The short-time scale unit is used to set a short-time window ΔT, which is on the order of minutes, and to calculate the voltage fluctuation suppression rate and the average voltage qualification rate. The long-term scale unit is used to preset a long-term evaluation period T, which is on the order of hours, to calculate the overall network loss change and equipment operation cost.

[0019] The feedback optimization module includes Pareto front units and contribution vector units; The Pareto front unit is used to restrict the two-dimensional sample points of the adjusted voltage management strategy from being on the Pareto front when the staff adjusts the voltage management strategy. The contribution vector unit is used to select controllable devices for priority use in descending order of contribution vectors when staff adjust the voltage management strategy.

[0020] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention forms a complete closed-loop process from data collection, effectiveness evaluation, conflict identification, contribution quantification to strategy adjustment, ensuring that voltage management strategies can be dynamically optimized according to the real-time status of the distribution network, avoiding the problem that "one-off strategies" cannot adapt to system fluctuations.

[0021] 2. This invention simultaneously considers the three core objectives of voltage quality, network loss, and control cost. It achieves "multi-objective optimization" through methods such as synergy judgment and Pareto front construction, thus solving the one-sidedness of existing technologies that "emphasize voltage and neglect network loss" or "emphasize effectiveness and neglect cost". Attached Figure Description

[0022] Figure 1 This is a module distribution diagram of the closed-loop optimization system for distribution network voltage management based on dynamic effectiveness evaluation, as described in this invention. Figure 2 This is a schematic diagram illustrating the steps of the closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation, as described in this invention. Detailed Implementation

[0023] 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.

[0024] Example: Figures 1-2 As shown, the present invention provides a technical solution. A closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation, the method comprising the following steps: S100. Collect working data of all nodes in the distribution network, status data and command values ​​of all controllable devices, and time-align all collected data; associate all collected data with the topology of the distribution network to form a "data-device-topology" mapping relationship. The specific steps to establish the mapping relationship between "data-device-topology" are as follows: S101. Collect operational data from all network nodes in the distribution network's SCADA, AMI, PMU, and distributed power monitoring systems. This operational data includes voltage, active power, and reactive power. Controllable devices include capacitor banks, static var generators, voltage regulating transformers, and photovoltaic inverters. All collected data is unified to the same timestamp. Integrating multi-source data from SCADA, AMI, PMU, and distributed power monitoring systems covers core parameters such as voltage and active or reactive power, avoiding information gaps from single data sources and providing a complete data foundation for subsequent analysis. Unifying all collected data to the same timestamp solves the time asynchrony problem of data from multiple devices and systems, ensuring the accuracy of subsequent index calculations and sensitivity analyses.

[0025] S102. Associate all collected data with the topology of the distribution network, mapping each collected data point to the physical devices and nodes of the distribution network, forming a "data-device-topology" mapping relationship. Through the "data-device-topology" mapping, abstract data is bound to the physical devices and node locations of the distribution network, giving subsequent analysis a clear physical orientation and preventing data from becoming disconnected from the actual system.

