Intelligent compensation box-type substation and compensation control method and system thereof

Abstract

This invention relates to the field of intelligent distribution network automation control technology, specifically to an intelligent compensation box-type substation and its compensation control method and system. The method is executed by the controller: acquiring grid operating parameters and the status of capacitor compensation branches; extracting transient distortion characteristics; if there is a sudden change in reactive power deficit and distortion characteristics exist, compensation processing is suspended; if not suspended, candidate schemes are generated and voltage changes are predicted based on equivalent short-circuit impedance; if limits are exceeded, degradation or blocking is implemented, and the target scheme is determined by combining the branch's abnormal history; the target compensation scheme is executed, and the system's equivalent short-circuit impedance and judgment threshold are updated based on steady-state actual feedback. This invention achieves transient pseudo-demand identification, reduces the risk of bus overvoltage exceeding limits, blocks repeated switching oscillations, and improves the system's adaptive correction capability and long-term operational safety under complex operating conditions.

Description

Technical Field

[0001] This invention relates to the field of intelligent power distribution network automation control technology, specifically to an intelligent compensation box-type substation and its compensation control method and system. Background Technology

[0002] Intelligent compensation prefabricated substations, as core nodes for power distribution and control in power distribution networks, are widely used in industrial parks, new energy facilities, and urban power distribution. With the development of intelligent power distribution technology, in addition to conventional transformers and distribution units, prefabricated substations typically integrate reactive power compensation modules and intelligent controllers, enabling local regulation of reactive power on the low-voltage bus side to improve power quality and maintain bus voltage stability.

[0003] Existing intelligent compensation prefabricated substation control schemes typically collect basic operating data such as voltage and current to calculate reactive power deficit and output switching commands based on a single static threshold or simple over-limit identification. While this approach is effective during periods of stable load change, in actual power grid operation, transient distortions occur when single-phase impact loads or asymmetrical disturbances affect the distribution network. During this process, drastic fluctuations in phase current and negative sequence current often generate brief abrupt changes in reactive power deficit. Most existing schemes lack the ability to effectively distinguish between transient distortion characteristics and true steady-state reactive power demand, making the compensation system prone to triggering blind and ineffective switching actions, thereby exacerbating bus voltage fluctuations.

[0004] On the other hand, prefabricated substations are often located at the end of the power grid or in high-impedance operating environments with a high proportion of renewable energy integration. The system's equivalent short-circuit impedance is often large and prone to fluctuations. Under such conditions, if static equal-capacity compensation logic is still used, directly connecting large-capacity capacitors can easily cause severe bus overvoltage. Existing technologies, when generating compensation actions, often only focus on reactive power capacity matching, lacking the ability to pre-assess and intervene in the risk of bus voltage rise after the compensation action is executed. This results in the controller being unable to detect the risk of exceeding limits before issuing commands, easily triggering underlying overvoltage protection actions, making it difficult to balance the sufficiency of reactive power compensation with the safety of bus voltage.

[0005] Furthermore, existing technologies, after detecting reactive power deficit and issuing commands, typically output actions cyclically based on the current state, lacking continuous tracking and closed-loop correction of the historical operating state of the compensation branch. Existing solutions often struggle to effectively avoid repeated over-limit tripping induced by the system environment and persistent anomalies caused by localized degradation of specific branches, easily leading to frequent switching oscillations in specific compensation branches. Simultaneously, the control parameters and judgment thresholds of existing systems are mostly fixed settings, unable to adaptively update based on actual steady-state feedback before and after actions, causing the compensation control model to gradually fail under complex and evolving power grid conditions.

[0006] Therefore, existing intelligent compensation box-type substations still have technical defects under complex operating conditions, such as difficulty in effectively distinguishing transient disturbances from actual reactive power demand, lack of pre-constraints on voltage overruns caused by compensation actions, susceptibility of specific compensation branches to frequent oscillations, and inability of control parameters to adapt to changes in operating conditions. Summary of the Invention

[0007] (i) The technical problem to be solved by the present invention is that: under complex operating conditions such as transient distortion and high impedance, existing intelligent compensation box-type substations have difficulty in effectively distinguishing between the transient reactive power deficit caused by disturbance and the actual steady-state demand. Furthermore, they lack the ability to pre-constrain the risk of bus voltage rise before compensation action and have difficulty in timely correcting the underlying control parameters based on actual steady-state feedback after action. This makes it easy for compensation action to be falsely triggered and directly induce bus overvoltage exceeding the limit, thereby causing repeated switching oscillations of specific compensation branches and the gradual failure of the control model.

[0008] (II) Technical Solution To address the aforementioned technical problems, this invention provides a compensation control method for an intelligent compensation box-type substation. The substation includes a low-voltage busbar, multiple capacitor compensation branches, and a controller. The method is characterized by the controller executing the following steps: S1. Obtain power grid operation parameters including three-phase voltage, three-phase current, reactive power calculation values ​​and state quantities characterizing the degree of three-phase imbalance, as well as the operation status of each capacitor compensation branch; S2. Extract transient distortion features based on the power grid operating parameters and determine the current reactive power deficit; if the rate of change of the current reactive power deficit exceeds the abrupt change threshold and the transient distortion features exist, then suspend the compensation processing for the current reactive power deficit. By extracting transient distortion features in step S2 and combining them with the reactive power deficit change rate for joint gating, compensation processing is actively suspended when a sudden change and distortion conditions are met. This effectively separates the transient reactive power deficit changes induced by grid disturbances from the actual steady-state reactive power demand, preventing the controller from blindly triggering invalid capacitor switching based solely on instantaneous calculation deviations, thus suppressing erroneous compensation actions under complex operating conditions from the source.

[0009] S3. If the compensation process for the current reactive power deficit is not suspended, a candidate compensation scheme is generated, and the predicted voltage change is calculated based on the system equivalent short-circuit impedance parameter, the candidate compensation capacity, and the current bus voltage. If the predicted voltage change exceeds the safety threshold, the candidate compensation scheme is downgraded or blocked. The target compensation scheme is determined from the candidate compensation scheme or the processed candidate compensation scheme, based on the abnormal operation history of each capacitor compensation branch. Predicting voltage changes based on the system's equivalent short-circuit impedance parameters and performing degradation or blocking actions when limits are exceeded allows compensation commands to be issued based on the constraints of quantitative assessment of bus overvoltage risk. This avoids forcibly activating large-capacity capacitors beyond the actual carrying capacity of the distribution network, which could directly induce bus overvoltage exceeding limits. Simultaneously, determining the target scheme by combining the abnormal operating history of the capacitor compensation branch can transform the local degradation state of the underlying hardware into anti-oscillation constraints for the top-level dispatching system. This prevents the control system from repeatedly reclosing defective branches that are prone to triggering protection actions without memory, and blocks the cyclic switching oscillations caused by specific branch anomalies.

[0010] S4. After determining the target compensation scheme, control the capacitor compensation branch corresponding to the target compensation scheme to perform the compensation action, and collect the actual voltage change and actual reactive power change after reaching steady state; based on the deviation between the actual voltage change and the predicted voltage change and the actual reactive power change, update the system equivalent short-circuit impedance parameter, the safety threshold and the sudden change threshold.

