Power quality comprehensive treatment device

Abstract

The present application relates to the technical field of new energy reactive power compensation, and provides a power quality comprehensive treatment device, which comprises an electric energy monitoring system, current sampling of a 10kV bus is realized through a current transformer, voltage signal collection is realized through a voltage relay, protection signal monitoring is realized through a signal relay, and comprehensive operation parameters including system voltage, current and harmonic characteristics are formed; a controllable static compensation system is used for driving capacitor to form a capacitive compensation branch based on the comprehensive operation parameters through intelligent switching control of a contactor, a series resonance suppression loop formed by an inductor is combined, a stepped reactive power compensation network with harmonic suppression function is established; a dynamic reactive power regulation system is used for connecting a static reactive power generator after realizing voltage matching through a transformer according to real-time compensation demand, and generating compensation current with opposite phase to the system reactive power component. The present application can realize fine adjustment, has high compensation precision and good compensation effect.

Description

Technical Field

[0001] This invention relates to the field of reactive power compensation technology for new energy sources, and in particular to a comprehensive power quality management device. Background Technology

[0002] Due to the rapid development of new energy sources in various regions, traditional complete sets of parallel capacitor banks (TBB) high-voltage reactive power compensation devices can no longer meet the new reactive power compensation requirements brought about by the grid connection of photovoltaic power generation. Only a combination of static and dynamic systems can meet the compensation requirements. TBB high-voltage reactive power compensation devices can only be fixed and switched in groups, with large level differences. When the load fluctuates rapidly, the switching speed cannot keep up, resulting in relatively poor compensation effect.

[0003] Existing technology 1, publication number CN120389394A, discloses a reactive power compensation control system and method based on artificial intelligence autonomous learning, including a data acquisition layer, an AI decision-making layer, a compensation execution layer, and a feedback optimization layer. The data acquisition layer includes a sensor module and a data acquisition module, with the sensor module used to collect current and voltage data in the circuit. While it can establish a reactive power prediction model through AI autonomous learning to achieve dynamic optimization of the compensation strategy; secondly, it can match the production load type in real time, i.e., a deviation <0.05 is judged as belonging to the same type of product, ensuring the accuracy of the compensation strategy; in addition, it can realize the instantaneous reactive power prediction of SVG and the steady-state compensation coordination control of TBB / TSC, improving the response speed and accuracy of hybrid compensation; finally, through post-compensation verification and model update mechanisms, it ensures the adaptability and stability of the system in long-term operation; however, relying on AI autonomous learning results in high system complexity, and the response speed is affected by the algorithm's computation time, which may not meet the ultra-fast compensation requirements of certain scenarios; continuous training and optimization of the model is required, leading to high maintenance costs.

[0004] Existing technology 2, publication number CN106374494A, discloses a hybrid compensation system and method for railway power supply networks, including a control system, a converter, a reactor, and a filter module. One end of the converter is grounded, and the other end is connected in series with the reactor and the filter module and then connected to the traction power supply network. The control system is used to detect the voltage and current characteristics of the traction power supply network and provide control pulses to the converter. Although it has advantages such as a wide harmonic compensation range, good filtering performance, large compensation capacity, high cost performance, good output voltage characteristics, and compensation performance unaffected by the grid frequency, easy implementation, and simultaneous dynamic compensation of harmonics and reactive power, making it suitable for high-voltage and high-power load applications, it is mainly designed for railway traction power supply networks, limiting its applicability and making it unsuitable for direct application to conventional 10kV distribution networks. Furthermore, harmonic compensation relies on fixed filter modules, making it difficult to adapt to dynamic changes in the harmonic content of the grid.

[0005] Prior art 3, publication number CN104362631B, discloses a control method for a high-power dynamic hybrid compensation system for reactive power and harmonics. This hybrid compensation system includes an SVG (Static Var Generator), a TSF (Transient Function Filter) group, and a controller. The SVG and each TSF are connected in parallel to the power grid. Using the three-phase current of the power grid between the SVG and the TSF group as the detection object, the controller processes the voltage and current values ​​collected by each sensor and obtains the TSF switching signal based on hybrid logic, thereby achieving graded compensation of most of the reactive power of the load and filtering out specific harmonics. Furthermore, the control parameters of the SVG are obtained based on instantaneous power theory, thereby continuously compensating for the remaining reactive power and harmonics of the load. Although it can completely eliminate the impact of passive device switching on the power grid and prevent its harmonic current from flowing into the power grid, it saves energy, reduces emissions, and is environmentally friendly, with good social and economic benefits and a wide range of applications, especially suitable for use in the oil drilling field; however, when using thyristor-switched filters with SVG+TSF hybrid compensation, there is still a certain mechanical response delay when TSF switching, and it is impossible to achieve completely shock-free dynamic compensation; harmonic filtering relies on TSF groups with fixed parameters, and its adaptability to harmonic spectrum changes is weak.

[0006] Current technologies 1, 2, and 3 suffer from contradictions between response speed and compensation accuracy, issues regarding the coordination of static and dynamic compensation, and the compatibility of harmonic suppression and reactive power compensation. Therefore, this invention provides a comprehensive power quality management device. Summary of the Invention

[0007] To address the aforementioned technical problems, the present invention provides a comprehensive power quality management device, comprising:

[0008] The dynamic reactive power regulation system is used to generate a compensation current with a phase opposite to the reactive power component of the system by means of a stepped reactive power compensation network based on real-time compensation needs, voltage matching through a transformer, and then connection to a static var generator through an isolating switch. The system utilizes the PWM control technology of the IGBT power module to generate a compensation current that is in the opposite phase to the reactive power component of the system, thus forming a dynamic compensation output with continuous adjustment capability.