[0026] S200. After the distribution network staff implement the voltage management strategy output by the voltage management system, a core indicator evaluation mechanism with different time scales is constructed to calculate the core indicators at different time scales. The specific steps for calculating the core indicators at different time scales are as follows: S201. After the distribution network voltage management strategy is implemented, a time-scale core indicator evaluation mechanism is constructed. The time-scale core indicators include instantaneous scale indicators, short-time scale indicators and long-time scale indicators. For instantaneous indicators including voltage over-limit elimination rate and node voltage deviation, the number of all voltage over-limit nodes in the distribution network after the implementation of the voltage management strategy is collected, with the collection time limited to 3 seconds before and after the implementation of the voltage management strategy. The voltage over-limit elimination rate is then calculated using the following formula: ; In the formula, R elim (t) represents the voltage over-limit elimination rate, N viol (t0-) represents the number of all voltage-over-limit nodes in the distribution network before the voltage management strategy is implemented, N viol (t0+) represents the number of all voltage-over-limit nodes in the distribution network after the voltage management strategy is implemented; To calculate the node voltage deviation in the distribution network after the implementation of the voltage management strategy, the rated reference voltage of the distribution network is extracted, and the node voltage deviation is calculated using the following formula: ; In the formula, D key (t) represents the node voltage deviation of the distribution network, U i (t) represents the voltage deviation at the i-th node in the distribution network after the voltage management strategy is implemented, U ref Ω represents the rated reference voltage, and Ω represents the total number of nodes in the distribution network. S202. For short-time scale indicators including voltage fluctuation suppression rate and average voltage compliance rate, a short-time window ΔT is set, with the short-time window being on the minute level. The average standard deviation of the entire network voltage within the short-time window before and after the implementation of the voltage management strategy is collected; the voltage fluctuation suppression rate is calculated using the following formula: ; In the formula, R supp (△T) represents the voltage fluctuation suppression rate. This represents the average standard deviation of the entire network voltage within a short time window after the implementation of the voltage management strategy. This represents the average standard deviation of the entire network voltage within a short time window before the implementation of the voltage management strategy; Preset voltage acceptable range for distribution network [U] min U max ], U min U represents the lower limit of the acceptable voltage range. max This represents the upper limit of the voltage acceptable range. Within a short time window, the time the voltage at each node remains within the acceptable range is collected, and the average voltage acceptance rate is calculated using the following formula: ; In the formula, R qualofoed (△T) represents the average voltage qualification rate, T q，i (△T) represents the time during which the voltage of the i-th node is within the acceptable range when collected within a short time window; S203. For long-term indicators including changes in overall network losses and equipment operation costs, a long-term evaluation period T is preset, which is on the order of hours. The total active power loss of the distribution network at time t before and after the implementation of the voltage management strategy is collected, and the change in overall network losses is calculated using the following formula: ; In the formula, △P loss (T) represents the change in overall network loss, P loss，after (t) represents the total active power loss of the distribution network at time t after the voltage management strategy is implemented, P loss，before (t) represents the total active power loss of the distribution network at time t before the voltage management strategy is implemented, and T represents the long-term evaluation period and is used as the integration interval; Based on expert experience, the wear cost and operating loss cost per unit of reactive power regulation of each controllable device after a single action are extracted and quantified to calculate the cost of device action. The formula is as follows: ; In the formula, C action (T) represents the cost of equipment operation, c wear，k N represents the wear and tear cost of the k-th controllable device after a single action. act，k (T) represents the number of actions of the k-th controllable device during the long-term evaluation cycle, and c energy，k |△Q represents the operating loss cost per unit of reactive power regulation of the k-th controllable device. k (T) represents the cumulative absolute value of reactive power regulation of the k-th controllable device during the long-term evaluation period.

[0027] The indicators are broken down into instantaneous, short-term, and long-term scales, corresponding to different scenario requirements such as "rapid elimination of voltage exceedances," "voltage fluctuation suppression," and "long-term network loss and equipment loss control," avoiding the one-sidedness of assessments based on a single time scale. The calculation logic of each scale indicator is clearly defined, transforming the "voltage management effect" from a qualitative description into quantitative data, facilitating subsequent performance comparisons and strategy adjustments. The instantaneous scale (within 3 seconds) matches the action effects of fast-response devices such as SVG, while the long-term scale adapts to the cumulative impact of slow-response devices such as OLTC, ensuring a precise match between indicator assessment and equipment action characteristics, avoiding the irrationality of "using long-term indicators to judge instantaneous equipment."

[0028] S300: Normalize the core indicators at different time scales, assign weights to the core indicators at different time scales and perform weighted fusion to obtain a comprehensive performance value, and dynamically adjust the weights according to the current distribution network status. The specific steps for assigning weights to core indicators at different time scales and then weighting and merging them to obtain the comprehensive performance value are as follows: S301. Normalize the core indicators of different time scales to the [0,1] interval to eliminate the difference in dimensions. For the three core indicators of node voltage deviation, comprehensive network loss change and equipment operation cost, when normalizing, subtract the initial normalized value from 1 to obtain the final normalized value. Weights are assigned to core indicators at different time scales. First, time scale weights are allocated. Then, the weights of the two core indicators at the same time scale are redistributed. Finally, the six core indicators across the three time scales are weighted and fused using the weights, as shown in the formula: ; In the formula, P(t) represents the overall effectiveness value, and w inst w short w long These represent the weights of the instantaneous scale index, the short-term scale index, and the long-term scale index, respectively. , , These represent the normalized vectors of instantaneous, short-term, and long-term scale indicators, respectively. Through weighted fusion, a single "comprehensive effectiveness value" is obtained, condensing the multi-dimensional and multi-time-scale evaluation results into easily understandable and comparable core indicators, thus reducing the decision-making complexity for staff.