[0011] By collecting the actual changes after steady state and updating the equivalent short-circuit impedance parameters and core thresholds of the system based on the deviation between the actual and predicted values ​​in a closed loop, the actual network strength evolution of the distribution network and the error feedback of the deduction model can be transformed into the adaptive correction driving force of the control parameters. This avoids long-term decision-making inaccuracies caused by dynamic changes in the network topology or the solidification of static parameters, thereby comprehensively improving the bottom-level self-learning ability and life-cycle operational stability of the intelligent compensation box-type substation in complex high-impedance environments.

[0012] Furthermore, in step S2, the power grid operating parameters used to extract the transient distortion characteristics include the phase current change rate and the negative sequence current component change rate. When the rate of change of the phase current is greater than the first preset rate of change limit and the rate of change of the negative sequence current component is greater than the second preset rate of change limit, the transient distortion feature is determined to exist.

[0013] By using both the phase current change rate and the negative sequence current component change rate as extraction parameters, the determination of transient distortion can be based on the combined degree of instantaneous current change and asymmetric change. This avoids inaccurate identification caused by isolated judgment based solely on the reactive power deficit itself or single-phase current fluctuations. It helps to distinguish abnormal responses caused by complex operating conditions such as single-phase impacts or asymmetric faults from normal fluctuations of three-phase symmetrical loads, thereby improving the system's accuracy in identifying transient reactive power deficit changes, reducing the probability of compensation actions being falsely triggered, and providing a more reliable data foundation for subsequent suspension processing.

[0014] Furthermore, in step S2, after suspending the compensation process for the current reactive power deficit, the rate of change of the current reactive power deficit, the rate of change of the phase current, and the rate of change of the negative sequence current component are continuously monitored during the first observation period. When the rate of change of the current reactive power deficit is less than the first recovery threshold, the rate of change of the phase current is less than the second recovery threshold, and the rate of change of the negative sequence current component is less than the third recovery threshold within at least two consecutive sampling sub-windows, the suspension state of the compensation processing for the current reactive power deficit is released.

[0015] By continuously monitoring the three key parameters mentioned above during the first observation period, and using the simultaneous lowering of all three parameters below the corresponding recovery threshold within at least two consecutive sampling sub-windows as the trigger condition for lifting the suspension, the recovery of compensation processing can be based on continuous confirmation that the grid disturbance has stabilized. This avoids prematurely lifting the suspension state by making isolated judgments based solely on a transient decline at a single moment or a decrease in the fluctuation of a single parameter. It helps to distinguish between the transient tail fluctuations of the grid disturbance that have not yet completely subsided and the real reactive power demand that has truly recovered and stabilized. This prevents the control system from falling into repeated suspension and lifting action oscillations at the edge of transient distortion decay, thereby improving the accuracy and stability of compensation recovery timing determination under complex operating conditions, and providing a reliable steady-state data environment for the generation of subsequent compensation schemes.

[0016] Furthermore, in step S3, the calculation of the predicted voltage change based on the system's equivalent short-circuit impedance parameters, candidate compensation capacity, and current bus voltage specifically includes: Calculate the ratio of the current bus voltage to the system rated voltage, and correct the candidate compensation capacity based on the square of the ratio to obtain the corrected compensation capacity; Divide the corrected compensation capacity by the current bus voltage to obtain the theoretical reactive current; The predicted voltage change is obtained by multiplying the theoretical reactive current by the equivalent short-circuit impedance parameter of the system.

[0017] By calculating the square of the ratio of the current bus voltage to the system rated voltage to correct the candidate compensation capacity, the electrical characteristic that the actual reactive power output of the capacitor is proportional to the square of the operating voltage can be transformed into a basis for pre-calculation capacity correction, avoiding prediction distortion caused by static estimation based solely on the rated capacity on the nameplate of the compensation equipment. Furthermore, by converting the corrected compensation capacity into theoretical reactive current and multiplying it by the system's equivalent short-circuit impedance parameter, the prediction of voltage changes can be based on the joint constraints of the actual output of the equipment and the real-time strength of the power grid. This transforms the potential impact of compensation actions on the bus voltage into quantifiable risk characterization results, which helps to more accurately reflect the actual voltage rise after the capacitor is put into operation under specific operating conditions. This improves the reliability of the pre-assessment of overvoltage exceedance risks and provides precise quantitative support for subsequent downgrading or blocking interventions of candidate compensation schemes.

[0018] Furthermore, in step S3, the downgrading or blocking process for the candidate compensation scheme specifically includes: The candidate compensation capacity is divided into multiple sub-compensation capacities, and the corresponding single-step predicted voltage change is calculated based on each of the sub-compensation capacities. If there is a capacity splitting method that ensures that the single-step predicted voltage change corresponding to each of the sub-compensation capacities does not exceed the safety threshold, then the candidate compensation scheme is changed to a progressive execution scheme that deploys each of the sub-compensation capacities in steps according to a preset time interval, as the degradation process. If there is no capacity splitting method that ensures that the single-step predicted voltage change corresponding to each of the sub-compensation capacities does not exceed the safety threshold, then the candidate compensation scheme is intercepted as the blocking process.

[0019] By splitting the candidate compensation capacity into multiple sub-compensation capacities and calculating the corresponding single-step predicted voltage change for each, the concentrated voltage surge caused by a one-time large-capacity input can be decomposed into multiple controllable small-amplitude voltage responses. This avoids the instantaneous over-limit of the bus voltage caused by rigid open-loop switching based solely on the total reactive power deficit. Furthermore, by performing either step-by-step progressive input degradation processing or direct interception blocking processing based on the comparison results of the single-step predicted voltage change and the safety threshold, the final compensation command can be issued based on the joint constraints of step-by-step capacity extrapolation and the ultimate carrying capacity of the power grid. This prevents the controller from blindly outputting dangerous compensation actions when the overvoltage risk is known, and helps to establish a dynamic balance between sufficient reactive power compensation and bus voltage safety, thereby improving the execution flexibility and boundary self-protection capability of the intelligent compensation box-type substation under complex high-impedance conditions.

[0020] Further, in step S3, determining the target compensation scheme from the candidate compensation schemes or the processed candidate compensation schemes by combining the abnormal operation history of each of the capacitor compensation branches specifically includes: Monitor the actual operational response of each of the capacitor compensation branches; If it is detected that a specific capacitor compensation branch triggers overvoltage protection and is disconnected within a preset short time window after the closing action, a penalty mark is recorded in the abnormal operation history corresponding to the specific capacitor compensation branch. The penalty sleep duration of the specific capacitor compensation branch is determined based on the cumulative frequency of the penalty flag, and the dynamic availability level of the specific capacitor compensation branch is reduced within the penalty sleep duration. Candidate compensation schemes or processed candidate compensation schemes that include capacitor compensation branches within the penalty sleep duration are eliminated, and the target compensation scheme is determined from the remaining candidate compensation schemes or processed candidate compensation schemes.

[0021] By monitoring the actual action response of the branch and recording the abnormal clearing behavior that triggers overvoltage protection within a short time window as a penalty mark, the unstable characteristics of the underlying equipment in actual operation can be transformed into historical constraint information for participation in top-level scheduling. This avoids the controller repeatedly calling compensation branches that have already shown local degradation by performing memoryless cyclic switching based solely on static reactive power deficit. Furthermore, by dynamically determining the sleep duration based on the cumulative frequency of penalty marks and eliminating candidate schemes containing such sleep branches, the determination of the final target scheme can be based on the joint constraints of global reactive power demand and the health status of individual branches. This is beneficial for forcibly removing specific abnormal branches that are prone to triggering protection actions from the compensation sequence within a preset time period, preventing the control system from performing destructive repeated reclosing actions on the same defective branch in a very short time. This effectively curbs the continuous action oscillation caused by local anomalies and improves the self-healing scheduling capability of smart substations under non-ideal equipment conditions and the service life of electrical equipment.

