Weak power grid voltage stabilizing method and system based on cable impedance matching

Abstract

The application discloses a weak electric network voltage stabilization method and system based on cable impedance matching, and belongs to the technical field of weak electric network voltage stabilization, and comprises the following steps: acquiring a compensation current of a grid-connected point and a feeder line impedance profile between the grid-connected point and an equivalent power source of a system, determining voltage drop of the compensation current along the feeder line distribution, and taking a maximum drop node as a target node; decomposing the compensation current into a sensitivity matrix according to harmonic order and sequence components; under the constraint that a compensation amount at the grid-connected point is unchanged, redistributing amplitudes and phases of each component to suppress voltage drop of the target node; comparing actual and expected voltage drops after the redistribution, correcting the sensitivity matrix or updating the impedance profile according to the deviation, and reselecting the target node, and iterating until the drop is not greater than a set threshold. The application can suppress upstream voltage drop without reducing the grid-connected point management capacity, does not need to install sensors along the line, and can adapt to working conditions.

Description

Technical Field

[0001] This invention relates to the field of low-voltage grid voltage stabilization technology, and in particular to a low-voltage grid voltage stabilization method and system based on cable impedance matching. Background Technology

[0002] In power supply systems, cables serve as the fundamental physical components for power transmission, and their equivalent impedance exhibits distributed parameter characteristics along the cable's length. In short-circuit or high-voltage power grids using large-section conductors, the cable impedance between the converter connection point (grid connection point) and the upstream equivalent power source is relatively small, and the voltage difference between the two ends of the cable is negligible. The electrical effect of the converter at the grid connection point is approximately equivalent to its direct effect on the upstream power source. Under these conditions, the cable, as a transmission channel, has a well-matched impedance with the upstream power source, and the voltage drop along the cable caused by the compensation current flowing through it does not have a significant impact.

[0003] With the increasing demand for distributed power supply and remote load power, the cables used in weakly fed power grids exhibit characteristics such as high impedance per unit length, long laying distances, and frequent branching and model changes along the cable. The equivalent impedance of these cables deviates significantly from the matching state with the upstream equivalent power source—that is, the cable impedance looking towards the system from the grid connection point is no longer proportional to the equivalent impedance of the upstream system, and the cable's own series impedance becomes a non-negligible physical factor. Under these cable conditions, the compensation current injected by the converter to address power quality issues (including voltage deviation, harmonic distortion, and three-phase imbalance) at the grid connection point will cause a significant voltage drop along the cable due to the series impedance along the cable.

[0004] The voltage drop caused by the compensation current across the distributed impedance along the cable leads to new voltage deviations and waveform distortions at an intermediate bus node upstream of the grid connection point that lacks voltage monitoring. This problem is not apparent in strong power grids where the cable impedance is well matched to the equivalent impedance of the upstream system. However, in weakly fed grids—where the cable impedance is no longer commensurate with the equivalent impedance of the upstream system, and the cable itself becomes the primary carrier of the voltage drop—it becomes a key factor limiting the spread of the compensation effect along the line.

[0005] In existing technologies, improvement methods mainly focus on adjusting the physical parameters and configuration of cables. Examples include: increasing the cross-sectional area of ​​the cable conductor to reduce impedance per unit length; shortening the cable laying distance between the compensation device and the upstream power source; adding reactive power compensation components or voltage regulating components at appropriate locations along the cable to share the voltage drop along the line; and replacing the original cable with a special cable with a low-impedance structure. All of these methods aim to match the cable impedance with the upstream equivalent power source by changing the physical parameters of the cable or the structure of the cable channel, thereby suppressing voltage drops caused by the compensation current along the line. However, the aforementioned improvement methods have the following shortcomings in engineering implementation: First, increasing the conductor cross-section and replacing special cables are limited by existing cable laying conditions, routing channels, and renovation costs, making it difficult to implement widely on existing lines; Second, adding compensation components along the line requires corresponding physical installation space and operation and maintenance conditions in the middle of the cable, which is difficult to implement in a large number of outdoor or underground cables; Third, the aforementioned methods are all one-time adjustments to the physical parameters of the cable and cannot respond to the time-varying equivalent impedance of the cable caused by operating conditions, and the matching state between the cable impedance and the upstream equivalent power supply deviates again after the operating conditions change. Summary of the Invention

[0006] The purpose of this invention is to overcome the above-mentioned shortcomings and provide a method and system for stabilizing weak power grids based on cable impedance matching, so as to solve the problems mentioned in the background art.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a method for stabilizing a weak power grid based on cable impedance matching, comprising: Obtain the compensation current at the grid connection point and the impedance profile along the feeder between the grid connection point and the system equivalent power source. Based on the impedance profile along the feeder and the compensation current, determine the voltage drop distributed along the feeder when the compensation current flows through the feeder, and take the node with the largest voltage drop as the target node. The feeder segment between the grid connection point and the target node is recorded as the first feeder segment. The compensation current is decomposed into several current components, and the voltage drop change caused by the unit change of each current component on the first feeder segment is determined. The sensitivity matrix is ​​constructed using these components as elements. The redistribution current parameters are obtained by minimizing the voltage drop amplitude at the target node based on the sensitivity matrix. Based on the sensitivity matrix and the redistribution current parameters, the expected voltage drop of the target node is calculated, and the compensation current is redistributed and injected into the feeder using the redistribution current parameters. After the redistribution is executed, the actual voltage drop generated by the target node is measured. The actual voltage drop is compared with the expected voltage drop to obtain the voltage drop deviation. When the voltage drop deviation is not greater than the set deviation threshold, the sensitivity matrix is ​​corrected according to the voltage drop deviation. When the voltage drop deviation is greater than the set deviation threshold, the impedance profile along the feeder is updated and the target node is re-determined. When the target node before and after the update is inconsistent, the sensitivity matrix is ​​reconstructed according to the updated target node, and the redistribution is executed again until the voltage drop amplitude of the target node is not greater than the set amplitude threshold.

[0008] Preferably, obtaining the compensation current at the grid connection point and the line impedance profile along the feeder from the grid connection point to the system's equivalent power source specifically includes: Based on the frequency components contained in the compensation current and the voltage response corresponding to each frequency component at the grid connection point, the frequency characteristics of the input impedance from the grid connection point to the system side are determined. The feeder line between the grid connection point and the system equivalent power supply is divided into several segments. The equivalent impedance of each segment is solved by inversion using the frequency characteristics of the input impedance at each frequency component as constraints, and the impedance profile along the feeder line is obtained.

[0009] Preferably, the frequency characteristics of the input impedance are determined by the frequency components contained in the compensation current and the corresponding voltage response, specifically including: The compensation current is used as the excitation for the inherent frequency components of harmonic current suppression, negative sequence and zero sequence compensation and voltage regulation, without injecting a separate detection signal into the feeder. The ratio of the change in voltage response at each frequency component at the grid connection point to the change in compensation current corresponding to that frequency component is used as the value of the input impedance at that frequency component. When the frequency component of the compensation current changes with the operating conditions, the input impedance of the newly appearing frequency component of the compensation current under the operating conditions is taken and the frequency characteristic of the input impedance is added.

[0010] Preferably, the inversion solution for the equivalent impedance of each segment specifically includes: Each segment along the feeder is equivalent to a two-port network containing series impedance and ground admittance. The cascade relationship of each two-port network is used to establish the correspondence between the equivalent impedance of each segment and the input impedance of the grid connection point. The input impedance is calculated at each frequency component based on the correspondence. The calculated values ​​are compared with the measured values ​​of the frequency characteristics of the input impedance. The equivalent impedance of each segment is adjusted based on the deviation obtained from the comparison until the deviation meets the set error condition. The difference in equivalent impedance between adjacent segments is limited to a range determined by the line type and segment length of the feeder.

[0011] Preferably, the process of constructing the sensitivity matrix includes: The compensation current is decomposed into several current components according to harmonic order and sequence component, and each current component corresponds to a combination of a harmonic order and a sequence component. For each current component, based on the value of the equivalent impedance of each segment of the first feeder segment in the impedance profile along the feeder under the harmonic order corresponding to the current component, the voltage drop change caused by the unit amplitude change and unit phase change of the current component on the first feeder segment is determined. The sensitivity matrix is ​​constructed using the voltage drop change corresponding to each current component as elements, such that each element in the sensitivity matrix corresponds to a current component, and the element represents the influence relationship between the change of the corresponding current component and the voltage drop of the first feeder segment.

[0012] Preferably, the redistribution current parameters are obtained by solving the sensitivity matrix, and the specific process includes: The composite compensation amount corresponding to the compensation current at the grid connection point for controlling voltage deviation, harmonic distortion and three-phase imbalance is determined, and the composite compensation amount remains unchanged before and after redistribution, which is used as a constraint condition for solving the redistribution current parameters. Based on the sensitivity matrix, the correspondence between the total voltage drop at the target node and the amplitude and phase of each current component is determined; Under the constraints, the amplitude and phase of each current component are solved, and the amplitude and phase of each current component obtained are used as the redistribution current parameters. When there are no amplitude adjustment amounts and phase adjustment amounts that satisfy the constraints, a set of amplitude adjustment amounts and phase adjustment amounts is solved to make the total voltage drop as close as possible to the voltage drop limit, and this set of amplitude adjustment amounts and phase adjustment amounts is used as the redistribution current parameters.