[0029] S302. Set a weighted constraint mechanism, using the proportion of voltage-limit-exceeding nodes in the distribution network to the total number of nodes as the severity of voltage limits in the distribution network. Preset an over-limit threshold. When the severity exceeds the threshold, the weight of the instantaneous scale indicator should be the maximum; when the severity is less than the threshold, the weight of the long-term scale indicator should be the maximum. inst +w short +w long =1. Adjust the weights of indicators at each time scale according to the real-time status of the distribution network to avoid a "one-size-fits-all" assessment caused by fixed weights.

[0030] S400 quantifies the three objectives of voltage quality, network loss and control cost of the distribution network into vector form, and judges whether the voltage management strategy causes objective conflict by calculating the synergy of the objective vector. The specific steps for determining whether a voltage management strategy causes target conflict by calculating the degree of cooperation of the target vector are as follows: S401. Quantify the three objectives of distribution network voltage quality, network loss, and control cost into vector form. The voltage quality is the reciprocal of the average voltage qualification rate, and the network loss is the total active power loss P. lossThe control cost is the equipment operation cost; the target vectors before and after the voltage management strategy execution are obtained by extracting the three target quantizations before and after the voltage management strategy execution. and ; S402. Calculate the degree of synergy between the target vector before and after the implementation of the voltage control strategy. The formula is: ; In the formula, r represents the degree of synergy between the target vectors before and after the voltage governance strategy is implemented. r∈[-1,1], when r≈1, it indicates strong synergy, and the voltage governance strategy is judged to be good and without conflict; when r≈-1, it indicates strong conflict, and the voltage governance strategy is judged to harm the other targets; when r≈0, it indicates no correlation, which is approximately equal to the judgment that -0.5 and 0.5 are used as classification nodes and rounded off.

[0031] The three core objectives of voltage quality, network loss, and control cost are quantified into vectors. By analyzing the degree of synergy, it is possible to determine whether the strategy is "one thing is taken into account but another is not", thus avoiding the passive situation of discovering conflicts only after the strategy is implemented.

[0032] S500: When the voltage management strategy is executed, calculate the sensitivity matrix of the action of each controllable device to the voltage of each node in the distribution network and the total active power loss, and use the sensitivity matrix to quantify the contribution of each controllable device. The specific steps for quantifying the contribution of each controllable device using the sensitivity matrix are as follows: S501. In the power distribution network system, based on the power flow equations, the sensitivity matrix of the action of each controllable device to the voltage of each node and the total active power loss of the distribution network is calculated using numerical methods when the voltage management strategy is implemented. The contribution of each controllable device is quantified using the sensitivity matrix, and the formula is as follows: ; In the formula, C dev，k This represents the contribution of the k-th controllable device. This represents the sensitivity matrix of the action of the k-th controllable device to the voltage of each node in the distribution network. This represents the sensitivity matrix Δu to the action of the k-th controllable device and the total active power loss of the distribution network. k This represents the action quantity of the k-th controllable device. This represents the voltage management adjustment amount at node i of the voltage management strategy; the action amount is extracted from the voltage management strategy; the sensitivity matrix clarifies the specific impact of the action of a single controllable device on the voltage of each node and the total network loss, solving the problem of "not knowing which device contributes the most to the management effect" and avoiding blind adjustment.

[0033] When the contribution is positive, it is determined that the corresponding controllable equipment is moving in the correct direction, and the voltage control effect is good; when the contribution is negative, it is determined that the corresponding controllable equipment is moving in the wrong direction, and the voltage control effect is poor. Converting equipment contribution into quantifiable data makes it easier to identify high-quality equipment with "high contribution and low cost" and inefficient equipment with "low contribution and high cost," providing "precise control targets" for subsequent strategy optimization.