[0022] Furthermore, the intelligent compensation box-type substation also includes a static var generator module connected to the low-voltage busbar; The method further includes: when the specific capacitor compensation branch is within the penalty sleep period, if the current reactive power deficit is not completely eliminated by the target compensation scheme, the uneliminated reactive power deficit is allocated to the static var generator module, and the static var generator module is scheduled to output a flexible reactive current that matches the uneliminated reactive power deficit, so as to replace the specific capacitor compensation branch within the penalty sleep period in performing reactive power compensation.

[0023] By allocating the remaining reactive power deficit to the static var generator (SVR) module during the penalty sleep period of a specific capacitor compensation branch, and scheduling its output to provide matching flexible reactive current for alternative compensation, the capacity loss caused by the forced shutdown of discrete capacitor branches can be transformed into a continuously adjustable active compensation command. This avoids the persistent compensation gap caused by directly isolating locally deteriorated equipment. Furthermore, using flexible reactive current to replace specific sleep branches for compensation allows the overall compensation action to be based on the complementary advantages of step-type passive switching and continuous flexible adjustment. This is beneficial for maintaining the hardware isolation of abnormal branches while supplementing the compensation tasks that discrete capacitor branches cannot handle, achieving a smooth relay of the remaining reactive power deficit, thereby improving the full-capacity compensation guarantee capability and control continuity of the intelligent prefabricated substation under locally constrained equipment conditions.

[0024] Further, in step S4, updating the system equivalent short-circuit impedance parameter, the safety threshold, and the sudden change threshold based on the deviation between the actual voltage change and the predicted voltage change, and the actual reactive power change, specifically includes: Based on the actual voltage change and the actual reactive power change, calculate the measured short-circuit impedance value corresponding to the current action; The measured short-circuit impedance value and the original equivalent short-circuit impedance parameter of the system are weighted and smoothed by filtering, and the calculation result is used as the updated equivalent short-circuit impedance parameter of the system. If the absolute value of the deviation is greater than the preset error tolerance limit, the safety threshold is decreased by a preset first correction step size, and the mutation threshold is increased by a preset second correction step size.

[0025] By inferring the measured short-circuit impedance value based on actual voltage and reactive power changes and then applying weighted smoothing filtering, the actual power grid strength feedback can be transformed into dynamic evolution parameters of the underlying simulation model. This avoids prediction inaccuracies caused by open-loop estimation based solely on fixed impedance, allowing the controller's equivalent power grid model to gradually approximate the actual operating state. Furthermore, when the prediction deviation exceeds the error tolerance limit, by reducing the safety threshold and increasing the mutation threshold, the control action can be based on conservative defensive constraints under model inaccuracy conditions. This avoids maintaining a lenient assessment under conditions of large deviations, which could induce false switching. This is beneficial for increasing the constraint strength on overvoltage risk and reducing sensitivity to short-term abnormal fluctuations in subsequent control, thereby enhancing the adaptive capability of parameter correction and improving the operational reliability throughout the entire life cycle under complex operating conditions.

[0026] The present invention also provides a compensation control system for an intelligent compensation box-type substation, the substation including a low-voltage bus and multiple capacitor compensation branches connected to the low-voltage bus; The compensation control system includes: The parameter acquisition module is used to acquire power grid operation parameters including three-phase voltage, three-phase current, reactive power calculation value and state quantities characterizing the degree of three-phase imbalance, as well as the operation status of each capacitor compensation branch. The distortion determination and suspension module is used to extract transient distortion features based on the power grid operating parameters and determine the current reactive power deficit; if the rate of change of the current reactive power deficit exceeds the abrupt change threshold and the transient distortion features exist, the compensation processing for the current reactive power deficit is suspended. The scheme deduction module is used to generate candidate compensation schemes if the compensation process for the current reactive power deficit is not suspended, and to calculate the predicted voltage change based on the system equivalent short-circuit impedance parameters, candidate compensation capacity and current bus voltage. A degradation or blocking module is used to perform degradation or blocking processing on the candidate compensation scheme if the predicted voltage change exceeds a safety threshold. The target scheme determination module is used to determine the target compensation scheme from the candidate compensation schemes or the processed candidate compensation schemes by combining the abnormal operation history of each of the capacitor compensation branches. The execution and acquisition module is used to control the capacitor compensation branch corresponding to the target compensation scheme to perform compensation actions after the target compensation scheme is determined, and to acquire the actual voltage change and actual reactive power change after reaching steady state. The parameter closed-loop update module is used to update the system equivalent short-circuit impedance parameter, the safety threshold, and the mutation threshold based on the deviation between the actual voltage change and the predicted voltage change, and the actual reactive power change.

[0027] Through the underlying collaboration between the parameter acquisition module and the distortion judgment and suspension module, an initial defense mechanism against transient fluctuations can be built at the front end of the system control architecture, preventing the hardware controller from blindly responding to temporary reactive power deficits caused by disturbances such as single-phase impacts. Furthermore, through the series interaction between the scheme deduction module and the degradation or blocking module, the capacitor output characteristics and grid strength parameters are transformed into pre-deduction constraints, ensuring that compensation actions are based on a quantitative assessment of proactive overvoltage prevention, preventing the system from outputting dangerous commands that cause physical bus limits to exceed. Simultaneously, the target scheme determination module incorporates abnormal operating history, giving the control system the ability to identify and isolate locally deteriorated branches, thus preventing repeated reclosing oscillations in the underlying hardware. Finally, the underlying feedback data stream, composed of the execution and acquisition module and the parameter closed-loop update module, transforms the steady-state deviation of actual actions into a driving force for correcting control thresholds and impedance models, overcoming the adaptive deficiencies caused by the long-term fixation of control parameters in traditional hardware devices. This comprehensively improves the closed-loop self-healing capability and long-term operational reliability of the intelligent substation control system under complex and evolving operating conditions.

[0028] The present invention also provides an intelligent compensation box-type substation, including a box, and a transformer, a low-voltage bus, a data acquisition device, multiple capacitor compensation branches and a controller installed in the box; The transformer and the plurality of capacitor compensation branches are all connected to the low-voltage busbar; The data acquisition device is connected to the low-voltage bus and the controller respectively; The controller is connected to each of the capacitor compensation branches, and the controller is configured to perform the compensation control method of the intelligent compensation box-type substation as described in any of the above.

[0029] By highly integrating data acquisition devices, controllers, and multiple capacitor compensation branches within the enclosure, and configuring the controller to execute the aforementioned advanced compensation control methods, the physical hardware architecture of the substation and the closed-loop adaptive control logic can achieve deep synergy, avoiding the shortcomings of the open-loop isolated operation of monitoring units and execution hardware in traditional substations. This architecture enables the controller to directly implement global unified scheduling of each capacitor compensation branch based on the underlying electrical state obtained by the data acquisition device, including transient disturbance rejection, overvoltage inference blocking, and abnormal state stripping. This facilitates the direct integration of complex grid disturbance identification and equipment self-healing defense mechanisms into the substation itself, thereby providing highly secure and reliable reactive power compensation and voltage stability support for complex high-impedance distribution networks in a compact physical form.