[0013] Preferably, when the voltage drop deviation is not greater than a set deviation threshold, the sensitivity matrix is ​​corrected based on the voltage drop deviation, specifically including: The deviation of the sensitivity matrix relative to its true value is calculated based on the amplitude and phase of each current component in the redistributed current parameters during the correction process and the voltage drop deviation, and is denoted as the sensitivity deviation. The decrease in voltage drop deviation after the current correction relative to the voltage drop deviation after the previous correction is denoted as the deviation reduction. Based on the deviation reduction and the pre-established correspondence between the deviation reduction and the correction coefficient, the correction coefficient for this correction process is determined. In the aforementioned correspondence, the larger the reduction, the larger the corresponding correction coefficient. The corresponding elements of the sensitivity matrix are corrected using the correction coefficient and the sensitivity deviation to obtain the corrected sensitivity matrix; Based on the corrected sensitivity matrix, the voltage drop deviation of this correction process is determined; the voltage drop deviation of this correction process is compared with the voltage drop deviation of the previous correction process to obtain the reduction amount of this correction process, and the reduction amount of this correction process is used to determine the correction coefficient of the next correction process; If the voltage drop deviation does not decrease after a set number of consecutive correction processes, the termination condition is met, and it is determined that the sensitivity deviation exceeds the correctable range of the sensitivity matrix, thus terminating the correction process.

[0014] Preferably, when the voltage drop deviation exceeds a set deviation threshold, the impedance profile along the feeder is updated and the target node is redefined. The specific process includes: When the voltage drop deviation is greater than the set deviation threshold, the residual of each segment in the inversion is determined; The residual is the deviation between the calculated input impedance value and the measured input impedance value of each segment; The measured value of the input impedance is the ratio of the voltage response of each frequency component at the grid connection point to the corresponding compensation current. The input impedance calculation value is the input impedance calculated based on the equivalent impedance of each segment obtained by inversion, and the correspondence between the equivalent impedance of each segment and the input impedance of the grid connection point. The equivalent impedance of each segment is re-inverted to obtain the updated impedance profile along the feeder line. Based on the updated feeder impedance profile, the voltage drop along the feeder caused by the compensation current in each segment of the feeder is re-determined, and the node with the largest voltage drop is re-acquired. When the node with the largest voltage drop that is reacquired is inconsistent with the current target node, the node with the largest voltage drop that is reacquired is used as the updated target node, and the feeder segment between the grid connection point and the updated target node is updated to the first feeder segment. When they are consistent, the target node remains unchanged.

[0015] Preferably, the method further includes responding to changes in the compensation current according to a set period, specifically including: A tuning period is set. Within each tuning period, the compensation current at the grid connection point and its frequency components are used as the basis for determining the frequency characteristics of the input impedance, the impedance profile along the feeder, the target node, and the sensitivity matrix within that tuning period. The sum of the absolute values ​​of the differences in the frequency characteristics of the input impedance determined by two adjacent tuning cycles at each frequency component is taken as the degree of change of the frequency characteristics of the input impedance. When the degree of change exceeds a set change threshold, the tuning period is shortened. When the degree of change is not greater than the set change threshold, the adjustment period is extended; Between two adjacent tuning cycles, when a new frequency component appears in the compensation current, the new frequency component is added to the value point of the frequency characteristic of the input impedance.

[0016] In addition, this invention also discloses a low-voltage grid stabilization system based on cable impedance matching, used to implement the above-mentioned low-voltage grid stabilization method based on cable impedance matching, specifically including: The target node determination module is used to obtain the compensation current at the grid connection point and the impedance profile along the feeder between the grid connection point and the system equivalent power source. Based on the impedance profile along the feeder and the compensation current, it determines the voltage drop distributed along the feeder when the compensation current flows through the feeder, and takes the node with the largest voltage drop as the target node. The feeder segment between the grid connection point and the target node is recorded as the first feeder segment. The sensitivity matrix construction module is used to decompose the compensation current into several current components, determine the voltage drop change caused by the unit change of each current component on the first feeder segment, and construct the sensitivity matrix using these components as elements. The redistribution parameter solving module is used to solve for the redistribution current parameters based on the sensitivity matrix, with the goal of minimizing the voltage drop amplitude at the target node. The redistribution injection execution module is used to calculate the expected voltage drop of the target node based on the sensitivity matrix and the redistribution current parameters, and to redistribute the compensation current and inject it into the feeder using the redistribution current parameters. The deviation assessment and iterative update module is used to measure the actual voltage drop generated by the target node after the redistribution is executed, compare the actual voltage drop with the expected voltage drop to obtain the voltage drop deviation, and when the voltage drop deviation is not greater than a set deviation threshold, correct the sensitivity matrix according to the voltage drop deviation. When the voltage drop deviation is greater than the set deviation threshold, update the impedance profile along the feeder and redetermine the target node. When the target node before and after the update is inconsistent, reconstruct the sensitivity matrix according to the updated target node and perform the redistribution again until the voltage drop amplitude of the target node is not greater than the set amplitude threshold.

[0017] Beneficial effects of this invention: 1. The present invention provides a weak grid voltage stabilization method based on cable impedance matching. By redistributing the compensation current according to harmonic order and sequence components, the voltage drop generated at the upstream intermediate bus node when the compensation current flows through the distributed impedance of the feeder remains unchanged at the grid connection point while maintaining the combined compensation amount of the compensation current at the grid connection point. Since the combined compensation amount corresponding to the treatment of voltage deviation, harmonic distortion, and three-phase imbalance at the grid connection point remains equal before and after redistribution, the power quality treatment effect at the grid connection point is not weakened by redistribution, thus achieving the suppression of upstream voltage drop without sacrificing the treatment capability at the grid connection point. Compared to methods that suppress voltage drops along a cable by altering its physical parameters, such as increasing the cross-section of the cable conductor, shortening the cable laying distance, adding compensation components along the line, or replacing special cables, this invention does not require changes to the physical parameters of the cable or the structure of the cable channel. It is not limited by existing cable laying conditions, routing channels, or modification costs, and can be implemented on existing lines. Compared to methods that reduce voltage drops along a line by limiting the amplitude of the compensation current, thereby sacrificing the grid connection point's governance capabilities, this invention achieves suppression under the constraint of unchanged synthetic compensation amount, thus balancing governance capabilities and voltage drop suppression.

[0018] 2. This invention can obtain the impedance profile along the feeder line by relying solely on multi-frequency measurements at a single point of grid connection, eliminating the need to install voltage or current sensors along the energized feeder line. Since the equivalent impedance of each segment along the feeder line is encoded with different weights in the frequency-dependent characteristics of the input impedance at the grid connection point, this invention uses multiple frequency components contained in the compensation current as excitation. The equivalent impedance of each segment is inverted by combining the input impedances of each frequency component at the grid connection point, thus obtaining the spatial impedance distribution that would otherwise require multi-point measurements along the line, with only a single-point measurement at the grid connection point. Compared to existing improvement methods that require adding components in the middle of the cable, thus demanding corresponding physical installation space and operation and maintenance conditions, this invention does not require any measurement or compensation devices along the feeder line, avoiding the construction difficulties and operational risks associated with installing devices in a large number of outdoor or underground cables. Furthermore, this invention reuses the multi-frequency components inherent in the compensation current for power quality management as the inversion excitation, without injecting additional detection signals into the feeder, ensuring that the impedance inversion does not cause secondary pollution to the feeder's power quality.

[0019] 3. This invention compares the actual voltage drop of the target node with the expected voltage drop after redistribution. Based on the magnitude of the voltage drop deviation, it corrects the deviation through two routes: sensitivity matrix correction or feeder impedance profile update. The target node is then redefined after the redistribution changes the voltage drop distribution along the feeder. When the voltage drop deviation is small, sensitivity matrix correction ensures the accuracy of the sensitivity matrix used for redistribution. When the voltage drop deviation is large, feeder impedance profile update and target node reselection ensure the suppression target is always the upstream intermediate bus node with the most severe current voltage drop. Compared to existing improvement methods, which involve one-time adjustments to cable physical parameters and cannot respond to time-varying cable equivalent impedance caused by operating conditions, this invention uses the above comparison and correction to form a closed loop. This ensures that the voltage drop suppression effect caused by the compensation current is maintained even when the cable equivalent impedance changes with operating conditions, overcoming the problem of one-time adjustments leading to a deviation in the matching state after changes in operating conditions.

[0020] 4. This invention responds to changes in the compensation current according to a set period and adaptively adjusts the set period based on the degree of change in the input impedance frequency characteristics. Within each set period, the input impedance frequency characteristics, the impedance profile along the feeder, the target node, and the sensitivity matrix are redefined using the compensation current and its constituent frequency components. When the change in the input impedance frequency characteristics is significant, the set period is shortened to accelerate tracking; when the change is minor, the set period is extended to reduce recalculation frequency. Therefore, this invention can both promptly follow the time-varying equivalent impedance of the cable and maintain the suppression effect on upstream voltage dips when the compensation current continuously changes with load and operating conditions, and reduce redundant calculations when operating conditions are stable, thus balancing the adaptability to changes in operating conditions with the economy of computation. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the process for a weak power grid voltage stabilization method based on cable impedance matching provided in an embodiment of this application.

[0022] Figure 2 A schematic diagram of a low-voltage grid voltage stabilization system module based on cable impedance matching provided in an embodiment of this application.

[0023] Figure 3 The equivalent circuit diagram for segments along the feeder line is shown.

[0024] Figure 4 The waveforms show the time-domain simulation comparison between the target node voltage drop and the synthetic compensation. Detailed Implementation

[0025] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0026] Please see Figure 1 As shown, the voltage stabilization method for weak power grids based on cable impedance matching specifically includes: Obtain the compensation current at the grid connection point and the impedance profile along the feeder between the grid connection point and the system equivalent power source. Based on the impedance profile along the feeder and the compensation current, determine the voltage drop distributed along the feeder when the compensation current flows through the feeder, and take the node with the largest voltage drop as the target node. The feeder segment between the grid connection point and the target node is recorded as the first feeder segment.