[0034] S600: Construct Pareto frontier sample points for "efficiency-cost" and combine the contributions of all controllable equipment into vectors to form equipment-level energy efficiency labels; staff adjust voltage management strategies based on Pareto frontier sample points and energy efficiency labels.

[0035] The specific steps taken by staff to adjust the voltage management strategy based on Pareto front sample points and energy efficiency labels are as follows: S601. After each voltage management strategy is implemented, the evaluated comprehensive effectiveness value and equipment action cost are stored as a two-dimensional sample point in the historical sample database. The power distribution network system performs Pareto non-dominated sorting on the historical sample database daily to construct the Pareto front. When staff adjust the voltage management strategy, the two-dimensional sample point of the adjusted voltage management strategy is restricted to be on the Pareto front. The Pareto front sample point clarifies the optimal trade-off between effectiveness and cost, providing staff with an intuitive decision boundary and avoiding "blindly pursuing high effectiveness while ignoring costs" or "over-controlling costs while sacrificing effectiveness".

[0036] S602. After the voltage management strategy is implemented, a contribution vector is generated for all controllable devices in the distribution network. This contribution vector is stored, and when staff adjust the voltage management strategy, controllable devices are selected and prioritized based on their contribution vectors, from largest to smallest. Equipment-level energy efficiency labels integrate the contribution and cost of individual devices, allowing staff to quickly identify the "optimal energy efficiency device combination" and reduce the trial-and-error costs of strategy adjustments. The distribution network voltage governance closed-loop optimization system based on dynamic effectiveness evaluation includes a data acquisition module, a time-scale evaluation module, a comprehensive effectiveness calculation module, a strategy conflict judgment module, a contribution analysis module, and a feedback optimization module. The data acquisition module is used to collect working data of all nodes in the distribution network, status data of all controllable devices, and command values, and to perform time alignment on all collected data. The time-scale evaluation module is used to construct a time-scale core indicator evaluation mechanism after the distribution network staff executes the voltage management strategy output by the voltage management system, and to calculate the core indicators at different time scales respectively. The comprehensive performance calculation module is used to normalize the core indicators at different time scales, assign weights to the core indicators at different time scales and perform weighted fusion to obtain the comprehensive performance value, and dynamically adjust the weights according to the current distribution network status. The strategy conflict judgment module is used to quantify the three objectives of voltage quality, network loss and control cost of the distribution network into vector form, and to judge whether the voltage management strategy causes target conflict by calculating the synergy of the target vector. The contribution analysis module is used to calculate the sensitivity matrix of the action of each controllable device to the voltage of each node in the distribution network and the total active power loss when the voltage management strategy is executed, and to quantify the contribution of each controllable device using the sensitivity matrix. The feedback optimization module is used to construct Pareto frontier sample points of "efficiency-cost", and combine the contributions of all controllable equipment into vectors to form equipment-level energy efficiency labels; staff adjust the voltage management strategy based on the Pareto frontier sample points and energy efficiency labels.

[0037] The time-scale evaluation module includes instantaneous scale units, short-time scale units, and long-time scale units; The instantaneous scale unit is used to limit the acquisition time to 3 seconds before and after the voltage management strategy is executed, and to calculate the voltage over-limit elimination rate and node voltage deviation. The short-time scale unit is used to set a short-time window ΔT, which is on the order of minutes, and to calculate the voltage fluctuation suppression rate and the average voltage qualification rate. The long-term scale unit is used to preset a long-term evaluation period T, which is on the order of hours, to calculate the overall network loss change and equipment operation cost.

[0038] The feedback optimization module includes Pareto front units and contribution vector units; The Pareto front unit is used to restrict the two-dimensional sample points of the adjusted voltage management strategy from being on the Pareto front when the staff adjusts the voltage management strategy. The contribution vector unit is used to select controllable devices for priority use in descending order of contribution vectors when staff adjust the voltage management strategy. Example

[0039] Taking a 10kV urban-rural fringe distribution network as the application object, the distribution network includes 30 nodes and 12 feeders. The controllable equipment configuration is as follows: Static Var Generator (SVG): 2 units; On-Load Tap Changer (OLTC): 1 unit; Capacitor Bank: 3 groups; Photovoltaic Inverter: 5 units.