[0030] (III) Beneficial effects of the present invention: By jointly identifying the transient distortion characteristics of the power grid and the changes in reactive power deficit, and by performing bus voltage extrapolation and pre-constraint based on the equivalent short-circuit impedance of the system before compensation action, and by combining the abnormal operation history of each compensation branch and the actual steady-state feedback to determine the target scheme and update the control parameters, the invention achieves reliable identification of transient reactive power deficit changes and real steady-state requirements under complex operating conditions, reduces the risk of bus overvoltage exceeding the limit caused by blindly investing large-capacity capacitors, blocks repeated switching oscillations caused by local abnormalities in branches, and improves the adaptive correction capability and long-term operational safety of the compensation control model under the dynamic evolution of power grid operating conditions. Attached Figure Description

[0031] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0032] Figure 1 A schematic block diagram of the structure of an intelligent compensation box-type substation and its compensation control system provided in one embodiment of the present invention; Figure 2 This is a flowchart illustrating a compensation control method for an intelligent compensation box-type substation according to an embodiment of the present invention. Detailed Implementation

[0033] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific Implementation

[0034] like Figure 1 As shown, this embodiment provides an intelligent compensation prefabricated substation and its compensation control system and method. This intelligent compensation prefabricated substation is preferably deployed at the end of industrial parks, new energy grid connection points, or areas with concentrated heavy loads to adapt to the complex operating conditions in such scenarios, where the system's equivalent short-circuit impedance is large and transient distortion is easily induced by the starting of large motors or single-phase faults. The intelligent compensation prefabricated substation includes a enclosure, and a transformer, low-voltage busbar, data acquisition device, multiple capacitor compensation branches, and a controller housed within the enclosure. The high-voltage side of the transformer is connected to the external distribution network, while its low-voltage side and multiple capacitor compensation branches are connected to the low-voltage busbar, enabling the low-voltage side power output from the transformer to be distributed to the load side via the low-voltage busbar, and the reactive power on the low-voltage busbar side to be regulated in stages by the multiple capacitor compensation branches. The data acquisition device is connected to both the low-voltage bus and the controller to collect power grid operating parameters and the operating status of each capacitor compensation branch on the low-voltage bus side, and then transmits these data to the controller. The controller is connected to each capacitor compensation branch to control the activation and deactivation of each capacitor compensation branch based on the collected operating data. Thus, an integrated compensation control architecture consisting of a primary electrical unit, a parameter acquisition unit, and a compensation control unit is formed within the enclosure, providing the hardware foundation for the execution of subsequent compensation control methods.

[0035] Specifically, multiple capacitor compensation branches can have different capacities, for example, configured using a binary encoding method, or they can be multiple discrete compensation units with the same capacity. Each capacitor compensation branch can include a series-connected switching switch, a series reactor, and a power capacitor, wherein the switching switch is preferably a thyristor contactless switch or a composite switch to achieve rapid zero-crossing switching. The controller's control connection with each capacitor compensation branch can specifically be manifested as a connection to the trigger terminal of the aforementioned switching switch, thereby enabling the issuance of activation or deactivation commands to the corresponding branch. Since the control method of this invention needs to perform screening, degradation, blocking, and target scheme determination on candidate compensation schemes under different operating conditions, the grouping and independent switching capability of multiple capacitor compensation branches constitute the physical basis for subsequent candidate compensation scheme generation and target compensation scheme execution.

[0036] In a preferred embodiment, the intelligent compensation prefabricated substation further includes a static var generator (SVG) module connected to the low-voltage busbar. This SVG module preferably employs a converter circuit structure based on fully controlled power electronic devices such as IGBTs. This allows the controller to schedule the SVG module to output continuously adjustable flexible reactive current when a specific capacitor compensation branch is in a penalty dormant state and the target compensation scheme fails to completely eliminate the current reactive power deficit. This enables the SVG module to continue compensating for the remaining compensation tasks that the discrete capacitor compensation branch cannot handle. By integrating the SVG module within the prefabricated substation, discrete step-type passive compensation and continuous flexible active compensation can work in synergy, providing hardware support for compensating for the remaining deficit after the removal of abnormal branches.

[0037] In this embodiment, the data acquisition device includes a voltage sampling unit, such as a three-phase PT, a current sampling unit, such as a three-phase CT, and a branch status acquisition unit. To accurately capture transient disturbances, the data acquisition device preferably has high-frequency sampling capability, for example, a sampling frequency set to no less than 3.2kHz. The voltage and current sampling units are used to acquire the three-phase voltage and current on the low-voltage bus side, providing basic waveform data for reactive power calculations and state variables characterizing the degree of three-phase imbalance, such as the negative sequence current component. The branch status acquisition unit is used to acquire the opening and closing status and overcurrent / overvoltage protection status of each capacitor-compensated branch, so that the controller can simultaneously consider the bus operating status and the physical availability status of the branches when generating candidate compensation schemes and determining the target compensation scheme. In some embodiments, the state variables characterizing the degree of three-phase imbalance and the reactive power calculation values ​​can be calculated in real-time online by the controller, such as a DSP or ARM processor, based on the acquired three-phase high-frequency data, or they can be calculated by the preprocessing chip inside the data acquisition device and directly sent to the controller. In other words, this embodiment does not limit the above parameters to be directly output by a single piece of hardware, but only requires that the controller be able to obtain the corresponding power grid operating parameters and branch operating status for executing the compensation control method.

[0038] In this embodiment, the controller integrates a compensation control system corresponding to the aforementioned compensation control method. This compensation control system includes a parameter acquisition module, a distortion determination and suspension module, a scheme deduction module, a degradation or blocking module, a target scheme determination module, an execution and acquisition module, and a parameter closed-loop update module. The system comprises the following modules: a parameter acquisition module for receiving operational data from a data acquisition device; a distortion determination and suspension module for extracting transient distortion features based on operational parameters and determining the current reactive power deficit, suspending compensation processing when the rate of change of the deficit exceeds the mutation threshold and transient distortion features are present; a scheme deduction module for generating candidate compensation schemes when compensation processing is not suspended, and calculating the predicted voltage change based on the system's equivalent short-circuit impedance parameters, candidate compensation capacity, and current bus voltage; a degradation or blocking module for performing degradation or blocking processing when the predicted voltage change exceeds the safety threshold; a target scheme determination module for determining the final target compensation scheme by combining the abnormal operation history of each capacitor compensation branch; an execution and acquisition module for controlling the hardware branch to perform compensation actions and acquiring the actual voltage change and actual reactive power change after reaching steady state; and a parameter closed-loop update module for back-calculating and updating the system's equivalent short-circuit impedance parameters, safety threshold, and mutation threshold based on the prediction deviation and actual reactive power change. The above modules can be centrally deployed in a single processing device in the form of software program units, firmware units, or hardware-software co-operation units, or they can be distributed according to function in multiple control boards composed of DSP and FPGA that are interconnected with each other, as long as they can complete the corresponding data processing, status determination, compensation deduction and parameter update functions.