[0009] Optional, dynamic reactive power regulation system, including:

[0010] The voltage conversion subsystem is used to acquire real-time operating data of the stepped reactive power compensation network. The transformer automatically adjusts the tap position according to the reactive power deficit to convert the voltage to the working voltage level that is compatible with the static var generator. After receiving the access command, the disconnecting switch first performs a pre-closure detection.

[0011] The current component compensation subsystem is used by the IGBT power module drive system to acquire the voltage and current waveform data of the current cycle. After processing by the instantaneous reactive power algorithm, the reactive current component that needs to be compensated is decomposed. Based on the reactive current component, three sets of time-varying modulation signals are generated. Each set of signals contains two control layers: fundamental compensation command and harmonic suppression command.

[0012] The switching sequence forming subsystem is used for three sets of time-varying modulation signals to enter the power unit for execution. The three sets of signals are staggered by 30 degrees in the time domain by carrier phase shifting to form an interleaved parallel switching sequence. The turn-on and turn-off time of each IGBT is determined by the intersection of the corresponding modulation signal and carrier. The generated pulse group is amplified by the gate drive to form a power output with a specific phase relationship.

[0013] The depth automatic adjustment subsystem is used to synthesize the pulse voltages output by each power unit through a filter reactor to form a continuous current waveform that is opposite in phase to the reactive current component. The continuous current waveform is compared with the target compensation amount in real time through closed-loop feedback, and the difference signal of the comparison is used to automatically adjust the PWM modulation depth.

[0014] Optional, a current component compensation subsystem includes:

[0015] The vector decomposition component is used to input synchronized voltage and current data into instantaneous power calculation and establish a dynamic reference coordinate system using the orthogonal decomposition principle. The rotation angle of the reference coordinate system is determined in real time by the positive sequence component of the voltage fundamental wave, forming an adaptive rotation calculation reference. The current vector is decomposed into two orthogonal projection components in the reference coordinate system.

[0016] The harmonic feature extraction component is used to extract the reactive current component, which then performs a spectrum scan on the reactive current; the scan results generate a harmonic energy distribution map.

[0017] The dynamic gain adjustment component is used to control the command generator to receive the decomposed reactive current components and harmonic distortion distribution information; to dynamically adjust the gain of the fundamental reactive component, the gain coefficient of which is determined by the ratio of the system reactive power deficit to the preset threshold, and to generate the basic compensation amplitude reference; at the same time, the harmonic components are processed by frequency domain weighting to suppress the preset characteristic harmonics.

[0018] The harmonic suppression component uses a layered coding structure for the three sets of time-varying modulated signals for output. The bottom layer of each set of three time-varying modulated signals is the fundamental compensation waveform, whose amplitude changes with the reference value after gain adjustment. The upper layer is superimposed with a phase-modulated harmonic suppression waveform, and the envelope of the suppression waveform matches the energy distribution of the characteristic harmonics. The three sets of time-varying modulated signals maintain a phase interlock relationship.

[0019] Optional, dynamic gain adjustment component, including:

[0020] The mode switching sub-component is used to obtain the separated quadrature reactive current components, compare the current reactive component amplitude with the preset reactive deficit threshold, and establish a dynamic adjustment coefficient calculation program.

[0021] The coefficient generation sub-component is used by the main channel to generate a coarse compensation coefficient based on the comparison results. The coarse compensation coefficient has a piecewise linear relationship with the deviation. The auxiliary channel introduces an inertia factor formed by historical compensation data to smooth the output of the main channel. The outputs of the two channels are weighted and superimposed at the fusion node. The weight value depends on the dynamic response requirements of the current system.

[0022] The dynamic correction sub-component is used to input the coarse gain coefficient after channel fusion into the anti-saturation limiting, and has an adaptive limiting boundary. The boundary value of the adaptive limiting boundary is dynamically corrected by the spectral purity index provided by the harmonic feature extraction component to obtain the dynamic gain coefficient.

[0023] The coupling operation sub-component generates dynamic gain coefficients which are then fed into the reference generation unit. The dynamic gain coefficients are coupled with the system's nominal compensation capacity. The output base compensation amplitude reference value has time-varying characteristics, and its rate of change is constrained by the system's transient response characteristics.

[0024] Optional, dynamically corrected subcomponent, including:

[0025] The initial amplitude limiting boundary value mapping module is used to acquire spectral purity index data from the harmonic energy distribution map, characterizing the current system's harmonic pollution level, including two key parameters: fundamental reactive power ratio and harmonic distortion rate. Boundary calculations are performed on these two key parameters, establishing a boundary amplitude benchmark value in the fundamental power ratio dimension and a boundary contraction coefficient in the harmonic distortion rate dimension. The fundamental reactive power ratio parameter in the spectral purity index is normalized and mapped to the initial amplitude limiting boundary value; simultaneously, the harmonic distortion rate parameter is logarithmically transformed to generate a boundary contraction factor.