[0027] The compensation current refers to the current injected into the feeder by the converter at the grid connection point to improve power quality (voltage deviation, harmonic distortion, three-phase imbalance). It is expressed in phasor form, and its amplitude and phase differ at different frequency components. The feeder impedance profile refers to the sequence of equivalent series impedances corresponding to each segment after the feeder from the grid connection point to the system's equivalent power source is divided into several segments, arranged along the feeder's line, denoted as Z1, Z2, ... Z... N Where N is the number of feeder segments, Z k Let be the complex equivalent impedance of the k-th segment. When the compensation current flows through the segmented impedances along the feeder, a voltage drop occurs at each node along the line. In the simplified case where the feeder admittance to ground is neglected and the compensation current is approximately equal in magnitude along each segment of the feeder, the cumulative voltage drop from the grid connection point to the k-th node can be expressed as the sum of the products of the compensation current and the segmented equivalent impedances between the grid connection point and the node, as shown in equation (1): (1) In the formula, The cumulative voltage drop from the grid connection point to the kth node is expressed in volts (V). The equivalent impedance of the i-th feeder segment is expressed in ohms (Ω). The compensation current phasor is in amperes (A); i is the feeder segment number, i = 1, 2, 3, ..., k; k is the node number. When the compensation current contains multiple frequency components, equation (1) holds for each frequency component, and the voltage drop generated by each frequency component is superimposed according to the phasor. Equation (1) is used to qualitatively explain the cause of voltage drop along the line, that is, the larger the compensation current and the larger the cumulative segment impedance it flows through, the larger the cumulative voltage drop generated; in this simplified case, the cumulative voltage drop increases monotonically with the increase of the node number.

[0028] Equation (1) does not take into account the ground admittance along the feeder and assumes that the compensation current is equal in magnitude along each segment of the feeder. However, in actual weak feeder networks, there is a non-negligible ground admittance along long-distance feeders. When a major harmonic source or fluctuating load is connected to a certain intermediate node along the feeder, the compensation current used to control the load will mainly flow into that node, causing the compensation current flowing through the feeder segments after that node to decrease accordingly. In order to take into account the current shunting effect of the above-mentioned ground admittance and the change of compensation current along the line, a two-port network model including series impedance and ground admittance is used to calculate the voltage and line current of each node segment by segment. Using the system equivalent power supply voltage and the compensation current injected into the feeder by the converter as known boundary conditions, starting from the system equivalent power supply side, the voltage and line current of the previous node are obtained segment by segment from the voltage and line current of the next node until the grid connection point, as shown in Equation (2): (2) In the formula, The series impedance of the i-th feeder segment is expressed in ohms (Ω). The ground admittance for the i-th feeder segment is expressed in Siemens (S). For the i-th segment input terminal (i-th segment input terminal) The voltage phasor of a node (1), in volts (V). The line current phasor at the input terminal of the i-th segment is expressed in amperes (A). This is the voltage phasor of the i-th segment output terminal (i-th node), in volts (V). Let be the line current phasor at the output terminal of the i-th segment, in amperes (A); i is the sequence number of the feeder segment, i = 1, 2, 3, ..., N. The recursion starts from the equivalent power source side of the system, using the equivalent power source voltage and the current across its internal impedance as the initial values. The voltage and line current of each intermediate node are calculated segment by segment towards the grid connection point. At intermediate nodes connected to harmonic sources or fluctuating loads, the load current flowing into that node is correspondingly included in the line current, making the line currents on either side of that node different. The voltage drop at each node is obtained from the difference between the grid connection point voltage and the voltage at that node. The equivalent power source voltage, its internal impedance, and the compensation current are all known quantities required for the recursion; the grid connection point voltage is obtained when the recursion reaches the grid connection point and is cross-checked with the measured value at the grid connection point.

[0029] As can be seen from equation (2), due to the shunting effect of the ground admittance and the reduction of the compensation current along the line after it flows into the intermediate load node, the voltage drop amplitude of each node no longer increases monotonically with the node number, but may reach the maximum value at a certain intermediate node. That is, the node with the most severe voltage deviation caused by the compensation current may be located in the middle of the feeder line, rather than at the end of the feeder line.

[0030] The simplified model described in Equation (1) is applicable to scenarios where the feeder length is short, the ground admittance is negligible, and the compensation current is approximately equal along each segment of the feeder; the complete model described in Equation (2) is applicable to scenarios where the feeder has a long distance, the ground admittance is not negligible, or the feeder has intermediate loads along the line causing the compensation current to change along the line. In implementation, the total feeder length and the ground admittance per unit length of the cable can be used to determine the compensation current: when the shunt current caused by the total ground admittance of the feeder does not exceed 5% of the compensation current, the simplified model of Equation (1) is used; otherwise, the complete model of Equation (2) is used. Unless otherwise specified, the voltage drop at each node in each embodiment of the present invention is calculated using the complete model of Equation (2).

[0031] Because there are harmonic sources or fluctuating loads connected to intermediate nodes along the feeder, the compensation current used to control these loads mainly flows into these nodes, reducing the line current in the downstream segments of these nodes. Therefore, the voltage drop amplitude of each node obtained from the deduction is not monotonically increasing along the line, but may reach its maximum value at these intermediate nodes.

[0032] For example, consider a specific scenario: Suppose the feeder is divided into 5 segments, and the series impedance (under fundamental frequency) of each segment is Z1 = 0.2 + j0.3Ω, Z2 = 0.3 + j0.4Ω, Z3 = 0.5 + j0.6Ω, Z4 = 0.3 + j0.4Ω, and Z5 = 0.2 + j0.3Ω, respectively. The corresponding ground admittance of each segment is taken into account; the compensation current injected at the grid connection point is 50A (fundamental positive sequence component). Suppose there is a significant fluctuating load connected at the 3rd node of the feeder, and the compensation current used to manage this load mainly flows into the 3rd node, so that the line current of the segments after the 3rd node is significantly reduced. According to the formula (2), the voltage drop amplitude of each node is calculated to be approximately 18V, 43V, 82V, 65V, and 40V, respectively. As can be seen, the downstream line current decreases after the load is connected to the third node, and the voltage drop reaches its maximum value at the third node. The voltage drop at subsequent nodes no longer increases but decreases instead. The third node is the extreme point of the voltage drop along the line. Therefore, the third node is determined as the target node, and the feeder segment between the grid connection point and the third node is denoted as the first feeder segment.

[0033] The voltage offset problem, which was originally scattered along the feeder and not monitored by voltage sensors, is focused on a specific target node. This simplifies the solution object of subsequent redistribution from the voltage distribution of all nodes along the feeder to the voltage drop of the target node, reducing the solution dimension.

[0034] Obtain the line impedance profile of the feeder between the grid connection point and the system equivalent power source, specifically including: Based on the frequency components contained in the compensation current and the voltage response of the corresponding frequency components at the grid connection point, the frequency characteristics of the input impedance from the grid connection point to the system side are determined.

[0035] The feeder line between the grid connection point and the system's equivalent power source is divided into several segments. The equivalent impedance of each segment is solved by inversion using the frequency characteristics of the input impedance at each frequency component as constraints, thus obtaining the impedance profile along the feeder line.

[0036] The frequency characteristic of input impedance refers to the curve of the equivalent impedance as seen from the grid connection point to the system side as a function of frequency. It is represented by a set of discrete frequency-impedance pairs, i.e., from frequency f1 to f2. M Z was measured on each side. in (f1) to Z in (f) M ), where M is the number of frequency points participating in the inversion, Z in (f) j () is the frequency f jThe input impedance at a given point is a complex value. Inversion solution refers to determining the impedance of each segment by reverse-engineering the correspondence between the known frequency characteristics of the input impedance and the unknown equivalent impedance of the feeder segments. This correspondence is determined by the propagation characteristics of distributed parameter lines: different frequency components have different attenuation constants and phase constants along the feeder, thus each feeder segment has different degrees of influence on the input impedance at different frequencies. The impedance of each segment can be solved by simultaneously solving the input impedances at multiple frequency points. The unique solution condition for the inversion is: the number of frequency points M involved in the inversion is not less than the number of feeder segments N, i.e., M is not less than N.

[0037] Under the engineering constraints of single-point measurement at the grid connection point, spatial distribution information that would otherwise require the deployment of multiple sensors along the line was obtained, avoiding the construction difficulties and operational risks of installing sensors on energized feeders.

[0038] For a specific scenario example: For a feeder divided into 5 segments, select 6 frequency points (M equals 6, greater than N equals 5, satisfying the inversion condition) containing the compensation current, namely the fundamental frequency 50Hz, the 5th harmonic 250Hz, the 7th harmonic 350Hz, the 11th harmonic 550Hz, the 13th harmonic 650Hz, and the 17th harmonic 850Hz. Measure the complex values ​​of the input impedance at each frequency, substitute them into the correspondence between the impedance of each segment and the input impedance at the grid connection point, and solve the system of equations to obtain the equivalent impedance of each segment under the fundamental frequency.

[0039] The frequency characteristics of the input impedance are determined by the frequency components contained in the compensation current and the corresponding voltage response, specifically including: The compensation current is used as the excitation for the inherent frequency components of harmonic current suppression, negative sequence and zero sequence compensation and voltage regulation, without injecting a separate detection signal into the feeder.

[0040] The ratio of the change in voltage response at each frequency component at the grid connection point to the change in compensation current corresponding to that frequency component is used as the input impedance value for that frequency component. In other words, before and after the compensation current changes at that frequency component, the ratio of the change in voltage response at the grid connection point to the change in compensation current is taken. Using this ratio eliminates the influence of the equivalent power supply's own voltage at that frequency component, ensuring the obtained value reflects the input impedance seen from the grid connection point towards the system side.