[0040] Collect all data from different systems and align them by time; based on the power distribution network GIS topology map, establish a node-feeder-equipment association table (e.g., node 8 belongs to feeder 3 and is associated with SVG-1 equipment; node 1 is a power supply node and is associated with OLTC equipment). The collected voltage and power data are mapped using a three-dimensional label of "device ID-node number-feeder number", for example, "SVG-1 (ID: S001)-node 8-feeder 3-voltage 1.02pu-active power 0.3MW-reactive power 4.2Mvar", to ensure that the data can be traced back to the specific physical location and equipment.

[0041] Calculation of core metrics across different time scales, with an example focusing on instantaneous metrics (collection time: within 3 seconds before and after strategy execution): Voltage over-limit elimination rate: Before strategy execution: There were 6 nodes that exceeded the voltage limit (nodes 7, 9, 21, 23, 25, and 29). After the strategy is executed: 1 node exceeds the voltage limit (node ​​29); Calculate: R elim (t)=(6-1) / 6×100%≈83.33%.

[0042] Node voltage deviation: Rated reference voltage U ref =10kV (1.0pu), total number of nodes Ω=30; Voltage U of each node after strategy execution i (t) Range: 0.95pu-1.04pu, maximum deviation |U i (t)-U ref |=0.05pu (node ​​29); Calculate: D key (t) = max|U i (t)-U ref |=0.05pu.

[0043] Subsequently, the voltage fluctuation suppression rate was calculated to be 53.13% and the average voltage qualification rate was 98.67% in the short-term scale indicators (short-term time window ΔT = 15 minutes). In the long-term scale indicator (long-term assessment period T=24 hours), the change in comprehensive network loss is -144MWh; the equipment operation cost is 1.99. After normalizing the above indicators, we get: Voltage over-limit elimination rate = 0.8333 Node voltage deviation = 1 - 0.5000 = 0.5000 Voltage fluctuation suppression rate = 0.5313 Average voltage pass rate = 0.9335 Total network loss change = 1 - 0.5200 = 0.4800 Equipment operation cost = 1 - 0.2475 = 0.7525 Overall performance value = (0.8333×0.2+0.5000×0.2)+(0.5313×0.15+0.9335×0.2)+(0.4800×0.15+0.7525×0.1)=(0.1667+0.1000)+(0.0797+0.1867)+(0.0720+0.0753)=0.2667+0.2664+0.1473=0.6804.

[0044] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation, characterized in that: The method includes the following steps: S100. Collect working data of all nodes in the distribution network, status data and command values ​​of all controllable devices, and time-align all collected data; associate all collected data with the topology of the distribution network to form a "data-device-topology" mapping relationship. S200. After the distribution network staff implement the voltage management strategy output by the voltage management system, a core indicator evaluation mechanism with different time scales is constructed to calculate the core indicators at different time scales. S300: Normalize the core indicators at different time scales, assign weights to the core indicators at different time scales and perform weighted fusion to obtain a comprehensive performance value, and dynamically adjust the weights according to the current distribution network status. S400 quantifies the three objectives of voltage quality, network loss and control cost of the distribution network into vector form, and judges whether the voltage management strategy causes objective conflict by calculating the synergy of the objective vector. S500: When the voltage management strategy is executed, calculate the sensitivity matrix of the action of each controllable device to the voltage of each node in the distribution network and the total active power loss, and use the sensitivity matrix to quantify the contribution of each controllable device. S600: Construct Pareto frontier sample points for "efficiency-cost" and combine the contributions of all controllable equipment into vectors to form equipment-level energy efficiency labels; staff adjust voltage management strategies based on Pareto frontier sample points and energy efficiency labels.

2. The closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation according to claim 1, characterized in that: The specific steps in S100 that constitute the mapping relationship of "data-device-topology" are as follows: S101. Collect working data of all network nodes in the SCADA, AMI, PMU, and distributed power monitoring systems of the distribution network. The working data includes voltage, active power, and reactive power. The controllable devices include capacitor banks, static var generators, voltage regulating transformers, and photovoltaic inverters. All collected data are unified to the same timestamp. S102. Associate all collected data with the topology of the distribution network, and map each collected data to the physical devices and nodes of the distribution network to form a "data-device-topology" mapping relationship.