[0026] The product operation module is used to multiply the initial amplitude limit boundary value with the boundary contraction factor to obtain the preliminary dynamic boundary; the boundary smoothing module combines historical boundary data to form an inertial adjustment effect; the adjustment trend correction of the preliminary dynamic boundary acceptance coefficient generation sub-component; when the compensation coefficient after the fusion of the main and auxiliary channels shows a continuous growth trend, the boundary value is preventively contracted.

[0027] The gain correction module generates dynamic limiting boundary values ​​which are then fed into the gain correction module. Coarse gain coefficients exceeding the boundary range are constrained to the boundary values, and a boundary over-limit flag is triggered. Information from the boundary over-limit flag is fed back to the mode switching subcomponent.

[0028] Optional, the initial limiting boundary value mapping module includes:

[0029] The conversion curve generation submodule receives the original fundamental reactive power ratio parameter from the harmonic feature extraction component and inputs it into the boundary reference generator, which has a configurable nonlinear conversion curve set inside.

[0030] The raw boundary value output submodule is used to input the standardized raw fundamental reactive power ratio data into the dynamic slope adjustment unit and synchronously receive the load rate; dynamically correct the slope characteristics of the conversion curve according to the load rate; and output the uncalibrated raw boundary value of the corrected conversion curve.

[0031] The boundary smoothing submodule is used to determine the reasonableness of the original boundary value entering the boundary. It calls the boundary records from the three furthest periods in the historical database to calculate the deviation between the current boundary value and the historical average. When the deviation exceeds the preset threshold, the boundary smoothing program is started, and a weighted average is used to bring the current boundary value closer to the historical median. Finally, the initial amplitude limit boundary value is output.

[0032] Optional, a boundary smoothing submodule, including:

[0033] A historical reference benchmark unit is established to call the boundary records of the most recent three operating cycles in the historical database and extract boundary value samples under the same operating conditions in each cycle; time decay weighted calculation is performed on the boundary value samples to generate a historical reference median with time-dependent characteristics;

[0034] The deviation measurement evaluation unit is used to compare the uncalibrated original boundary value with the obtained historical reference median. The difference comparison uses the relative deviation measurement method to calculate the percentage deviation of the current original boundary value from the historical median. When the percentage deviation exceeds the dynamic threshold range, the smoothing mechanism is triggered.

[0035] A progressive correction unit is implemented to perform multi-level weighted fusion processing on the original boundary value that triggers smoothing; the algebraic average of the current original boundary value and the historical reference median is taken as the intermediate value; the intermediate value is then weighted again with the historical reference median, and the weight coefficient is dynamically adjusted according to the magnitude of the previous deviation; a boundary change rate limit is applied to the result of the second weighting, so that the corrected original boundary value will not exceed the preset maximum adjustment range per step; after processing, the initial amplitude limit boundary value of the final output is obtained.

[0036] Optionally, a progressive correction unit is implemented, comprising:

[0037] A dynamic weight benchmark subunit is established to construct a nonlinear weight mapping based on the percentage amplitude difference; the previous deviation value is input into the mapping function to generate the initial weight allocation ratio; the generated initial weight allocation ratio is input into the weight compensation regulator and fine-tuned according to the characteristics of the current system operation stage.

[0038] A bidirectional weighted fusion sub-unit is executed, using the intermediate value obtained from the first round of processing as the current operating condition characteristic quantity, and the historical reference median value after compensation and adjustment as the stability benchmark quantity; the two are fused and calculated asymmetricly according to the final weight coefficient; the fusion result forms a second-weighted value with dynamic equilibrium characteristics;

[0039] An incremental safety constraint sub-unit is applied to input the quadratic weighted value into the boundary rate of change limiter and compare it with the final boundary value of the previous period.

[0040] Optionally, it also includes: a power monitoring system, which is used to sample the current of the 10kV bus through a current transformer, collect voltage signals through a voltage relay, and complete the protection signal monitoring in conjunction with a signal relay to form comprehensive operating parameters including system voltage, current and harmonic characteristics.

[0041] Optionally, it also includes: a controllable static compensation system, which, based on comprehensive operating parameters, drives capacitors to form capacitive compensation branches through intelligent switching control of contactors, and combines them with series resonant suppression circuits composed of inductors to establish a stepped reactive power compensation network with harmonic suppression function.

[0042] This invention employs a high-voltage reactive power dynamic hybrid compensation device combining a complete set of parallel capacitor compensation devices (TBB) and a static var generator (SVG). This avoids the disadvantages of the TBB and adds the advantages of the SVG. The SVG's dynamic compensation has a fast response speed, capable of compensating for loads with rapidly changing reactive power, while the static reactive power compensation of the TBB is used to compensate for loads with relatively stable reactive power and infrequent changes. The combined compensation allows for fine-tuning, high accuracy, and good results. This embodiment achieves fine-tuning, high accuracy, and good results; it improves the power factor through reactive power compensation, reducing the total supply current, improving power supply safety, and enhancing power quality; it effectively supports the load-side voltage, strengthening system voltage stability; it saves energy and reduces consumption; after the compensation device is put into operation, it increases the transformer's operating margin, reduces losses, and improves equipment utilization.