[0041] When the frequency components of the compensation current change with the operating conditions, the input impedance value of the newly emerging frequency components of the compensation current under the operating conditions is taken and added to the frequency characteristics of the input impedance, so that the frequency components contained in the frequency characteristics of the input impedance increase with the change of operating conditions.

[0042] The inherent frequency components of the compensation current originate from the different frequency components generated by the converter to accomplish different governance tasks: suppressing the k-th harmonic requires injecting the k-th harmonic current (commonly the 5th, 7th, 11th, 13th, 17th, etc.), correcting negative and zero sequence requires injecting the corresponding sequence component, and voltage stabilization requires injecting the fundamental frequency component. The voltage response of each frequency component at the grid connection point, i.e., the voltage component corresponding to that frequency in the grid connection point bus voltage, is obtained by performing Fourier decomposition on the grid connection point voltage. The input impedance at the frequency... The value of is determined by the following formula: (3) In the formula, For frequency Input impedance (unit: Ω). The change in voltage response of this frequency component at the grid connection point is represented by the phasor (unit: V). The phasor (unit: A) is used to compensate for the change in current at this frequency component.

[0043] For example, under a specific operating condition, the compensation current contains 5th, 7th, and 11th harmonic components, allowing for the acquisition of input impedance at three frequency points. However, when the load changes and the compensation current adds a 13th harmonic component, ... By incorporating frequency characteristics, the number of available frequency points can be expanded to four; subsequently, with the addition of the 17th harmonic, this can be further expanded to five. The increase in frequency points supports finer piecewise inversion, causing the impedance profile along the feeder to gradually become denser over time.

[0044] Impedance identification is completed without injecting additional disturbances into the power grid, avoiding secondary pollution of power quality caused by independent detection signals. This ensures that the excitation source used for impedance identification is completely consistent with the mitigation actions; the more sufficient the compensation, the richer the identification excitation, achieving a synergistic effect between mitigation and identification.

[0045] The inversion solution for the equivalent impedance of each segment specifically includes: Each segment along the feeder is equivalent to a two-port network containing series impedance and ground admittance. The cascade relationship of each two-port network is used to establish the correspondence between the equivalent impedance of each segment and the input impedance of the grid connection point.

[0046] Calculate the input impedance values ​​at each frequency component based on the corresponding relationship, compare the calculated values ​​with the measured values ​​of the frequency characteristics of the input impedance, and adjust the equivalent impedance of each segment according to the deviation obtained from the comparison until the deviation meets the set error condition.

[0047] The difference in equivalent impedance between adjacent segments is limited to a range determined by the feeder line type and segment length.

[0048] like Figure 3The diagram shows the equivalent circuit diagram of the feeder segment, which consists of three core parts from left to right: the converter grid connection point, the intermediate cascaded network, and the system equivalent power supply terminal. The observation port on the left... This represents the bus voltage at the grid connection point (PCC) where the converter is connected to the power grid. This represents the initial line current injected into the feeder from the grid connection point, i.e., the compensation current generated by the converter, which serves as the initial excitation source for subsequent segment-by-segment calculations. The intermediate cascaded network constitutes the equivalent model of the feeder segments, discretizing the long-distance weak feeder line into N cascaded electrical segments. Each segment is modeled as a two-port network containing series and parallel branches, where the series impedance... This represents the equivalent series resistance and inductive reactance of each feeder segment; when the compensation current flows through this section, a series voltage drop along the line will occur; admittance to ground. This represents the ground leakage effect caused by distributed capacitance, etc. If other loads are connected to the node, they can also be equivalently connected here. The currents of each level of the line are marked in the diagram. It is clearly shown that the ground admission branch is affected. Due to the shunting effect, the current in the main feeder changes after flowing through each node. This intuitively explains the physical mechanism by which the maximum voltage drop (target node) may occur in the middle of the line rather than at the end. (Right-side terminal) This represents the equivalent internal AC voltage source of the upstream main transformer or the high-voltage grid side. This represents the internal impedance of the equivalent power source of the upstream system.

[0049] A two-port network refers to the electrical model of each feeder segment, which includes a series impedance parameter Z. seg (Describes the series voltage drop along the line) and a ground admittance parameter Y seg (Describes leakage to ground along the line); cascading relationship refers to the overall transmission relationship formed segment by segment starting from the system's equivalent power source, which can be represented by multiplying the ABCD parameter matrices: (4) In the formula, A, B, C, and D are the transmission parameters of the total two-port network obtained after cascading each segment of the two-port network; The series impedance of the kth segment is expressed in ohms (Ω). is the ground admittance of the k-th segment, in Siemens (S); N is the total number of segments; the multiplication symbol indicates that the transmission parameter matrices from the 1st segment to the Nth segment are multiplied sequentially.

[0050] Input impedance at grid connection point It can be obtained from the cascade parameters: (5) In the formula, The input impedance is the line-of-sight impedance from the grid connection point to the system side, expressed in ohms (Ω). The internal impedance of the system's equivalent power source is expressed in ohms (Ω).

[0051] Inversion solution is given a set of measured values , inversely find each and To minimize the deviation function shown in the following equation: (6) In the formula, It is the deviation function. The input impedance is calculated based on the current segmented impedance and the corresponding relationship, and the unit is ohms (Ω). For frequency The measured input impedance is expressed in ohms (Ω); |·| represents the modulus of the complex number; M is the number of frequency points involved in the inversion. The inversion iteration continues until J does not exceed the set error threshold J. th In this embodiment, J terminates at a certain time. th One can take 1% of the sum of the squares of the measured impedance moduli.

[0052] The limitation on the impedance difference between adjacent segments is determined by the product of the feeder line type (e.g., the impedance per unit length of LGJ-185 conductor at 50Hz is approximately 0.17 + j0.40Ω / km) and the segment length (e.g., 500 meters per segment), with a typical allowable upper limit of 30% of the reference value within this range.

[0053] Consider a specific scenario: Assume a 5-segment feeder line, each using the LGJ-185 model and 500 meters long. The equivalent impedance reference value for each segment under fundamental frequency is 0.085 + j0.20Ω, and the maximum allowable deviation of adjacent segments from the reference is 0.026 + j0.06Ω. During the inversion iteration, if the impedance difference between two adjacent segments exceeds this range in any iteration, a constraint is imposed on this difference during the iteration process to maintain the physical rationality of the iteration result. Let J... th If we take 0.5Ω², we iterate until J equals 0.42Ω², which satisfies the error condition, and the inversion ends.

[0054] By using two-port cascade modeling, the inversion results have a clear physical correspondence rather than being purely mathematical fitting, thus improving the reliability of the inversion results. By physically constraining the impedance difference between adjacent segments, non-physical jumps that may occur in the inversion results are suppressed, thus improving the stability of the solution.

[0055] The compensation current is decomposed into several current components, and the voltage drop caused by the unit change of each current component on the first feeder segment is determined. The sensitivity matrix is ​​constructed using these components as elements.

[0056] The compensation current is decomposed into several current components according to harmonic order and sequence component. Each current component corresponds to a combination of a harmonic order and a sequence component.

[0057] For each current component, based on the equivalent impedance of each segment of the first feeder segment in the impedance profile along the feeder line at the harmonic order corresponding to the current component, the voltage drop caused by the unit amplitude change and unit phase change of the current component on the first feeder segment are determined.

[0058] A sensitivity matrix is ​​constructed using the voltage drop change corresponding to each current component as elements, so that each element in the sensitivity matrix corresponds to a current component, and the element represents the influence relationship between the change of the corresponding current component and the voltage drop of the first feeder segment.

[0059] Here, the current component refers to each component obtained by cross-decomposing the compensation current according to harmonic order and sequence component. Each component is jointly labeled by a harmonic order h and a sequence component s, denoted as hh. Where h takes the 1st, 5th, 7th, 11th, and 13th harmonic orders, and s takes one of the positive, negative, or zero sequence. The voltage drop of the first feeder segment is the sum of the voltage drop phasors generated by each current component on that segment. The segment impedance experienced by the h-th harmonic component should be taken as the value Z of the equivalent impedance of that segment at the h-th harmonic frequency. k (h), its relationship with the fundamental impedance is as follows: the resistive component changes little with frequency, and the reactive component is approximately proportional to frequency, i.e., Z k The reactance of (h) is approximately h times the fundamental reactance. The equivalent impedance Z of each segment at the h-th harmonic frequency is... k（h） The reactance of the cable increases proportionally to the fundamental reactance with respect to each harmonic order; that is, the reactance at the h-th harmonic is approximately h times the fundamental reactance. The resistance, when the skin effect is ignored, can be approximated as the same as the fundamental resistance. In implementations where the skin effect needs to be considered, the resistance can be increased relative to the fundamental resistance by a factor of half the harmonic order; that is, the resistance at the h-th harmonic is approximately h times the fundamental resistance. In scenarios where the cable cross-section is small and the skin effect is insignificant, using the fundamental value for the resistance is sufficient to meet the accuracy requirements.

[0060] In the sensitivity matrix, each element represents the change in the voltage drop phasor caused by a unit change in the complex phasor of the corresponding current component on the first feeder segment. Because Since the phasor is a complex number, its change encompasses both amplitude and phase changes. Therefore, this sensitivity is expressed in complex form, which is the rate of change of the voltage drop phasor of the first feeder segment with respect to the phasor of this current component. The sensitivity of the (h, s)th current component to the voltage drop of the first feeder segment is determined by the following formula: (7) In the formula, Let be the sensitivity corresponding to the (h, s)th current component, which is a complex number and the unit is ohms (Ω). This is the voltage drop phasor for the first feeder segment, in volts (V). The complex phasor of the current component is calibrated as h-th harmonic and s-sequence component, and the unit is ampere (A). The equivalent impedance of the k-th segment within the first feeder segment at the h-th harmonic frequency is expressed in ohms (Ω); K is the number of segments contained in the first feeder segment. This sensitivity is a complex number, which, as the ratio of complex phasors, uniformly characterizes the voltage drop phasor change caused by the change in the current component phasor: the magnitude change of the current component phasor causes the voltage drop phasor to change in the same direction along the component's direction, and the phase change of the current component phasor causes the voltage drop phasor to deflect in direction. Both are characterized by this complex sensitivity, eliminating the need to define separate sensitivities for amplitude and phase.