[0039] Based on the aforementioned physical structure and modular architecture, the intelligent compensation box-type substation in this embodiment does not merely execute static compensation logic based on a single threshold. Instead, through the coordinated operation of data acquisition devices, controllers, multiple capacitor compensation branches, and static var generator modules, the controller can construct a complete compensation control closed loop around the transient characteristics of the power grid on the low-voltage bus side, the operating status of each branch, and the feedback of compensation actions. That is, this embodiment first acquires operating parameters and branch status through hardware, and then each functional module of the system sequentially completes transient distortion identification, pre-operation risk simulation, scheme degradation or blocking, branch anomaly screening, SVG flexible connection compensation, and parameter closed-loop update, thereby providing a unified system foundation and hardware support for the specific implementation of each step of the subsequent compensation control method. The following provides a further detailed explanation of the specific execution process and arithmetic logic of the compensation control method described in this embodiment.

[0040] like Figure 2 As shown, the specific execution process of this compensation control method will be further explained below. This embodiment takes a substation located at the end of a power grid with high impedance and drastic load fluctuations as an example. For a normal, stable power grid environment, the execution logic is basically the same as the following process, except that some characteristic parameters will not trigger the corresponding suspension, interception, or degradation conditions. The controller executes the following compensation control method in each control cycle, as follows: Step S1: Obtain runtime parameters and establish current state In this embodiment, the controller continuously acquires the power grid operating parameters on the low-voltage bus side and the operating status of each capacitor compensation branch through a data acquisition device. These power grid operating parameters include at least three-phase voltage, three-phase current, calculated reactive power, and state variables characterizing the degree of three-phase imbalance. The operating status of each capacitor compensation branch includes at least the branch's current on / off status, most recent action status, and protection status. Preferably, the data acquisition device continuously samples the three-phase voltage and three-phase current on the low-voltage bus side according to a preset sampling period and sends the sampled data to the controller, which then performs parameter calculations and status updates under a unified time base.

[0041] Among them, three-phase voltage and three-phase current serve as the underlying sampled data to characterize the current operating condition of the low-voltage bus. The reactive power calculation value is obtained online by the controller based on the three-phase voltage and current to reflect the reactive power demand level on the low-voltage bus side within the current control cycle. The state variable characterizing the degree of three-phase imbalance is preferably the negative sequence current component; if necessary, it can be combined with other parameters reflecting the degree of three-phase imbalance for comprehensive characterization. The operating status of each capacitor compensation branch reflects whether the corresponding branch is currently in operation, whether protection actions have occurred, and whether it possesses the basic conditions to continue participating in compensation. By simultaneously establishing the bus electrical status and branch operating status within the same control cycle, the controller can obtain a unified state basis required for subsequent reactive power deficit determination and scheme selection.

[0042] Within the current control cycle, the controller also determines the current reactive power deficit based on the calculated reactive power value. Preferably, the calculated reactive power value within the current control cycle can be compared with a preset target compensation level to obtain the reactive power deficit that has not yet been eliminated. For example, in one implementation, the reactive power deficit Qgap(i) at the current time t(i) can be expressed as: Qgap(i) = Qtarget - Q(i) Where Qtarget represents the target reactive power value corresponding to the preset target compensation level, and Q(i) represents the calculated reactive power value at the current moment. To facilitate subsequent identification of short-term abnormal disturbances, this embodiment not only focuses on the absolute value of the current reactive power deficit but also on its rate of change between continuous control cycles. This rate of change can be calculated from the difference in reactive power deficit within adjacent time windows and the corresponding time interval, reflecting how quickly reactive power demand changes over a short period of time.

[0043] Step S2: Transient distortion feature extraction and compensation processing (suspended) After obtaining the grid operating parameters, the controller further extracts transient distortion features based on these parameters and determines whether the current reactive power deficit is an abnormal deficit that needs to be temporarily deferred. In this embodiment, the grid operating parameters used to extract transient distortion features include the phase current change rate and the negative sequence current component change rate. Preferably, the controller calculates the change rate of the phase current value and the negative sequence current component value within a preset time window based on adjacent sampling times, to form a judgment parameter reflecting the degree of short-term abrupt change.

[0044] The phase current change rate characterizes the degree of current jump on the load side within a short period of time. When there is a single-phase impact load, instantaneous commissioning of high-power equipment, or local short-circuit disturbance, the phase current change rate usually increases significantly. The negative sequence current component change rate characterizes the intensity of the three-phase current imbalance. When an asymmetrical disturbance or single-phase abnormal condition occurs in the power grid, the negative sequence current component change rate also increases synchronously. By simultaneously monitoring these two parameters, the controller can identify transient distortion states in the power grid from two dimensions: current abrupt change characteristics and imbalance change characteristics.

[0045] Specifically, the controller compares the above calculation results with preset judgment limits: when the phase current change rate is greater than the first preset change rate limit and the negative sequence current component change rate is greater than the second preset change rate limit, the controller determines that there is a transient distortion feature. Subsequently, the controller further compares the change rate of the current reactive power deficit with a preset abrupt change threshold; if the change rate of the current reactive power deficit exceeds the abrupt change threshold and the transient distortion feature is also determined to exist, the controller determines that the current reactive power deficit is an abnormal deficit that is not suitable for direct entry into the compensation decision. At this time, the system does not immediately enter the candidate compensation scheme generation process, but suspends the compensation processing for the current reactive power deficit, thereby avoiding the direct use of calculation anomalies caused by short-term disturbances as real steady-state compensation requirements and triggering blind switching.

[0046] This embodiment also includes compensation for suspension and recovery: After the compensation process is suspended, the controller does not immediately resume the compensation judgment, but instead enters a first observation period. During this first observation period, the controller continuously monitors the rate of change of the current reactive power deficit, the rate of change of the phase current, and the rate of change of the negative sequence current component. Preferably, this first observation period can be set as a monitoring interval consisting of multiple consecutive sampling sub-windows, for example, each sampling sub-window corresponds to one power frequency cycle, and the controller updates the current value of the corresponding parameter at the end of each sampling sub-window.

[0047] In this embodiment, the controller uses a continuous confirmation method to release the suspended state. The controller only releases the suspension of compensation processing for the current reactive power deficit when, within at least two consecutive sampling sub-windows, the rate of change of the current reactive power deficit is less than the first recovery threshold, the rate of change of the phase current is less than the second recovery threshold, and the rate of change of the negative sequence current component is less than the third recovery threshold. That is, releasing the suspension is not based on a brief dip at a single instantaneous sampling point, but rather on the premise that the relevant key parameters have stabilized below the corresponding recovery thresholds within multiple consecutive sampling sub-windows. This multi-parameter, continuous time-window confirmation method prevents the controller from prematurely resuming compensation processing before grid disturbances have fully attenuated, avoiding repeated switching between suspension, release, and re-suspension.

[0048] When the suspended state is lifted based on the above conditions, or when the controller confirms during the judgment that the current reactive power deficit does not simultaneously meet the conditions of mutation and distortion and is not suspended, the controller will use the current effective reactive power deficit as the input condition and sequentially enter the candidate scheme deduction and risk assessment process in step S3.

[0049] Step S3: Generation of candidate compensation schemes, risk simulation and determination of target scheme In step S2, if the controller confirms that the current reactive power deficit has not been suspended, or if the recovery conditions are met and the suspension is lifted within the first observation period, the controller will use the current effective reactive power deficit as input and sequentially enter step S3 to generate candidate compensation schemes, perform risk simulation, downgrade or block them, and determine the final executable target compensation scheme by combining the abnormal operation history of each capacitor compensation branch.