[0043] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0044] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0045] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0046] Figure 1 This is a block diagram of the power quality comprehensive management device in Embodiment 1 of the present invention;

[0047] Figure 2 This is a schematic diagram of the power quality comprehensive management device in Embodiment 1 of the present invention;

[0048] Figure 3 This is a block diagram of the power monitoring system in Embodiment 2 of the present invention;

[0049] Figure 4 This is a block diagram of the controllable static compensation system in Embodiment 5 of the present invention;

[0050] Figure 5 This is a block diagram of the dynamic reactive power regulation system in Embodiment 8 of the present invention;

[0051] Figure 6 This is a schematic diagram of the power quality comprehensive management device in Embodiment 15 of the present invention. Detailed Implementation

[0052] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0053] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to limit the embodiments of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0054] In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application. In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0055] Example 1: As Figure 1As shown, an embodiment of the present invention provides a comprehensive power quality management device, comprising:

[0056] The power monitoring system is used to sample the current of the 10kV bus through current transformers, acquire voltage signals through voltage relays, and complete protection signal monitoring in conjunction with signal relays, forming comprehensive operating parameters that include system voltage, current and harmonic characteristics.

[0057] The controllable static compensation system is used to drive capacitors to form capacitive compensation branches based on comprehensive operating parameters and intelligent switching control of contactors. Combined with series resonance suppression circuits formed by inductors, a stepped reactive power compensation network with harmonic suppression function is established.

[0058] The dynamic reactive power regulation system is used in stepped reactive power compensation networks. Based on real-time compensation requirements, after voltage matching is achieved through a transformer, it is connected to a static var generator through an isolating switch. Utilizing the PWM control technology of the IGBT power module, a compensation current with the opposite phase to the reactive component of the system is generated, forming a dynamic compensation output with continuous adjustment capability.

[0059] The working principle and beneficial effects of the above technical solution are as follows: The power monitoring system in this embodiment is used to sample the current of the 10kV bus through a current transformer, collect voltage signals through a voltage relay, and complete protection signal monitoring in conjunction with a signal relay, forming comprehensive operating parameters including system voltage, current, and harmonic characteristics; the controllable static compensation system is used to drive capacitors to form a capacitive compensation branch based on the comprehensive operating parameters, through intelligent switching control of contactors, and combined with a series resonance suppression circuit composed of inductors to establish a stepped reactive power compensation network with harmonic suppression function; the dynamic reactive power adjustment system is used to connect the stepped reactive power compensation network to a static var generator through a transformer after voltage matching according to real-time compensation requirements, and uses the PWM control technology of the IGBT power module to generate a compensation current with the opposite phase to the reactive power component of the system, forming a dynamic compensation output with continuous adjustment capability (the specific principle is as follows). Figure 2 (As shown in the diagram). The above scheme achieves comprehensive improvement in power quality. The power monitoring system continuously collects system operating parameters, providing data support for subsequent compensation; the controllable static compensation system establishes a basic reactive power compensation network based on these parameters, achieving stepped static compensation and suppressing harmonics; the dynamic reactive power regulation system performs refined dynamic regulation on this basis, further improving power quality through rapid response and continuous compensation. The organic combination of the three systems forms a complete comprehensive management scheme, ensuring both the stability of basic compensation and the accuracy of dynamic regulation, ultimately achieving comprehensive management effects such as improving the system power factor, suppressing voltage fluctuations, and filtering harmonic pollution, effectively improving the power grid's power supply quality and operating efficiency.

[0060] Example 2: As Figure 3 As shown, based on Embodiment 1, the power monitoring system provided in this embodiment of the invention includes:

[0061] The alignment and matching subsystem is used to sample the 10kV bus current throughout the entire cycle through the current transformer to obtain the raw current waveform data. At the same time, the voltage relay is used to obtain the instantaneous values ​​of the three-phase voltage. The two sets of voltage and current data enter the synchronization processing stage, and waveform alignment and time scale matching are completed within a time window of 10 milliseconds.

[0062] The harmonic distortion calculation subsystem is used for feature extraction of aligned and time-scaled voltage and current data. Based on an improved sliding window algorithm, it performs real-time spectrum calculation to extract the amplitude and phase information of the fundamental components of each phase voltage and current, and calculates the total harmonic distortion rate. At the same time, the protection status data provided by the signal relay is used as a correction factor in the calculation to ensure that the measurement results are automatically corrected under abnormal system conditions.

[0063] The parameter set formation subsystem is used to dynamically combine the fundamental component and harmonic components according to predetermined weighting coefficients. The voltage data adopts the root mean square synthesis method, and the current data adopts the peak hold method. During the fusion process, the participation weight of each harmonic is automatically adjusted according to the current load rate to form a preliminary operating parameter set containing 18-dimensional feature vectors such as fundamental voltage and current, 3rd-25th characteristic harmonic content, and three-phase unbalance.

[0064] The adaptive filtering subsystem is used to optimize the preliminary parameter set through adaptive filtering to eliminate the influence of measurement noise and transient interference. The optimized preliminary parameter set is compared with typical operating condition patterns in the historical database. The three most matching reference patterns are selected through similarity calculation, and their weighted average is taken as the final comprehensive operating parameter output.