[0061] It should be noted that the sensitivity described in this invention is uniformly defined in complex form, that is, the rate of change of the voltage drop phasor of the first feeder segment with respect to the complex phasor of the current component, without specifying unit changes for amplitude and phase respectively; the unit amplitude change and unit phase change mentioned above are merely a popular explanation of the physical meaning of this complex sensitivity, and in implementation, they are calculated together as a ratio of complex phasors. In the embodiment of obtaining sensitivity by numerical difference, the applied trial change can be taken as 1% to 5% of the current component's current amplitude.

[0062] The sensitivity of each current component is arranged according to the combination of its harmonics and order components to form a sensitivity matrix. The voltage drop at the target node (end of the first feeder segment) and each current component satisfy the following relationship: (8) In the formula, The voltage drop phasor of the target node, in volts (V). The sensitivity matrix is ​​determined by equation (7), and is a complex number with the unit ohms (Ω). It is a column vector composed of the magnitude and phase of each current component, and the unit is ampere (A).

[0063] The redistribution current parameters are obtained by minimizing the voltage drop at the target node and solving the sensitivity matrix.

[0064] The combined compensation amount corresponding to the compensation current at the grid connection point for controlling voltage deviation, harmonic distortion and three-phase imbalance is determined. The combined compensation amount remains unchanged before and after redistribution, which is used as a constraint condition for solving the redistribution current parameters.

[0065] Based on the sensitivity matrix, the correspondence between the total voltage drop at the target node and the amplitude and phase of each current component is determined.

[0066] Under constraints, the amplitude and phase of each current component are solved, and the obtained amplitude and phase of each current component are used as the redistribution current parameters.

[0067] When there are no amplitude and phase adjustment values ​​that satisfy the constraints, solve for a set of amplitude and phase adjustment values ​​to make the total voltage drop as close as possible to the voltage drop limit, and use this set of amplitude and phase adjustment values ​​as the redistribution current parameters.

[0068] The combined compensation amount of the compensation current at the grid connection point is determined by the phasor sum of each current component. The voltage drop generated by each current component on the first feeder segment depends not only on the magnitude of the component but also on its corresponding harmonic order. Since the equivalent impedance of each segment of the first feeder segment varies under different harmonic orders, the same phasor sum can be constructed from different harmonic-sequence component combinations, and these different combinations result in different voltage drop phasors on the first feeder segment. This degree of freedom allows, while maintaining the combined compensation amount (i.e., the grid connection point mitigation effect remains unchanged), the amplitude and phase distribution of each current component at different harmonic and sequence orders can be altered to change the combined voltage drop phasor at the target node. This causes the combined results to tend to cancel each other out, achieving the goal of suppressing the voltage drop at the target node. Therefore, the combined compensation amount and the target node voltage drop become two independent adjustable dimensions: the former ensures grid connection point mitigation, while the latter is independently adjustable through spectrum redistribution.

[0069] The composite compensation amount refers to the total component of the compensation current that plays a role in improving various power quality indicators at the grid connection point, i.e., the phasor sum of each current component at the grid connection point. The composite compensation amount remains unchanged before and after redistribution, meaning that the amplitude and phase of each current component are adjustable, but their sum at the grid connection point is constrained.

[0070] (9) In the formula, These are the phasors of each current component before redistribution. Let be the phasors of each current component after redistribution, all in amperes (A); summation iterates through all harmonic orders and sequence component combinations. This formula indicates that the sum of the phasors of each current component at the grid connection point after redistribution is equal to the sum of the phasors of each current component at the grid connection point before redistribution. That is, redistribution does not change the combined compensation amount of the compensation current at the grid connection point, thereby ensuring that the mitigation effect on voltage deviation, harmonic distortion, and three-phase imbalance at the grid connection point is not changed by redistribution.

[0071] The objective is to ensure that the voltage drop magnitude at the target node meets the constraint condition, i.e., the magnitude of the voltage drop at the target node obtained from equation (8) does not exceed the set voltage drop limit: (10) In the formula, This is the sensitivity matrix; This is a column vector formed by the redistributed current components, in amperes (A). To set the voltage drop limit, the unit is volts (V), and |·| represents the phasor magnitude. Solve for the current component column vector satisfying equation (10) under the equality constraints of equation (9). The amplitude and phase of each component obtained are the redistribution current parameters.

[0072] The set voltage sag limit is the maximum allowable voltage sag at the target node, determined based on the target node's tolerance range for voltage deviation; in practice, it can be taken as 3% to 5% of the target node's rated phase voltage. The smaller this limit value, the more stringent the requirement for suppressing voltage sag at the target node.

[0073] When there is no solution satisfying equation (10), that is, when the target node voltage drop cannot be reduced to within the limit under the constraint of constant synthetic compensation, the limit is relaxed. Solving under the relaxed limits yields a solution that reduces the target node voltage drop to the lowest achievable level under these constraints. Specifically, when the target node voltage drop cannot be reduced to within the initial limits due to limitations in system physical capabilities, under the constraint of constant synthetic compensation, a dynamic limit relaxation mechanism is triggered. The controller incrementally increments the current voltage sag limit in a preset, minute voltage step, and re-optimizes the solution after each increment until a valid solution satisfying all current constraints is found for the first time, at which point the iteration terminates. The parameters of each current component obtained in this critical state are the final redistribution current parameters. This ensures that, while maintaining the synthetic compensation amount unchanged, the voltage sag of the target node is suppressed to the absolute minimum level achievable under the current physical conditions.

[0074] In terms of solution implementation, according to equation (8), the voltage drop at the target node is a linear mapping between the sensitivity matrix and the phasors of each current component. Therefore, under the equality constraint of the unchanged synthetic compensation amount shown in equation (9), the phasors of each current component that satisfy the constraint condition shown in equation (10) for the voltage drop at the target node can be obtained by linear solution with equality constraints. For example, the Lagrange multiplier method can be used to incorporate the synthetic compensation amount constraint into the solution, or the constraint can be substituted into the elimination and the remaining degrees of freedom can be solved. When the number of current components is greater than the degrees of freedom occupied by the synthetic compensation amount constraint, there are multiple sets of solutions for the current component phasors that satisfy the constraints. The set that minimizes the voltage drop at the target node can be selected from among them.

[0075] Based on the sensitivity matrix and the redistribution current parameters, the expected voltage drop of the target node is calculated, and the compensation current is redistributed and injected into the feeder using the redistribution current parameters.

[0076] The expected voltage sag refers to the target node voltage sag calculated according to equation (8) based on the currently used sensitivity matrix and the solved redistribution current parameters before redistribution is executed, denoted as... It is the voltage drop that should have occurred at the target node after redistribution, calculated based on the current sensitivity matrix, and is calculated as follows: (11) In the formula, The expected voltage drop phasor for the target node, in volts (V). This is the sensitivity matrix currently in use; This is a column vector consisting of the current components corresponding to the redistribution current parameters, in amperes (A).

[0077] The execution of redistribution refers to generating a new compensation current reference based on the amplitude and phase of each current component determined by the redistribution current parameters, and injecting it into the feeder through the converter control loop; since the phasor sum of each current component remains equal to the combined compensation amount, the treatment effect at the grid connection point is not affected.

[0078] After the redistribution is executed, the actual voltage drop generated by the target node is measured. The actual voltage drop is compared with the expected voltage drop to obtain the voltage drop deviation. When the voltage drop deviation is not greater than the set deviation threshold, the sensitivity matrix is ​​corrected according to the voltage drop deviation. When the voltage drop deviation is greater than the set deviation threshold, the impedance profile along the feeder is updated and the target node is re-determined. When the target node before and after the update is inconsistent, the sensitivity matrix is ​​reconstructed according to the updated target node, and redistribution is performed again until the voltage drop amplitude of the target node is not greater than the set amplitude threshold.

[0079] Wherein, actual voltage dip refers to the actual voltage dip generated at the target node after the redistribution is executed, denoted as Since the target node is the upstream intermediate bus node without a voltage sensor, its actual voltage drop cannot be directly measured. Instead, after the redistribution is executed, the impedance profile along the feeder is re-inverted based on the new round of measurements at the grid connection point, and then calculated according to formula (2) based on the new profile and the compensation current after redistribution. The voltage drop deviation refers to the difference between the actual voltage drop and the expected voltage drop, denoted as . : (12) In the formula, Let be the voltage drop deviation phasor of the t-th correction process, in volts (V). The actual voltage drop phasor of the target node after the t-th round of redistribution is expressed in volts (V). Let t be the expected voltage drop phasor calculated using the current sensitivity matrix in the t-th round, in volts (V), where t is the round number.

[0080] Voltage sag deviation reflects the degree of mismatch between the sensitivity matrix and the impedance profile relative to reality, and its magnitude causes current splitting between the two paths: when | |Not greater than the set deviation threshold e th When the mismatch is determined to be minor and the problem lies at the sensitivity matrix level, the sensitivity matrix correction route is initiated; when | |Greater than e th When the impedance profile layer with more severe mismatch is determined to be located further upstream, the profile update and target node reselection route is initiated. A deviation threshold e is set. th Three percent of the rated phase voltage of the target node can be taken. After processing both routes, redistribution is performed again using the corrected sensitivity matrix, and it is determined whether the voltage drop amplitude of the target node is not greater than the set amplitude threshold U. th (3% of the rated phase voltage can be taken), if it is not reached, continue the iteration.