[0050] Specifically, the controller first generates candidate compensation schemes based on the current reactive power deficit and the operating status of each capacitor compensation branch. Preferably, the controller uses the current reactive power deficit that has not yet been eliminated as the target compensation amount, and combines it with the current on / off status, protection status, and availability status of each capacitor compensation branch to select one or more branches from multiple capacitor compensation branches to form candidate compensation schemes. Each candidate compensation scheme corresponds to a candidate compensation capacity, which characterizes the theoretical compensation capability that the scheme can provide when executed. This candidate compensation capacity can be obtained by combining the rated compensation capacities of each capacitor compensation branch constituting the candidate compensation scheme, or it can be obtained by adjusting the rated capacity based on the current status of the branch.

[0051] After generating candidate compensation schemes, the controller does not directly issue switching actions based on the candidate compensation capacity. Instead, it first calculates the predicted voltage change based on the system's equivalent short-circuit impedance parameters, the candidate compensation capacity, and the current bus voltage to determine the potential bus voltage response after the candidate compensation scheme is executed. The system's equivalent short-circuit impedance parameters can be the initial system setting or updated based on actual feedback after previous compensation actions. To ensure the risk simulation results reflect the actual effective output of the capacitor compensation branch under current operating conditions, the controller first calculates the ratio of the current bus voltage to the system's rated voltage and then corrects the candidate compensation capacity based on the square of this ratio, obtaining the corrected compensation capacity. In engineering principle, capacitors are essentially constant impedance elements, and their reactive power output is proportional to the square of the operating voltage. Through this correction, the controller transforms the static nameplate capacity into an effective compensation capacity under real power grid conditions. Preferably, the corrected compensation capacity Q_adj can be expressed as: Q_adj = Q_c × (U_cur / U_n)^2 Where Q_c is the candidate compensation capacity, U_cur is the current bus voltage, and U_n is the system rated voltage.

[0052] Subsequently, the controller divides the corrected compensation capacity by the current bus voltage to obtain the theoretical reactive current I_q, that is: I_q = Q_adj / U_cur Multiplying the theoretical reactive current by the system's equivalent short-circuit impedance parameter Z_eq yields the predicted voltage change ΔU_pre, i.e.: ΔU_pre = I_q × Z_eq Therefore, the controller can predict the potential bus voltage rise caused by the candidate compensation scheme before the actual compensation action is executed. The controller compares the predicted voltage change with a preset safety threshold. If the predicted voltage change does not exceed the safety threshold, it indicates that the candidate compensation scheme meets the basic bus voltage safety constraints under the current operating conditions and can continue to participate in the subsequent target scheme selection; if the predicted voltage change exceeds the safety threshold, the controller does not directly execute the candidate compensation scheme, but further performs degradation or blocking processing on the candidate compensation scheme to prevent forced closing from inducing underlying overvoltage protection.

[0053] Specifically, the controller breaks down the candidate compensation capacity into multiple sub-compensation capacities and calculates the corresponding single-step predicted voltage change based on each sub-compensation capacity. If a capacity splitting method exists that ensures the single-step predicted voltage change for each sub-compensation capacity does not exceed the safety threshold, the controller changes the original candidate compensation scheme to a progressive execution scheme that deploys each sub-compensation capacity in stages according to a preset time interval, as a degradation process. Conversely, if no capacity splitting method exists that ensures the single-step predicted voltage change for each sub-compensation capacity does not exceed the safety threshold (typically occurring under extreme high-impedance weak network conditions), the controller directly intercepts the candidate compensation scheme and treats it as a blocking result. Thus, the originally large-capacity compensation action, which was originally deployed all at once, is flexibly transformed into multiple time-sharing, tiered, and controlled small-capacity compensation actions, or directly shielded when the risk is unacceptable, thereby ensuring that the compensation action is based on the safety constraints of the bus voltage.

[0054] After completing the aforementioned risk handling, the controller further combines the abnormal operation history of each capacitor compensation branch to determine the target compensation scheme from the candidate compensation schemes or the processed candidate compensation schemes. To this end, the controller continuously monitors the actual action response of each capacitor compensation branch during operation. If a specific capacitor compensation branch is detected to have triggered overvoltage protection and been disconnected within a preset short time window after a closing action, the controller records a penalty flag in the abnormal operation history corresponding to that specific capacitor compensation branch. As the controller continuously monitors the operation process, this penalty flag accumulates in the abnormal operation history. The controller determines the penalty sleep duration for the specific capacitor compensation branch based on the cumulative frequency of the penalty flag and reduces the dynamic availability level of that specific capacitor compensation branch during this penalty sleep duration. Subsequently, the controller forcibly eliminates candidate compensation schemes or processed candidate schemes that include capacitor compensation branches within this penalty sleep duration and determines the final target compensation scheme from the remaining schemes. Therefore, the final target compensation scheme is not determined solely based on the reactive power capacity matching degree, but is simultaneously constrained by both the pre-commissioning voltage risk and the historical health status of the branch, thus completely severing the vicious cycle of oscillation tripping caused by the control system repeatedly calling defective hardware without memory.

[0055] In a preferred embodiment, for the aforementioned implementation environment equipped with a static var generator (SVA) module, when a specific capacitor compensation branch is within its penalty sleep period, and the target compensation scheme fails to completely eliminate the current reactive power deficit, the controller further allocates the remaining reactive power deficit to the SVA module and schedules the SVA module to output a flexible reactive current matching the remaining reactive power deficit, thereby replacing the specific capacitor compensation branch in its penalty sleep state to perform reactive power compensation. Thus, after the abnormal branch is removed, the system does not directly lose its ability to compensate for the remaining deficit due to the temporary withdrawal of the compensation branch. Instead, the SVA module provides continuously adjustable flexible reactive power compensation, smoothly continuing the remaining compensation tasks that the discrete capacitor compensation branch cannot handle.

[0056] After completing the aforementioned risk defense constraints and hardware health screening, and determining the final executable target compensation scheme, the controller will sequentially enter step S4 to execute the underlying physical actions and initiate the closed-loop update process of control parameters.

[0057] Step S4: Compensation action execution and parameter closed-loop update After determining the target compensation scheme in step S3, the controller sequentially proceeds to step S4 to execute compensation actions on the capacitor compensation branch corresponding to the target compensation scheme. After the actions are completed and the system reaches a new steady state, the equivalent short-circuit impedance parameters, safety threshold, and sudden change threshold of the system are updated in a closed loop based on the actual feedback. Thus, the controller no longer relies on a one-time static decision but can continuously revise the underlying deduction model and judgment threshold based on the actual execution results of the compensation actions.

[0058] Specifically, the controller issues activation commands to the corresponding hardware branches based on the target compensation scheme. If the target compensation scheme is a progressive execution scheme after degradation processing, the controller sequentially controls the activation of the capacitor compensation branches corresponding to each sub-compensation capacity according to a preset time interval; if the target compensation scheme is a candidate compensation scheme without degradation processing, the controller directly executes the centralized compensation action according to the branch combination corresponding to the scheme. In this way, the target compensation scheme is accurately implemented into a specific physical switching process.

[0059] After issuing the action command, the controller does not immediately collect data, but introduces a steady-state waiting delay to avoid the transient transition process at the moment the capacitor closes. When the voltage and reactive power change trends on the low-voltage bus side remain within the allowable fluctuation range within a preset time window, the controller determines that the current compensation action has entered the steady-state stage. At this time, the controller collects the actual voltage change and the actual reactive power change after reaching the steady state. The actual voltage change can be obtained from the bus voltage difference before and after the compensation action, and the actual reactive power change can be obtained from the reactive power difference before and after the compensation action.