3. The closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation according to claim 2, characterized in that: The specific steps for calculating the core indicators at different time scales in S200 are as follows: S201. After the distribution network voltage management strategy is implemented, a time-scale core indicator evaluation mechanism is constructed. The time-scale core indicators include instantaneous scale indicators, short-time scale indicators and long-time scale indicators. For instantaneous scale indicators including voltage over-limit elimination rate and node voltage deviation, the number of all voltage over-limit nodes in the distribution network before and after the implementation of the voltage management strategy is collected, and the collection time is limited to 3 seconds before and after the implementation of the voltage management strategy, and the voltage over-limit elimination rate is calculated. Calculate the node voltage deviation in the distribution network after the voltage management strategy is implemented, extract the rated reference voltage of the distribution network, and calculate the node voltage deviation. S202. For short-time scale indicators including voltage fluctuation suppression rate and average voltage qualification rate, set a short-time window ΔT, which is on the minute level, and collect the average standard deviation of the voltage of the entire network within the short-time window before and after the implementation of the voltage management strategy in the distribution network; calculate the voltage fluctuation suppression rate. Preset voltage acceptable range for distribution network [U] min U max ], U min U represents the lower limit of the acceptable voltage range. max This represents the upper limit of the voltage acceptable range. The average voltage acceptable rate is calculated by collecting the time during which the voltage of each node is within the voltage acceptable range within a short time window. S203. For long-term indicators including changes in comprehensive network losses and equipment operation costs, a long-term evaluation period T is preset. The long-term evaluation period is on the hour level. The total active power loss of the distribution network at time t before and after the voltage management strategy is implemented is collected, and the changes in comprehensive network losses are calculated. Based on expert experience, the wear cost and operating loss cost per unit of reactive power regulation of each controllable device after a single action are extracted and quantified to calculate the cost of device action.

4. The closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation according to claim 3, characterized in that: The specific steps in S300 for assigning weights to core indicators at different time scales and then weighting and fusing them to obtain the comprehensive performance value are as follows: S301. Normalize the core indicators of different time scales to the [0,1] interval to eliminate the difference in dimensions. For the three core indicators of node voltage deviation, comprehensive network loss change and equipment operation cost, when normalizing, subtract the initial normalized value from 1 to obtain the final normalized value. Weights are assigned to core indicators at different time scales. First, time scale weights are allocated. Then, the weights of the two core indicators at the same time scale are redistributed. Finally, the six core indicators across the three time scales are weighted and fused using the weights, as shown in the formula: ； In the formula, P(t) represents the overall effectiveness value, and w inst w short w long These represent the weights of the instantaneous scale index, the short-term scale index, and the long-term scale index, respectively. , , These represent the normalized vectors of instantaneous scale index, short-term scale index, and long-term scale index, respectively. S302. Set a weight constraint mechanism, using the proportion of the number of voltage over-limit nodes in the distribution network to the total number of nodes as the severity of over-limit in the distribution network, preset an over-limit threshold, and when the severity of over-limit exceeds the over-limit threshold, the instantaneous scale index weight is required to be the maximum weight. When the severity of exceeding the limit is less than the threshold, the long-term scale indicator should have the highest weight. inst +w short +w long =1.

5. The closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation according to claim 4, characterized in that: The specific steps in S400 for determining whether the voltage management strategy causes target conflict by calculating the degree of cooperation of the target vector are as follows: S401. Quantify the three objectives of distribution network voltage quality, network loss, and control cost into vector form. The voltage quality is the reciprocal of the average voltage qualification rate, and the network loss is the total active power loss P. loss The control cost is the equipment operation cost; the target vectors before and after the voltage management strategy execution are obtained by extracting the three target quantizations before and after the voltage management strategy execution. and ; S402. Calculate the degree of cooperation r of the target vectors before and after the voltage control strategy is implemented, where r∈[-1,1]. When r≈1, it indicates strong cooperation, and the voltage control strategy is good and there is no conflict. When r≈-1, it indicates strong conflict, and the voltage control strategy harms the other targets. When r≈0, it indicates no correlation.