[0081] When the voltage drop deviation is not greater than the set deviation threshold, the sensitivity matrix is ​​corrected based on the sensitivity deviation, specifically including: The deviation of the sensitivity matrix relative to its true value is calculated based on the amplitude and phase of each current component in the redistributed current parameters during the correction process, as well as the voltage drop deviation. This deviation is denoted as the sensitivity deviation.

[0082] The decrease in voltage drop deviation after the current correction relative to the voltage drop deviation after the previous correction is denoted as the deviation reduction. Based on the deviation reduction and the pre-established correspondence between the deviation reduction and the correction coefficient, the correction coefficient for this correction process is determined. The correction coefficient for the first correction process is taken as a preset initial value.

[0083] In the correspondence, the larger the reduction, the larger the corresponding correction coefficient.

[0084] The corresponding elements of the sensitivity matrix are corrected using correction coefficients and sensitivity deviations to obtain the corrected sensitivity matrix.

[0085] Based on the corrected sensitivity matrix, the voltage drop deviation for this correction process is determined. The voltage drop deviation for this correction process is compared with that for the previous correction process to obtain the reduction amount. This reduction amount is used to determine the correction coefficient for the next correction process.

[0086] If the voltage drop deviation does not decrease after a set number of consecutive correction cycles, the termination condition is met, and the sensitivity deviation is determined to exceed the correctable range of the sensitivity matrix, thus terminating the correction process.

[0087] Sensitivity deviation refers to the deviation of the sensitivity matrix from its true value, denoted as . The reasoning behind this is based on the following relationship: the voltage sag at the target node is caused by the sensitivity matrix acting on the current component; therefore, the voltage sag deviation is equal to the sensitivity deviation acting on the known current component, i.e.: (13) In the formula, Let be the voltage drop deviation phasor for the t-th round, in volts (V). The sensitivity deviation is the deviation of the sensitivity matrix in round t from its true value. This is the column vector of current components corresponding to the redistributed current parameters in round t, in amperes (A). Since only one set is provided per single correction... and Not enough to uniquely determine Therefore, we need to collect data from the most recent several correction processes and solve the system of equations simultaneously. Let's assume we collect data from the most recent P correction processes. The following equation is derived by inverse least squares: (14) In the formula, The sensitivity deviation is the deviation of the sensitivity matrix in round t from its true value. It is a matrix composed of the voltage drop deviations of the t-th round and the preceding P rounds arranged in columns; The column vector of current components corresponding to the P-wheel. The matrix is ​​formed by arranging columns; the superscript + indicates the generalized inverse (pseudo-inverse) of the matrix; P is the number of correction processes involved in the inverse calculation, which should not be less than the number of current components n.

[0088] The sensitivity matrix is ​​corrected according to the correction factor and the sensitivity deviation: (15) In the formula, This is the corrected sensitivity matrix for the next round; This represents the sensitivity matrix before the t-th round of correction. is the correction coefficient for round t, with a value greater than 0 and less than 1; The deviation of the sensitivity matrix in round t from its true value. Since there is no previous deviation to compare, the correction coefficient in the first correction process is set to a preset initial value, which can be between 0.2 and 0.4. In this embodiment, it is set to 0.3. If the preset initial value is too large, it will easily cause oscillation in the first correction round, and if it is too small, the first round will converge slowly.

[0089] The correction factor α is determined by a pre-established correspondence based on the reduction in deviation. The reduction in deviation refers to the decrease in the magnitude of the voltage sag deviation after this correction compared to the previous one, denoted as d: (16) In the formula, This represents the decrease in deviation during round t, expressed in volts (V). For the tth Voltage drop deviation phasor of round 1; Let be the voltage drop deviation phasor of the t-th round; |·| represents the magnitude of the phasor.

[0090] In practical engineering implementation, the pre-established correspondence between the deviation reduction and the correction coefficient is a physical constraint rule constructed through offline transient simulation fitting of the feeder impedance model or statistical analysis of historical operating data under multiple operating conditions. Specifically, this correspondence can be configured as an interval lookup table rule for continuous mathematical function mapping. When using function mapping, the correction coefficient is set as a linear proportional function of the deviation reduction, thereby ensuring that the system has a fast transient response and convergence speed when the deviation reduction is significant. When using interval lookup table rules, based on the operational accuracy boundary of the underlying physical controller, the effective physical distribution range of the deviation reduction is divided into several continuous numerical intervals, and a fixed correction coefficient value with a step-like monotonically increasing value is configured for each interval. Regardless of the mapping mechanism used, the control principle that the larger the reduction, the larger the corresponding correction coefficient is strictly followed, thereby effectively avoiding discrete oscillations of the target node voltage at the end of the optimization process while ensuring the control system's rapid adaptive tracking of the time-varying impedance characteristics of long-distance weak-feed grids.

[0091] When the voltage drop deviation exceeds the set deviation threshold, the impedance profile along the feeder is updated and the target node is redefined. The specific process includes: When the voltage drop deviation exceeds the set deviation threshold, the residual of each segment in the inversion is determined.

[0092] The residual is the deviation between the calculated input impedance value and the measured input impedance value of each segment.

[0093] The measured input impedance is the ratio of the voltage response of each frequency component at the grid connection point to the corresponding compensation current.

[0094] The input impedance is calculated based on the equivalent impedance of each segment obtained through inversion, and the corresponding relationship between the equivalent impedance of each segment and the input impedance at the grid connection point.

[0095] The equivalent impedance of each segment is re-inverted to obtain the updated impedance profile along the feeder.

[0096] Based on the updated impedance profile along the feeder, the voltage drop along the feeder caused by the compensation current in each segment is re-determined, and the node with the largest voltage drop is re-acquired.

[0097] When the node with the largest voltage drop that is reacquired is inconsistent with the current target node, the node with the largest voltage drop that is reacquired is taken as the updated target node, and the feeder segment between the grid connection point and the updated target node is updated as the first feeder segment.

[0098] When there is consistency, keep the target node unchanged.

[0099] The residual refers to the deviation between the calculated input impedance value obtained by substituting the equivalent impedance of a certain segment into the correspondence between the equivalent impedance of each segment and the input impedance at the grid connection point, and the measured input impedance value at the corresponding frequency of that segment. It is used to measure the accuracy of the segment inversion result. The residual of the k-th segment is determined by the following formula: (17) In the formula, The residual for the k-th segment is expressed in ohms (Ω). The input impedance is calculated based on the segmented equivalent impedance obtained from the inversion and by formulas (4) and (5), with the unit being ohms (Ω). For the corresponding frequency The measured input impedance is expressed in ohms (Ω); |·| represents the modulus of the complex number. The measured input impedance is obtained from equation (3), which is the ratio of the voltage response of the frequency component at the grid connection point to the corresponding compensation current.

[0100] During the re-inversion, for residuals greater than the set residual threshold r th The segment is further divided into multiple sub-segments to improve the inversion precision of the segment; when the input impedance values ​​on the existing frequency components are insufficient to determine the equivalent impedance of each segment, excitation frequency components are added within the frequency range contained in the compensation current. In this embodiment, the residual threshold r th The average impedance modulus of each segment can be taken as 10%. Based on the further segmented segments and the supplemented frequency components, the updated impedance profile along the feeder is obtained by re-inversion. According to this profile, the voltage drop of the compensation current along the feeder is recalculated according to Equation (2), and the node with the largest voltage drop is taken as the newly determined target node. If the node is inconsistent with the current target node, the target node and the first feeder segment are updated; if they are consistent, they remain unchanged.

[0101] It also includes responding to changes in the compensation current according to the setting period, the specific process of which includes: Set a tuning period. Within each tuning period, the compensation current at the grid connection point and its frequency components are used as the basis for determining the frequency characteristics of the input impedance, the impedance profile along the feeder, the target node, and the sensitivity matrix within that tuning period.

[0102] The degree of change in the frequency characteristics of the input impedance is taken as the sum of the absolute values ​​of the differences in the frequency components of the input impedance determined by two adjacent tuning periods.

[0103] When the degree of change exceeds the set change threshold, the tuning cycle is shortened.

[0104] When the degree of change is not greater than the set change threshold, extend the tuning period.

[0105] Between two adjacent tuning cycles, when a new frequency component appears in the compensation current, the new frequency component is added to the value point of the frequency characteristic of the input impedance.

[0106] The tuning period refers to the time interval at which the method re-determines the input impedance frequency characteristics, the impedance profile along the feeder, the target node, and the sensitivity matrix. Within each tuning period, the compensation current and its frequency components at the grid connection point are used as the basis for determining these quantities, thus periodically updating them as the compensation current changes. The degree of change in the input impedance frequency characteristics is characterized by the sum of the absolute values ​​of the differences between the input impedance frequency characteristics determined in two adjacent tuning periods at each frequency component. Specifically, the values ​​of the input impedance determined in the current tuning period at each frequency component are subtracted one by one from the values ​​determined in the previous tuning period at the corresponding frequency components. The magnitudes of these differences are then added together, and the sum represents the degree of change in the input impedance frequency characteristics. A greater degree of change indicates a faster change in the feeder impedance with operating conditions.

[0107] The adjustment of the tuning period is performed between a preset upper and lower limit. Each adjustment shortens or lengthens the period by a set step size. The lower limit can be 10 seconds, the upper limit can be 300 seconds, and the adjustment step size can be 30 seconds. These values ​​can be adjusted according to the rate of change of the on-site operating conditions. A change threshold is set to determine the degree of change in the input impedance frequency characteristics on which the shortening or lengthening of the tuning period is based. The unit is ohms, and it can be taken as 5% of the sum of the magnitudes of the input impedance frequency characteristics of the previous tuning period. If the change is greater than this threshold, the tuning period is shortened; if it is not greater than this threshold, the tuning period is lengthened.