[0060] In this embodiment, the controller further calculates the measured short-circuit impedance value corresponding to the current action based on the actual voltage change and the actual reactive power change. The specific derivation is as follows: The controller first converts the actual reactive power change ΔQ_act into the actual reactive current I_q,act based on the current steady-state bus voltage U_cur. I_q,act = ΔQ_act / U_cur Subsequently, based on the actual voltage change ΔU_act and the actual reactive current I_q,act, the measured short-circuit impedance value Z_meas corresponding to the current action is calculated: Z_meas = ΔU_act / I_q,act Its engineering implications are: to inversely determine the actual response strength of the power grid to reactive power disturbances under the current operating conditions through the actual execution results.

[0061] After obtaining the measured short-circuit impedance value, to avoid the controller making excessive corrections to the impedance model due to single measurement errors or occasional power grid fluctuations, the controller employs a first-order inertial filtering algorithm to perform weighted smoothing filtering calculations on the measured short-circuit impedance value and the original system equivalent short-circuit impedance parameters, and uses the calculation results as the updated system equivalent short-circuit impedance parameters. This weighted smoothing filtering calculation can be expressed as: Z_eq,new = α × Z_meas + (1 - α) × Z_eq,old Where Z_eq,new represents the updated equivalent short-circuit impedance parameter of the system, Z_meas represents the measured short-circuit impedance value, Z_eq,old represents the original equivalent short-circuit impedance parameter of the system before the update, and α is a preset smoothing filter coefficient (with a value range of 0 < α < 1). The updated Z_eq,new will be overwritten to the internal storage area, making the simulation model smoothly reflect the actual field conditions.

[0062] After updating the impedance parameters, the controller further verifies the accuracy of the preceding risk projection. The controller calculates the deviation e between the actual voltage change ΔU_act and the predicted voltage change ΔU_pre in step S3: e = | ΔU_act - ΔU_pre | The controller compares the deviation *e* with a preset error tolerance limit. When the deviation *e* exceeds the error tolerance limit, it indicates a significant deviation between the mathematical model under the current operating condition and the actual power grid environment. In this case, the controller decreases the safety threshold by a preset first correction step size and increases the mutation threshold by a preset second correction step size. Decreasing the safety threshold makes the constraints on bus overvoltage risk more stringent in subsequent control; increasing the mutation threshold reduces the controller's sensitivity to short-term abnormal fluctuations, preventing frequent suspension or erroneous compensation actions due to low thresholds under inaccurate model conditions. Therefore, the controller can proactively tighten the action boundaries when large prediction deviations occur, improving the stability and reliability of decision-making under complex operating conditions.

[0063] To further illustrate the complete execution process of the compensation control method in this embodiment, a detailed explanation is provided below in conjunction with a specific application scenario.

[0064] Taking a substation located at the end of a high-impedance industrial park as an example. At a certain operating moment, a high-power single-phase load is connected to the low-voltage bus. The instantaneous connection of this load causes phase current fluctuations and an increase in the negative-sequence current component. After the data acquisition device obtains the above-mentioned power grid operating parameters, the controller calculates the rate of change of the current reactive power deficit, the rate of change of the phase current, and the rate of change of the negative-sequence current component. When both the rate of change of the phase current and the rate of change of the negative-sequence current component are greater than their corresponding preset rate of change limits, the controller determines that there is a transient distortion characteristic; at the same time, the rate of change of the current reactive power deficit also exceeds the abrupt change threshold. At this time, the controller determines that the reactive power deficit is an abnormal deficit induced by a short-term disturbance, and then suspends the compensation processing for the current reactive power deficit, maintaining the current state of each capacitor compensation branch unchanged.

[0065] After executing the suspension operation, the controller enters the first observation period. Within two consecutive sampling sub-windows, as the single-phase load transient process ends and enters a stable operating state, the controller monitors that the rate of change of the current reactive power deficit, the rate of change of the phase current, and the rate of change of the negative sequence current component are all lower than the corresponding first, second, and third recovery thresholds, respectively. Once the release condition is met, the controller releases the suspension state and generates candidate compensation schemes using the current reactive power deficit of the system as input parameters.

[0066] Assume the generated candidate compensation scheme calls the first and second capacitor compensation branches. Based on the current bus voltage and the system's equivalent short-circuit impedance parameters, the controller performs a squared correction on the total capacity of the candidate compensation scheme and predicts the voltage change. If the calculated predicted voltage change corresponding to a one-time concentrated activation of the candidate compensation scheme exceeds a safety threshold, the controller performs a degradation process, changing the scheme to a progressive execution scheme that activates the first and second capacitor compensation branches in stages according to preset time intervals.

[0067] Before implementing the scheme, the controller further verifies the abnormal operating history of each branch. If it is found that the second capacitor compensation branch is in a penalty sleep period due to overvoltage protection, the controller removes the second capacitor compensation branch from the above scheme and determines that only the first capacitor compensation branch is to be used as the target compensation scheme. For the remaining reactive power deficit that is not eliminated after removing the second capacitor compensation branch, the controller allocates its value to the static var generator module and schedules the static var generator module to output a matching flexible reactive current to achieve continuous compensation with the first capacitor compensation branch.

[0068] The controller directs the first capacitor compensation branch and the static var generator module to perform actions. After the electrical transition process on the low-voltage bus side ends and a new steady state is reached, the controller collects the actual voltage change and the actual reactive power change, calculates the measured short-circuit impedance value corresponding to the current action, and performs weighted smoothing filtering with the original impedance parameters to update the system's equivalent short-circuit impedance parameters. Simultaneously, the controller calculates that the absolute value of the deviation between the actual voltage change and the predicted voltage change in the simulation phase is greater than the error tolerance limit. The controller decreases the safety threshold and increases the mutation threshold according to a preset step size to apply the updated safety constraint threshold and mutation judgment criteria in the next control cycle.

[0069] Through the complete operation process described above, the intelligent compensation box-type substation and its compensation control method provided in this embodiment can sequentially perform transient distortion identification, pre-commissioning simulation, branch screening and parameter closed-loop update under conditions such as high impedance weak network environment and local equipment aging, thereby avoiding oscillation and overvoltage tripping caused by open-loop response and static parameters, and realizing stable operation of the compensation system under complex conditions.

[0070] The embodiments described above are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still make modifications to the specific implementation methods, parameter settings, or component forms, or make equivalent substitutions for the corresponding technical features; all variations and substitutions that do not depart from the technical concept and essence of the present invention should fall within the protection scope of the present invention.

Claims

1. A method for compensation control of an intelligent-compensated box-type substation, the substation comprising a low-voltage bus, a plurality of capacitor compensation branches, and a controller, the method being characterized by, The controller performs the following steps: S1. Obtain power grid operation parameters including three-phase voltage, three-phase current, reactive power calculation values ​​and state quantities characterizing the degree of three-phase imbalance, as well as the operation status of each capacitor compensation branch; S2. Extract transient distortion features based on the power grid operating parameters and determine the current reactive power deficit; if the rate of change of the current reactive power deficit exceeds the abrupt change threshold and the transient distortion features exist, then suspend the compensation processing for the current reactive power deficit. S3. If the compensation process for the current reactive power deficit is not suspended, a candidate compensation scheme is generated, and the predicted voltage change is calculated based on the system equivalent short-circuit impedance parameter, the candidate compensation capacity, and the current bus voltage. If the predicted voltage change exceeds the safety threshold, the candidate compensation scheme is downgraded or blocked. The target compensation scheme is determined from the candidate compensation scheme or the processed candidate compensation scheme, based on the abnormal operation history of each capacitor compensation branch. S4. After determining the target compensation scheme, control the capacitor compensation branch corresponding to the target compensation scheme to perform the compensation action, and collect the actual voltage change and actual reactive power change after reaching steady state. Based on the deviation between the actual voltage change and the predicted voltage change, and the actual reactive power change, the system equivalent short-circuit impedance parameter, the safety threshold, and the mutation threshold are updated.