[0065] The working principle and beneficial effects of the above technical solution are as follows: The alignment and matching subsystem of this embodiment is used to sample the 10kV bus current throughout the entire cycle through the current transformer to obtain the original current waveform data, and at the same time, the voltage relay is used to obtain the instantaneous values ​​of the three-phase voltage; the two sets of voltage and current data enter the synchronization processing stage, and waveform alignment and time-scale matching are completed within a time window of 10 milliseconds; the harmonic distortion calculation subsystem is used to extract the voltage and current data after alignment and time-scale matching, and to perform real-time spectrum calculation based on the improved sliding window algorithm to extract the amplitude and phase information of the fundamental components of each phase voltage and current, and at the same time calculate the total harmonic distortion rate; at the same time, the protection status data provided by the signal relay is used as a correction factor to participate in the calculation to ensure automatic correction in abnormal system conditions. The system generates positive measurement results. A parameter set formation subsystem dynamically combines the fundamental and harmonic components according to predetermined weighting coefficients. Voltage data uses the root mean square synthesis method, and current data uses the peak hold method. During the fusion process, the participation weights of each harmonic are automatically adjusted based on the current load rate, forming a preliminary operating parameter set containing 18-dimensional feature vectors, including fundamental voltage and current, 3rd-25th harmonic content, and three-phase imbalance. An adaptive filtering subsystem optimizes the preliminary parameter set through adaptive filtering, eliminating the influence of measurement noise and transient interference. The optimized preliminary parameter set is compared with typical operating conditions in the historical database. The three most matching reference modes are selected through similarity calculation, and their weighted average is used as the final comprehensive operating parameter output. This scheme achieves accurate monitoring and parameter extraction of the 10kV bus power quality. The alignment and matching subsystem ensures the spatiotemporal consistency of voltage and current data, providing an accurate basic data source for subsequent processing. The harmonic distortion calculation subsystem performs spectral decomposition on the synchronous data to obtain core characteristic quantities reflecting power quality. The parameter set formation subsystem integrates the decomposed features according to dynamic weights to construct a parameter set that comprehensively characterizes the system state. Its load rate adaptive weight adjustment mechanism improves the effectiveness of the parameters. The adaptive filtering subsystem optimizes the parameter set through pattern matching, and the final output comprehensive operating parameters eliminate measurement errors while retaining the characteristics of the actual operating conditions.

[0066] In summary, this embodiment transforms raw electrical quantities into a multi-dimensional parameter set, including fundamental characteristics, harmonic features, and three-phase balance, through a progressively advancing data processing flow. This provides accurate decision-making basis for subsequent compensation control. The data processing links formed by each subsystem have self-correcting capabilities, enhancing the system's anti-interference ability while ensuring measurement accuracy, and achieving efficient conversion from raw data to usable parameters.

[0067] Example 3: Based on Example 2, the harmonic distortion calculation subsystem provided in this embodiment of the invention includes:

[0068] A dual-layer storage component is used to acquire voltage and current data, which is then processed in 10-millisecond increments into a sliding window buffer. The sliding window buffer employs a dual-layer storage structure: the first layer retains the complete voltage and current waveform data for the current window, while the second layer stores the historical values ​​of characteristic parameters from the previous five windows.

[0069] At the beginning of each processing cycle, dynamic baseline calibration is performed on the original voltage and current waveforms of the first layer of the buffer. The calibration process uses the historical feature parameters stored in the second layer to predict the fundamental component of the current window and subtracts the fundamental trend component from the original voltage and current waveforms to obtain the net fluctuation data.

[0070] An automatic sampling component is used to input calibrated voltage and current waveform data into variable density sampling. The sampling strategy is dynamically adjusted according to the protection status data provided by the signal relay: uniform sampling is used in normal conditions, and the sampling density is automatically increased in the waveform distortion section in abnormal conditions.

[0071] The feature separation component is used to perform feature separation calculations after completing variable density sampling. It adopts recursive spectral analysis to extract each harmonic component sequentially starting from the low frequency band. Each harmonic extraction is based on the previous calculation result and error compensation is performed to form a recursive calculation chain.

[0072] The parameter calculation component is used to extract the fundamental component using a dual-channel verification mechanism. The main channel obtains the fundamental parameters based on the results of the recursive calculation chain; the auxiliary channel obtains the fundamental phase information by performing zero-crossing detection on the original voltage and current waveforms; the results of the two channels are cross-validated and the final fundamental component amplitude and phase are output; the total harmonic distortion rate is calculated using dynamic bandwidth integration, and the integration frequency band is automatically adjusted according to the current load rate, focusing on low-frequency integration under light load and expanding to high-frequency integration under heavy load.