[0108] like Figure 4The figure shows a time-domain simulation comparison waveform of the target node voltage sag and the synthetic compensation. In this embodiment, a simulation model containing converter control logic and long-distance feeder distribution parameters is constructed based on discrete time steps. During the simulation environment construction phase, the series impedance and ground admittance reference parameters of each feeder segment are preset according to the physical characteristics of the real weak-feed power grid, and nonlinear disturbance loads are set at the corresponding feeder nodes. The sampling frequency and control execution cycle of the converter controller model are set with reference to the computing power boundary of industrial-grade digital signal processors in a real hardware environment to ensure that the discrete-time series data output by the simulation system can accurately map the underlying digital physical control process.

[0109] It should be noted that the simulation model in this embodiment is built on the Discrete-Time Transient Simulation Platform (MATLAB). To eliminate the idealization distortion caused by the integration of continuous systems, the simulation solver adopts a fixed-step discrete solver algorithm. The basic time simulation operation step size is strictly preset to 10μs. This extremely small step size ensures that within a global simulation cycle of 1.5s, it can capture the microscopic waveform distortion caused by high-frequency harmonic transient response and control delay with high fidelity, generating a total of 150,000 discrete data operation nodes, providing underlying data support for high-resolution waveform output.

[0110] The system's equivalent feeder is set as a 2.5km long weak current feeder line, spatially divided into 5 discrete two-port network segments, each 500m long. The cable physical model is equivalent to the electrical parameters of standard LGJ-185 steel-cored aluminum stranded wire: under a fundamental frequency of 50Hz, the equivalent series impedance reference parameter for a single feeder segment is preset to 0.085 + j0.20Ω. Considering the distributed capacitance to ground and leakage effect of long-distance lines, the ground admittance reference parameter for each segment is preset to j1.57 × 10⁻⁶. -6 S. Under this distributed parameter architecture, the equivalent reactance of higher harmonics will be amplified strictly according to the frequency, objectively constructing the standing wave effect and attenuation boundary in long cable transmission.

[0111] To reproduce the mid-section current shunting effect along the feeder, an uncontrolled rectified nonlinear load with a base power of 50kW was centrally connected at the third node of the feeder topology (i.e., the target node, 1.5km from the grid connection point). This load naturally generated 5th (250Hz) and 7th (350Hz) harmonic currents with an amplitude ratio of 30%. Furthermore, to verify the dynamic robustness of the redistribution algorithm under complex operating conditions (i.e., the oscillation interval after 1.0s in the waveform diagram), a low-frequency envelope amplitude-modulated disturbance with a frequency of 10Hz and a modulation depth of 15% was superimposed onto the nonlinear load at 1.0s, thus physically realistically reproducing the extreme challenges of periodic load fluctuations under harsh weak grid conditions.

[0112] The hardware environment of the grid-connected converter is built according to the standard specifications of industrial-grade high-capacity active power quality management converters. The main circuit adopts a three-phase two-level voltage-source converter topology, with a system rated compensation capacity set at 100kVA and the DC-side energy storage bus voltage maintained at 800V. The grid-connected side adopts an LCL-type passive low-pass filter structure, with the inverter-side inductor preset to 1.2mH, the grid-side inductor preset to 0.4mH, and the filter capacitor preset to 20μF. This combination of hardware parameters defines the highest frequency band and maximum current ramp-up rate of the converter's physical output.

[0113] The timing logic of the control system fully conforms to the physical boundaries of an industrial-grade 32-bit floating-point digital signal processor (TMS320F28335). The analog-to-digital converter (ADC) is preset to a sampling frequency of 20kHz for the grid connection point voltage and compensation current. The core interrupt execution cycle of the entire controller is strictly locked to 100μs.

[0114] When acquiring the dataset for the aforementioned time-domain response comparison charts, the system simulation runtime was set to cover three continuous dynamic evolution stages: startup transient, optimal control intervention, and steady-state high-frequency disturbance. The simulation test started at initial time zero and maintained a traditional fixed-proportion compensation strategy for the first 0.5 seconds to establish the initial high-level reference state for voltage drop. At 0.5 seconds, the system triggered the dynamic spectrum redistribution control mechanism of this embodiment, and the controller began performing dimensionality reduction optimization calculations based on the real-time inverted sensitivity matrix. To further verify the dynamic robustness of the control algorithm, the system actively injected high-frequency harmonic disturbance conditions into the power grid model at 1 second of simulation runtime.

[0115] The chart on the left details the time-domain evolution of the actual voltage drop at the target node. The horizontal axis of this coordinate system represents simulation time, and the vertical axis represents the voltage drop amplitude, in volts. The legend includes a solid line representing the optimized control of this invention, a dashed line representing traditional fixed compensation, and a dotted line representing the system safety limit. During the system startup and initial operation phase from 0.0 to 0.5 seconds, the voltage drop rapidly climbs and stabilizes at a physical high of approximately 82V, significantly deviating from the target safety baseline set at approximately 52V. When the system reaches the 0.5-second node, the system precisely triggers the dynamic spectrum redistribution mechanism. As shown in the enlarged view in the middle, the solid line representing this invention smoothly and without overshoot drops to near the 52V safety limit within a very short transient adjustment time, achieving a close fit; while the dashed line, which does not employ this method, remains unaffected, consistently maintaining an excessive 82V. Between 1.0 and 1.5 seconds, the system faces continuous high-frequency operating disturbances. The rightmost magnified view further confirms that, although the waveform exhibits obvious sinusoidal envelope oscillations, the voltage drop centerline controlled by this invention is still strongly clamped at the target limit, achieving spatial-level precise suppression of voltage deviation at remote nodes.

[0116] The horizontal axis of the right-hand chart is synchronized with the left-hand chart, while the vertical axis represents the total composite phasor magnitude of each current component injected into the converter feeder at the grid connection point, in amperes. In the legend, the response curve after redistribution is represented by a thin solid line, while the comparison curve without redistribution is represented by a thick dashed line. Throughout the entire simulation period from 0.0 to 1.5 seconds, the solid and dashed lines show absolute overlap across all data intervals, including the three microscopic magnified views. Specifically, at 0.5 seconds, at the same moment the voltage drop on the left side causes a sharp downward adjustment, the magnified view on the right shows the total current magnitude remaining stable between approximately 73A and 74A, without any local abrupt changes or power oscillations caused by remote voltage control. Even in the high-frequency disturbance steady state after 1.0 seconds, the current magnitude oscillation trajectories of both remain perfectly aligned.

[0117] By combining the joint simulation data from both sides, this set of waveforms provides a self-consistent closed-loop proof at the underlying physical mechanism. The continuously overlapping current trajectory on the right side confirms that the system does not achieve voltage regulation in the time domain by simply and crudely increasing the overall output capacity of the converter; the total local power quality management of the converter at the grid connection point is strictly locked as a rigid constraint. Under these extremely constrained boundary conditions, the successful suppression of the voltage drop on the left side indicates the successful realization of independent dimensionality reduction control of the spatial voltage drop along the feeder.

[0118] Please see Figure 2 As shown, a low-voltage power grid stabilization system based on cable impedance matching specifically includes: The target node determination module is used to obtain the compensation current at the grid connection point and the impedance profile along the feeder between the grid connection point and the system equivalent power source. Based on the impedance profile along the feeder and the compensation current, it determines the voltage drop distributed along the feeder when the compensation current flows through the feeder, and takes the node with the largest voltage drop as the target node. The feeder segment between the grid connection point and the target node is recorded as the first feeder segment.

[0119] The sensitivity matrix construction module is used to decompose the compensation current into several current components, determine the voltage drop change caused by the unit change of each current component on the first feeder segment, and construct the sensitivity matrix using these components as elements.

[0120] The redistribution parameter solving module is used to solve for the redistribution current parameters based on the sensitivity matrix, with the goal of minimizing the voltage drop amplitude at the target node.

[0121] The redistribution injection execution module is used to calculate the expected voltage drop of the target node based on the sensitivity matrix and the redistribution current parameters, and to redistribute the compensation current and inject it into the feeder using the redistribution current parameters.

[0122] The deviation assessment and iterative update module is used to measure the actual voltage drop generated by the target node after the redistribution is executed, compare the actual voltage drop with the expected voltage drop to obtain the voltage drop deviation, and correct the sensitivity matrix based on the voltage drop deviation when the voltage drop deviation is not greater than the set deviation threshold. When the voltage drop deviation is greater than the set deviation threshold, the impedance profile along the feeder is updated and the target node is re-determined. When the target node before and after the update is inconsistent, the sensitivity matrix is ​​reconstructed based on the updated target node and redistribution is performed again until the voltage drop amplitude of the target node is not greater than the set amplitude threshold.

[0123] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the above functions can be divided into different functional modules to complete all or part of the functions described above.

[0124] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between devices or units through some interfaces, and may be electrical, mechanical, or other forms.

[0125] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A method for voltage stabilization in a weak power grid based on cable impedance matching, characterized in that, include: Obtain the compensation current at the grid connection point and the impedance profile along the feeder between the grid connection point and the system equivalent power source. Based on the impedance profile along the feeder and the compensation current, determine the voltage drop distributed along the feeder when the compensation current flows through the feeder, and take the node with the largest voltage drop as the target node. The feeder segment between the grid connection point and the target node is recorded as the first feeder segment. The compensation current is decomposed into several current components, and the voltage drop change caused by the unit change of each current component on the first feeder segment is determined. The sensitivity matrix is ​​constructed using these components as elements. The redistribution current parameters are obtained by minimizing the voltage drop amplitude at the target node based on the sensitivity matrix. Based on the sensitivity matrix and the redistribution current parameters, the expected voltage drop of the target node is calculated, and the compensation current is redistributed and injected into the feeder using the redistribution current parameters. After the redistribution is executed, the actual voltage drop generated by the target node is measured. The actual voltage drop is compared with the expected voltage drop to obtain the voltage drop deviation. When the voltage drop deviation is not greater than the set deviation threshold, the sensitivity matrix is ​​corrected according to the voltage drop deviation. When the voltage drop deviation is greater than the set deviation threshold, the impedance profile along the feeder is updated and the target node is re-determined. When the target node before and after the update is inconsistent, the sensitivity matrix is ​​reconstructed according to the updated target node, and the redistribution is executed again until the voltage drop amplitude of the target node is not greater than the set amplitude threshold.