2. The compensation control method of the smart compensation cubicle substation according to claim 1, characterized in that, In step S2, the power grid operating parameters used to extract the transient distortion characteristics include the phase current change rate and the negative sequence current component change rate. When the rate of change of the phase current is greater than the first preset rate of change limit and the rate of change of the negative sequence current component is greater than the second preset rate of change limit, the transient distortion feature is determined to exist.

3. The compensation control method of the smart compensation cubicle substation according to claim 2, characterized in that, In step S2, after suspending the compensation process for the current reactive power deficit, the rate of change of the current reactive power deficit, the rate of change of the phase current, and the rate of change of the negative sequence current component are continuously monitored during the first observation period. When the rate of change of the current reactive power deficit is less than the first recovery threshold, the rate of change of the phase current is less than the second recovery threshold, and the rate of change of the negative sequence current component is less than the third recovery threshold within at least two consecutive sampling sub-windows, the suspension state of the compensation processing for the current reactive power deficit is released.

4. The compensation control method of the smart compensation cubicle substation according to claim 1, characterized in that, In step S3, the calculation of the predicted voltage change based on the system's equivalent short-circuit impedance parameters, candidate compensation capacity, and current bus voltage specifically includes: Calculate the ratio of the current bus voltage to the system rated voltage, and correct the candidate compensation capacity based on the square of the ratio to obtain the corrected compensation capacity; Divide the corrected compensation capacity by the current bus voltage to obtain the theoretical reactive current; The predicted voltage change is obtained by multiplying the theoretical reactive current by the equivalent short-circuit impedance parameter of the system.

5. The method of compensation control of a smart-compensated box-type substation according to claim 1, characterized in that, In step S3, the process of downgrading or blocking the candidate compensation scheme specifically includes: The candidate compensation capacity is divided into multiple sub-compensation capacities, and the corresponding single-step predicted voltage change is calculated based on each of the sub-compensation capacities. If there is a capacity splitting method that ensures that the single-step predicted voltage change corresponding to each of the sub-compensation capacities does not exceed the safety threshold, then the candidate compensation scheme is changed to a progressive execution scheme that deploys each of the sub-compensation capacities in steps according to a preset time interval, as the degradation process. If there is no capacity splitting method that ensures that the single-step predicted voltage change corresponding to each of the sub-compensation capacities does not exceed the safety threshold, then the candidate compensation scheme is intercepted as the blocking process.

6. The method of compensation control of a smart-compensated box-type substation according to claim 1, wherein, In step S3, determining the target compensation scheme from the candidate compensation schemes or the processed candidate compensation schemes, based on the abnormal operation history of each of the capacitor compensation branches, specifically includes: Monitor the actual operational response of each of the capacitor compensation branches; If it is detected that a specific capacitor compensation branch triggers overvoltage protection and is disconnected within a preset short time window after the closing action, a penalty mark is recorded in the abnormal operation history corresponding to the specific capacitor compensation branch. The penalty sleep duration of the specific capacitor compensation branch is determined based on the cumulative frequency of the penalty flag, and the dynamic availability level of the specific capacitor compensation branch is reduced within the penalty sleep duration. Candidate compensation schemes or processed candidate compensation schemes that include capacitor compensation branches within the penalty sleep duration are eliminated, and the target compensation scheme is determined from the remaining candidate compensation schemes or processed candidate compensation schemes.

7. The method of compensation control of a smart-compensated box-type substation according to claim 6, characterized in that, The intelligent compensation box-type substation also includes a static var generator module connected to the low-voltage busbar; The method further includes: when the specific capacitor compensation branch is within the penalty sleep period, if the current reactive power deficit is not completely eliminated by the target compensation scheme, the uneliminated reactive power deficit is allocated to the static var generator module, and the static var generator module is scheduled to output a flexible reactive current that matches the uneliminated reactive power deficit, so as to replace the specific capacitor compensation branch within the penalty sleep period in performing reactive power compensation.

8. The method of compensation control of a smart-compensated box-type substation according to claim 1, characterized in that, In step S4, updating the system equivalent short-circuit impedance parameter, the safety threshold, and the sudden change threshold based on the deviation between the actual voltage change and the predicted voltage change, and the actual reactive power change, specifically includes: Based on the actual voltage change and the actual reactive power change, calculate the measured short-circuit impedance value corresponding to the current action; The measured short-circuit impedance value and the original equivalent short-circuit impedance parameter of the system are weighted and smoothed by filtering, and the calculation result is used as the updated equivalent short-circuit impedance parameter of the system. If the absolute value of the deviation is greater than the preset error tolerance limit, the safety threshold is decreased by a preset first correction step size, and the mutation threshold is increased by a preset second correction step size.

9. A compensation control system for an intelligent-compensated box-type substation, the substation comprising a low-voltage bus and a plurality of capacitor compensation branches connected to the low-voltage bus, characterized in that, The compensation control system includes: The parameter acquisition module is used to acquire power grid operation parameters including three-phase voltage, three-phase current, reactive power calculation value and state quantities characterizing the degree of three-phase imbalance, as well as the operation status of each capacitor compensation branch. The distortion determination and suspension module is used to extract transient distortion features based on the power grid operating parameters and determine the current reactive power deficit; if the rate of change of the current reactive power deficit exceeds the abrupt change threshold and the transient distortion features exist, the compensation processing for the current reactive power deficit is suspended. The scheme deduction module is used to generate candidate compensation schemes if the compensation process for the current reactive power deficit is not suspended, and to calculate the predicted voltage change based on the system equivalent short-circuit impedance parameters, candidate compensation capacity and current bus voltage. A degradation or blocking module is used to perform degradation or blocking processing on the candidate compensation scheme if the predicted voltage change exceeds a safety threshold. The target scheme determination module is used to determine the target compensation scheme from the candidate compensation schemes or the processed candidate compensation schemes by combining the abnormal operation history of each of the capacitor compensation branches. The execution and acquisition module is used to control the capacitor compensation branch corresponding to the target compensation scheme to perform compensation actions after the target compensation scheme is determined, and to acquire the actual voltage change and actual reactive power change after reaching steady state. The parameter closed-loop update module is used to update the system equivalent short-circuit impedance parameter, the safety threshold, and the mutation threshold based on the deviation between the actual voltage change and the predicted voltage change, and the actual reactive power change.

10. A smart compensation box-type substation, characterized in that, It includes a housing, and a transformer, low-voltage busbar, data acquisition device, multiple capacitor compensation branches and controller installed inside the housing; The transformer and the plurality of capacitor compensation branches are all connected to the low-voltage busbar; The data acquisition device is connected to the low-voltage bus and the controller respectively; The controller is connected to each of the capacitor compensation branches, and the controller is configured to perform the compensation control method of the intelligent compensation box-type substation as described in any one of claims 1 to 8.

Images