[0073] The working principle and beneficial effects of the above technical solution are as follows: The dual-layer storage component in this embodiment is used to acquire voltage and current data, and enters the sliding window buffer with a basic processing unit of 10 milliseconds. The sliding window buffer adopts a dual-layer storage structure. The first layer retains the complete voltage and current waveform data of the current window, and the second layer stores the historical values ​​of the feature parameters of the previous five windows. At the beginning of each processing cycle, dynamic baseline calibration is performed on the original voltage and current waveform of the first layer of the buffer. The calibration process uses the historical feature parameters stored in the second layer to predict the fundamental component of the current window, and subtracts the fundamental trend component from the original voltage and current waveform to obtain the net fluctuation data. The automatic sampling component is used for the calibrated voltage and current waveform data to enter variable density sampling. The sampling strategy is dynamically adjusted according to the protection status data provided by the signal relay: uniform sampling is used in the normal state, and the sampling density is automatically increased for waveform distortion sections in the abnormal state; feature separation. The component performs feature separation calculations after completing variable-density sampling, employing recursive spectral analysis to extract harmonic components sequentially starting from the low-frequency band. Each harmonic extraction is based on the previous calculation result with error compensation, forming a recursive calculation chain. The parameter calculation component uses a dual-channel verification mechanism to extract the fundamental component. The main channel obtains the fundamental parameters based on the results of the recursive calculation chain; the auxiliary channel obtains the fundamental phase information by performing zero-crossing detection on the original voltage and current waveforms. The results from the two channels are cross-validated to output the final fundamental component amplitude and phase. The total harmonic distortion rate is calculated using dynamic bandwidth integration, automatically adjusting the integration frequency band according to the current load rate, emphasizing low-frequency integration under light load and expanding to high-frequency integration under heavy load. This scheme achieves refined and adaptive processing of power quality monitoring through modular collaborative work; the dual-layer storage structure constructs a dynamic reference benchmark, establishing historical data support for fundamental calibration and effectively separating the true fluctuation components; the variable-density sampling mechanism intelligently adjusts the data acquisition density according to the system status, ensuring the integrity of basic data while enhancing the ability to capture abnormal waveform features. Feature separation employs recursive spectral analysis to form a computational chain, and achieves progressive and accurate extraction of harmonic components through error compensation; a dual-channel verification mechanism provides dual protection for fundamental parameters, and cross-verification significantly improves the reliability of fundamental detection; the dynamic bandwidth integration algorithm automatically optimizes the frequency band range according to the load conditions, ensuring that the total harmonic distortion calculation always closely matches the actual power state.

[0074] In summary, this embodiment achieves accurate separation and quantitative assessment of harmonic components, providing a complete solution for power quality analysis from data acquisition and feature extraction to parameter calculation. The coordinated design of each module ensures that the system can handle basic monitoring under steady-state conditions as well as quickly respond to transient anomalies, forming a comprehensive detection capability that takes into account both routine analysis and handling of special situations.

[0075] Example 4: Based on Example 3, the parameter calculation component provided in this embodiment of the invention includes:

[0076] The correction factor establishment sub-component is used to take the fundamental amplitude obtained by the recursive calculation chain as the main channel input, and the phase angle obtained by zero-crossing detection as the auxiliary channel input. The main channel establishes the amplitude correction factor by accumulating the residual error of the preceding harmonic separation stage. The auxiliary channel calculates the phase drift by using the polarity jump point spacing of the original waveform.

[0077] The cross-validation sub-component is used when two data streams enter the cross-validation stage. It performs phase sensitivity detection on the amplitude correction factor of the main channel. If the phase drift of the auxiliary channel exceeds the threshold, amplitude recalibration is triggered. At this time, the zero cross-interval of the auxiliary channel is mapped to the time domain window of the main channel to realign the fundamental frequency period boundary. After the verification is successful, the fused fundamental frequency parameters are output. The amplitude is taken from the error compensation result of the main channel, and the phase is combined with the jump point distribution of the auxiliary channel and the spectral positioning of the main channel.

[0078] The bandwidth integration sub-component is used for dynamic bandwidth integration based on the fundamental frequency parameter. When the load rate is lower than the load threshold, the integration interval is limited to an integer multiple of the fundamental frequency, focusing on low-frequency energy accumulation. When the load rate increases, discrete spectral lines of high-frequency harmonic groups are introduced, and the integration upper limit is dynamically expanded by calculating the harmonic distribution patterns recorded in the recursive calculation chain. The high-frequency integration weight is positively correlated with the gradient of the higher harmonic amplitude output by the feature separation component within the current window, ensuring that the total distortion rate calculation always tracks the actual spectral energy distribution.

[0079] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the correction factor establishment sub-component uses the fundamental amplitude obtained by the recursive calculation chain as the main channel input, and the phase angle obtained by zero-crossing detection as the auxiliary channel input; the main channel establishes the amplitude correction factor by accumulating the residual error of the preceding harmonic separation stage; the auxiliary channel calculates the phase drift using the polarity jump point spacing of the original waveform; the cross-validation sub-component is used for the two data streams to enter the cross-validation stage, and performs phase sensitivity detection on the amplitude correction factor of the main channel. If the phase drift of the auxiliary channel exceeds the threshold, amplitude recalibration is triggered; at this time, the zero-crossing interval of the auxiliary channel is mapped to the time domain window of the main channel for re-alignment. The fundamental frequency period boundary is defined. After successful verification, the fused fundamental frequency parameters are output, with the amplitude taken from the error compensation result of the autonomous channel and the phase calculated by combining the jump point distribution of the auxiliary channel and the spectral positioning of the main channel. The bandwidth integration subcomponent is used for dynamic bandwidth integration based on the fundamental frequency parameters. When the load rate is below the load threshold, the integration interval is limited to integer multiples of the fundamental frequency, focusing on low-frequency energy accumulation. When the load rate increases, discrete spectral lines of high-frequency harmonic groups are introduced, and the integration upper limit is dynamically expanded by calculating the harmonic distribution patterns recorded in the recursive calculation chain. The high-frequency integration weight is positively correlated with the gradient of the higher-order harmonic amplitude output by the feature separation component within the current window, ensuring that the total distortion rate calculation always tracks the actual spectral energy distribution. The above scheme achieves accurate calculation and dynamic optimization of power harmonic parameters through multi-module collaborative work. The correction factor establishment subcomponent constructs a dual-channel data acquisition mechanism, laying the foundation for interactive verification of the main and auxiliary channels. The cross-validation subcomponent significantly improves the accuracy of fundamental frequency parameter extraction by mutually verifying amplitude and phase information, and its period boundary realignment function effectively solves the phase drift problem in traditional methods. The bandwidth integrator component achieves adaptive spectral energy integration based on the fundamental parameters and harmonic distribution patterns of the preceding output. Its dynamic adjustment mechanism enables the distortion rate calculation to automatically optimize the frequency band according to load changes, ensuring calculation accuracy under light loads while also taking into account the high-frequency component capture capability under heavy load conditions.