2. The weak grid voltage stabilization method based on cable impedance matching as described in claim 1, characterized in that: The acquisition of the compensation current at the grid connection point and the line impedance profile along the feeder from the grid connection point to the system equivalent power source specifically includes: Based on the frequency components contained in the compensation current and the voltage response corresponding to each frequency component at the grid connection point, the frequency characteristics of the input impedance from the grid connection point to the system side are determined. The feeder line between the grid connection point and the system equivalent power supply is divided into several segments. The equivalent impedance of each segment is solved by inversion using the frequency characteristics of the input impedance at each frequency component as constraints, and the impedance profile along the feeder line is obtained.

3. The weak grid voltage stabilization method based on cable impedance matching as described in claim 2, characterized in that: The frequency characteristics of the input impedance are determined by the frequency components contained in the compensation current and the corresponding voltage response, specifically including: The compensation current is used as the excitation for the inherent frequency components of harmonic current suppression, negative sequence and zero sequence compensation and voltage regulation, without injecting a separate detection signal into the feeder. The ratio of the change in voltage response at each frequency component at the grid connection point to the change in compensation current corresponding to that frequency component is used as the value of the input impedance at that frequency component. When the frequency component of the compensation current changes with the operating conditions, the input impedance of the newly appearing frequency component of the compensation current under the operating conditions is taken and the frequency characteristic of the input impedance is added.

4. The weak grid voltage stabilization method based on cable impedance matching as described in claim 2, characterized in that: The inversion solution for the equivalent impedance of each segment includes: Each segment along the feeder is equivalent to a two-port network containing series impedance and ground admittance. The cascade relationship of each two-port network is used to establish the correspondence between the equivalent impedance of each segment and the input impedance of the grid connection point. The input impedance is calculated at each frequency component based on the correspondence. The calculated values ​​are compared with the measured values ​​of the frequency characteristics of the input impedance. The equivalent impedance of each segment is adjusted based on the deviation obtained from the comparison until the deviation meets the set error condition. The difference in equivalent impedance between adjacent segments is limited to a range determined by the line type and segment length of the feeder.

5. The voltage stabilization method for weak power grids based on cable impedance matching as described in claim 1, characterized in that: The specific process of constructing the sensitivity matrix includes: The compensation current is decomposed into several current components according to harmonic order and sequence component, and each current component corresponds to a combination of a harmonic order and a sequence component. For each current component, based on the value of the equivalent impedance of each segment of the first feeder segment in the impedance profile along the feeder under the harmonic order corresponding to the current component, the voltage drop change caused by the unit amplitude change and unit phase change of the current component on the first feeder segment is determined. The sensitivity matrix is ​​constructed using the voltage drop change corresponding to each current component as elements, such that each element in the sensitivity matrix corresponds to a current component, and the element represents the influence relationship between the change of the corresponding current component and the voltage drop of the first feeder segment.

6. The weak grid voltage stabilization method based on cable impedance matching as described in claim 5, characterized in that: The redistribution current parameters are obtained by solving the sensitivity matrix, and the specific process includes: The composite compensation amount corresponding to the compensation current at the grid connection point for controlling voltage deviation, harmonic distortion and three-phase imbalance is determined, and the composite compensation amount remains unchanged before and after redistribution, which is used as a constraint condition for solving the redistribution current parameters. Based on the sensitivity matrix, the correspondence between the total voltage drop at the target node and the amplitude and phase of each current component is determined; Under the constraints, the amplitude and phase of each current component are solved, and the amplitude and phase of each current component obtained are used as the redistribution current parameters. When there are no amplitude adjustment amounts and phase adjustment amounts that satisfy the constraints, a set of amplitude adjustment amounts and phase adjustment amounts is solved to make the total voltage drop as close as possible to the voltage drop limit, and this set of amplitude adjustment amounts and phase adjustment amounts is used as the redistribution current parameters.

7. The weak grid voltage stabilization method based on cable impedance matching as described in claim 6, characterized in that: When the voltage drop deviation is not greater than the set deviation threshold, the sensitivity matrix is ​​corrected based on the voltage drop deviation, specifically including: The deviation of the sensitivity matrix relative to its true value is calculated based on the amplitude and phase of each current component in the redistributed current parameters during the correction process and the voltage drop deviation, and is denoted as the sensitivity deviation. The decrease in voltage drop deviation after the current correction relative to the voltage drop deviation after the previous correction is denoted as the deviation reduction. Based on the deviation reduction and the pre-established correspondence between the deviation reduction and the correction coefficient, the correction coefficient for this correction process is determined. In the aforementioned correspondence, the larger the reduction, the larger the corresponding correction coefficient. The corresponding elements of the sensitivity matrix are corrected using the correction coefficient and the sensitivity deviation to obtain the corrected sensitivity matrix; Based on the corrected sensitivity matrix, the voltage drop deviation of this correction process is determined; the voltage drop deviation of this correction process is compared with the voltage drop deviation of the previous correction process to obtain the reduction amount of this correction process, and the reduction amount of this correction process is used to determine the correction coefficient of the next correction process; If the voltage drop deviation does not decrease after a set number of consecutive correction processes, the termination condition is met, and it is determined that the sensitivity deviation exceeds the correctable range of the sensitivity matrix, thus terminating the correction process.

8. The weak grid voltage stabilization method based on cable impedance matching as described in claim 2, characterized in that: When the voltage drop deviation exceeds the set deviation threshold, the impedance profile along the feeder is updated and the target node is redefined. The specific process includes: When the voltage drop deviation is greater than the set deviation threshold, the residual of each segment in the inversion is determined; The residual is the deviation between the calculated input impedance value and the measured input impedance value of each segment; The measured value of the input impedance is the ratio of the voltage response of each frequency component at the grid connection point to the corresponding compensation current. The input impedance calculation value is the input impedance calculated based on the equivalent impedance of each segment obtained by inversion, and the correspondence between the equivalent impedance of each segment and the input impedance of the grid connection point. The equivalent impedance of each segment is re-inverted to obtain the updated impedance profile along the feeder line. Based on the updated feeder impedance profile, the voltage drop along the feeder caused by the compensation current in each segment of the feeder is re-determined, and the node with the largest voltage drop is re-acquired. When the node with the largest voltage drop that is reacquired is inconsistent with the current target node, the node with the largest voltage drop that is reacquired is used as the updated target node, and the feeder segment between the grid connection point and the updated target node is updated to the first feeder segment. When they are consistent, the target node remains unchanged.

9. The voltage stabilization method for weak power grids based on cable impedance matching as described in claim 1, characterized in that: It also includes responding to changes in the compensation current according to a set period, the specific process of which includes: A tuning period is set. Within each tuning period, the compensation current at the grid connection point and its frequency components are used as the basis for determining the frequency characteristics of the input impedance, the impedance profile along the feeder, the target node, and the sensitivity matrix within that tuning period. The sum of the absolute values ​​of the differences in the frequency characteristics of the input impedance determined by two adjacent tuning cycles at each frequency component is taken as the degree of change of the frequency characteristics of the input impedance. When the degree of change exceeds a set change threshold, the tuning period is shortened. When the degree of change is not greater than the set change threshold, the adjustment period is extended; Between two adjacent tuning cycles, when a new frequency component appears in the compensation current, the new frequency component is added to the value point of the frequency characteristic of the input impedance.

10. A low-voltage grid voltage stabilization system based on cable impedance matching, used to implement the low-voltage grid voltage stabilization method based on cable impedance matching as described in any one of claims 1 to 9, characterized in that: Specifically, it includes: The target node determination module is used to obtain the compensation current at the grid connection point and the impedance profile along the feeder between the grid connection point and the system equivalent power source. Based on the impedance profile along the feeder and the compensation current, it determines the voltage drop distributed along the feeder when the compensation current flows through the feeder, and takes the node with the largest voltage drop as the target node. The feeder segment between the grid connection point and the target node is recorded as the first feeder segment. The sensitivity matrix construction module is used to decompose the compensation current into several current components, determine the voltage drop change caused by the unit change of each current component on the first feeder segment, and construct the sensitivity matrix using these components as elements. The redistribution parameter solving module is used to solve for the redistribution current parameters based on the sensitivity matrix, with the goal of minimizing the voltage drop amplitude at the target node. The redistribution injection execution module is used to calculate the expected voltage drop of the target node based on the sensitivity matrix and the redistribution current parameters, and to redistribute the compensation current and inject it into the feeder using the redistribution current parameters. The deviation assessment and iterative update module is used to measure the actual voltage drop generated by the target node after the redistribution is executed, compare the actual voltage drop with the expected voltage drop to obtain the voltage drop deviation, and when the voltage drop deviation is not greater than a set deviation threshold, correct the sensitivity matrix according to the voltage drop deviation. When the voltage drop deviation is greater than the set deviation threshold, update the impedance profile along the feeder and redetermine the target node. When the target node before and after the update is inconsistent, reconstruct the sensitivity matrix according to the updated target node and perform the redistribution again until the voltage drop amplitude of the target node is not greater than the set amplitude threshold.

Images