6. The closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation according to claim 5, characterized in that: The specific steps in S500 for quantifying the contribution of each controllable device using a sensitivity matrix are as follows: S501. Based on the power flow equation in the power distribution network system, the sensitivity matrix of the action of each controllable device to the voltage of each node and the total active power loss of the distribution network is calculated by numerical method when the voltage management strategy is executed. The contribution of each controllable device is quantified using a sensitivity matrix, and the action quantity is extracted from the voltage management strategy. When the contribution is positive, it is determined that the corresponding controllable equipment is moving in the correct direction and the voltage control effect is good; when the contribution is negative, it is determined that the corresponding controllable equipment is moving in the wrong direction and the voltage control effect is poor.

7. The closed-loop optimization method for distribution network voltage management based on dynamic effectiveness evaluation according to claim 6, characterized in that: The specific steps for staff to adjust the voltage management strategy based on Pareto frontier sample points and energy efficiency labels in S600 are as follows: S601. After each voltage management strategy is implemented, the comprehensive effectiveness value and equipment action cost are stored as a two-dimensional sample point in the historical sample database. The power distribution network system performs Pareto non-dominated sorting on the historical sample database every day to construct the Pareto front. When the staff adjusts the voltage management strategy, the two-dimensional sample point of the adjusted voltage management strategy is restricted to be on the Pareto front. S602. After the voltage management strategy is executed, all controllable devices in the distribution network will form a controllable device contribution vector. The contribution vector will be stored. When the staff adjusts the voltage management strategy, the controllable devices will be selected in descending order of contribution vectors for priority use.

8. A closed-loop optimization system for distribution network voltage management based on dynamic performance evaluation, characterized in that: The closed-loop optimization system for distribution network voltage management includes a data acquisition module, a time-scale evaluation module, a comprehensive effectiveness calculation module, a strategy conflict judgment module, a contribution analysis module, and a feedback optimization module. The data acquisition module is used to collect working data of all nodes in the distribution network, status data of all controllable devices, and command values, and to perform time alignment on all collected data. The time-scale evaluation module is used to construct a time-scale core indicator evaluation mechanism after the distribution network staff executes the voltage management strategy output by the voltage management system, and to calculate the core indicators at different time scales respectively. The comprehensive performance calculation module is used to normalize the core indicators at different time scales, assign weights to the core indicators at different time scales and perform weighted fusion to obtain the comprehensive performance value, and dynamically adjust the weights according to the current distribution network status. The strategy conflict judgment module is used to quantify the three objectives of voltage quality, network loss and control cost of the distribution network into vector form, and to judge whether the voltage management strategy causes target conflict by calculating the synergy of the target vector. The contribution analysis module is used to calculate the sensitivity matrix of the action of each controllable device to the voltage of each node in the distribution network and the total active power loss when the voltage management strategy is executed, and to quantify the contribution of each controllable device using the sensitivity matrix. The feedback optimization module is used to construct Pareto frontier sample points of "efficiency-cost", and combine the contributions of all controllable equipment into vectors to form equipment-level energy efficiency labels; staff adjust the voltage management strategy based on the Pareto frontier sample points and energy efficiency labels.

9. The closed-loop optimization system for distribution network voltage management based on dynamic performance evaluation according to claim 8, characterized in that: The time-scale evaluation module includes instantaneous scale units, short-time scale units, and long-time scale units; The instantaneous scale unit is used to limit the acquisition time to 3 seconds before and after the voltage management strategy is executed, and to calculate the voltage over-limit elimination rate and node voltage deviation. The short-time scale unit is used to set a short-time window ΔT, which is on the order of minutes, and to calculate the voltage fluctuation suppression rate and the average voltage qualification rate. The long-term scale unit is used to preset a long-term evaluation period T, which is on the order of hours, to calculate the overall network loss change and equipment operation cost.

10. The closed-loop optimization system for distribution network voltage management based on dynamic performance evaluation according to claim 8, characterized in that: The feedback optimization module includes a Pareto front unit and a contribution vector unit; The Pareto front unit is used to restrict the two-dimensional sample points of the adjusted voltage management strategy from being on the Pareto front when the staff adjusts the voltage management strategy. The contribution vector unit is used to select controllable devices for priority use in descending order of contribution vectors when staff adjust the voltage management strategy.

Images