[0080] In summary, this embodiment forms a closed-loop processing flow: the correction factor provides preprocessed data for cross-validation, the validation results guide the parameter setting of the bandwidth integral, and the distortion characteristics of the integral output are fed back to the previous harmonic separation process; the bidirectional coupling of the data stream enables the system to have parameter self-optimization capabilities, and finally realizes the coordinated and accurate calculation of the fundamental component and harmonic distortion rate, providing a reliable quantitative basis for power quality analysis.

[0081] Example 5: Figure 4 As shown, based on Embodiment 1, the controllable static compensation system provided in this embodiment of the invention includes:

[0082] The compensation command issuing subsystem uses the collected voltage and current data to analyze harmonic characteristics, decompose the fundamental reactive component and the content of each harmonic; based on the current load conditions, it determines the capacitive compensation demand and harmonic suppression priority, and generates segmented compensation commands.

[0083] The switching strategy initiation subsystem is used to initiate the intelligent switching strategy after receiving the segmented compensation command; for capacitive compensation branches, capacitor banks that match the fundamental reactive power demand are prioritized to form a basic compensation layer.

[0084] The automatic parameter adjustment subsystem is used to automatically adjust the inductor parameters of the series resonant suppression circuit according to the harmonic characteristics, so that it presents high impedance characteristics at the target harmonic frequency. When a certain harmonic exceeds the standard, a capacitor with a specific capacitance value is added and switched, so that the capacitive compensation branch and the inductor form a parallel resonance at the harmonic frequency, and the harmonic current is diverted to the internal consumption of the stepped reactive power compensation network.

[0085] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the compensation command issuing subsystem collects voltage and current data, analyzes harmonic characteristics, and decomposes the fundamental reactive component and the content of each harmonic. Based on the current load conditions, it determines the capacitive compensation demand and harmonic suppression priority, generating segmented compensation commands. The switching strategy initiation subsystem receives the segmented compensation commands and initiates an intelligent switching strategy. For capacitive compensation branches, capacitor banks matching the fundamental reactive demand are prioritized for activation, forming a basic compensation layer. The automatic parameter adjustment subsystem automatically adjusts the inductor parameters of the series resonant suppression circuit according to harmonic characteristics, making it exhibit high impedance characteristics at the target harmonic frequency. When a harmonic exceeds the standard, a capacitor with a specific capacitance value is added for switching, causing the capacitive compensation branch and the inductor to form parallel resonance at the harmonic frequency, diverting the harmonic current to the internal consumption of the stepped reactive power compensation network. The above solution achieves comprehensive power quality management through a layered and progressive control strategy. The compensation command issuing subsystem accurately identifies the operating conditions, and its segmented compensation commands provide decision-making basis for subsequent modules. The switching strategy's finite element system constructs a basic compensation architecture based on instructions. Through the coordinated operation of capacitive reactance branches and resonant suppression circuits, it satisfies both the fundamental reactive power compensation requirement and achieves selective suppression of specific harmonic frequencies. This ultimately forms a composite compensation scheme that balances economy and accuracy, achieving broadband harmonic mitigation while ensuring reactive power balance.

[0086] Example 6: Based on Example 5, the automatic parameter adjustment subsystem provided in this embodiment of the invention includes:

[0087] The spectral feature extraction component is used to trigger the deep harmonic governance mode when a specific harmonic component is found to exceed the standard (a harmonic component threshold is set); after the voltage and current data are separated by harmonic features, the energy distribution spectrum of the target harmonic frequency band is extracted into the compensation decision matrix;

[0088] The matching interval calculation component is used to mark the harmonic exceeding frequency band as a red warning area. At the same time, based on the fundamental reactive component decomposed in the previous stage, the resonant capacity matching interval corresponding to the exceeding frequency band is obtained. The value of the resonant capacity matching interval comes from the preset discretized capacitive reactance curve in the capacitor bank parameter library. Three sets of candidate capacitor values ​​are selected by the binary approximation algorithm.

[0089] The optimal capacitance value combination component is used to activate the fast response channel of the intelligent switching strategy start-up subsystem. It calls the topology information of the capacitor bank already invested in the basic compensation layer, and combines it with the real-time impedance characteristic curve of the inductor to construct a variable resonant network model. Through dynamic impedance matching, it locks the optimal capacitance value combination among the candidate capacitance values. It also satisfies two conditions: exhibiting capacitive compensation characteristics at the fundamental frequency and forming a precise parallel resonance with the inductor at the target harmonic frequency.

[0090] The phased execution governance component is used to initiate microsecond-level timing control and performs governance in two phases: In the first phase, the high-speed vacuum switch group is triggered to connect the capacitor with the selected capacitance value to the stepped reactive power compensation network. At this time, the power electronic interface unit monitors the transient process of the connection point in real time; In the second phase, when the phase angle of the network impedance is detected to reach the resonance critical point, the energy stored in the buffer reactor is released. The stored energy is the magnetic energy accumulated during the previous inductor parameter adjustment.

[0091] The working principle and beneficial effects of the above technical solution are as follows: The spectrum feature extraction component in this embodiment is used to trigger the deep harmonic governance mode when a specific harmonic component exceeds the standard (a harmonic component threshold is set); after harmonic feature separation, the energy distribution spectrum of the target harmonic frequency band is extracted into the compensation decision matrix; the matching interval calculation component is used to mark the harmonic exceeding frequency band as a red warning area, and simultaneously, based on the previously decomposed fundamental reactive component, to obtain the resonant capacity matching interval corresponding to the exceeding frequency band. The value of the resonant capacity matching interval comes from the preset discretized capacitive reactance curve in the capacitor bank parameter library, and three sets of candidate capacitor values ​​are selected through a binary approximation algorithm; the optimal capacitance value combination component is used to activate the fast response channel of the intelligent switching strategy start-up subsystem, calling... The topology information of the capacitor banks already deployed in the basic compensation layer, combined with the real-time impedance characteristic curves of the inductors, constructs a variable resonant network model. Through dynamic impedance matching, the optimal capacitance value combination is locked from the candidate capacitance values. This simultaneously satisfies two conditions: exhibiting capacitive compensation characteristics at the fundamental frequency and forming precise parallel resonance with the inductors at the target harmonic frequency. A phased execution governance component initiates microsecond-level timing control, executing governance in two phases: the first phase triggers a high-speed vacuum switch group, connecting the selected capacitance value capacitor to the stepped reactive power compensation network, while the power electronic interface unit monitors the transient process at the connection point in real time; the second phase, when the network impedance phase angle reaches the resonance critical point, releases the energy stored in the buffer reactor, which is the magnetic energy accumulated during the previous inductor parameter adjustment. The collaborative work of the components in this scheme achieves fully automated processing from detection to suppression. The compensation decision matrix established by the spectral feature extraction component provides the entire system with precise target positioning capabilities, enabling subsequent processing to address harmonic issues in specific frequency bands. The resonant capacity matching interval generated by the matching interval calculation component determines the parameter boundaries of the governance operation, ensuring that subsequent actions are always performed within the effective range. The variable resonant network model constructed using the optimal capacitance combination components achieves dynamic adaptation of the compensation device parameters to the grid state. Its adaptive characteristics enable the system to simultaneously meet the dual requirements of fundamental wave compensation and harmonic suppression. The phased control mechanism of the phased execution governance components ensures a smooth transition during the governance process, avoiding grid oscillations caused by sudden parameter changes.

[0092] In summary, the chain-like coordination in this embodiment forms a tiered governance structure: first, the problem is identified through feature analysis; then, a solution is determined through parameter calculation; and finally, the solution is implemented step by step to ensure effectiveness. The entire process achieves precise diversion and local consumption of harmonic energy while maintaining the system's basic reactive power compensation function. The system's self-adjustment capability stems from the data flow between components; the output of the preceding module always serves as the decision-making basis for the following module, forming a complete governance closed loop.

[0093] Example 7: Based on Example 6, the optimal capacitance combination component provided in this embodiment of the invention includes:

[0094] The information curve acquisition sub-component is used to call the capacitor bank topology information recorded by the basic compensation layer when constructing the variable resonant network model. This includes the switching status of each branch, capacitance parameters, and historical operating data. At the same time, it collects the impedance characteristic curve of the inductor under the current operating conditions, obtains real-time data points through high-frequency sampling, and forms a continuous impedance spectrum after cubic spline interpolation.

[0095] The dual-layer mapping sub-component is used to input the capacitor bank topology information and inductor impedance characteristic curves into the network reconstruction engine and process them using a dual-layer mapping mechanism. The first layer mapping establishes the equivalent capacitive reactance network at the fundamental frequency and automatically corrects the equivalent connection mode of the capacitor bank based on the fundamental reactive component obtained from the previous harmonic characteristics. The second layer mapping establishes a dynamic impedance matching matrix within the candidate capacitance value range for the target harmonic frequency band.

[0096] The virtual switching simulation sub-component is used for cross-validation of the dynamic impedance matching process. It substitutes three sets of candidate capacitor values ​​into the variable resonant network for virtual switching simulation. Each simulation includes two verification steps: the first step verifies the fundamental compensation effect by calculating the trend of the overall power factor of the system after the capacitor value is switched; the second step verifies the harmonic suppression effect by calculating the percentage of residual energy of the target harmonic under this configuration through inverse fast Fourier transform.

[0097] The comprehensive evaluation sub-component is used by the optimal capacitance value filter to receive all the output data from the virtual simulation and apply a weighted score to comprehensively evaluate each group of candidate values. The scoring criteria are the parameters of the red warning area in the compensation decision matrix, of which harmonic suppression accounts for 70% and fundamental compensation accounts for 30%. The scoring process incorporates the aging coefficient in the capacitor bank topology information to compensate for the capacity decay of capacitor units that have been operating for a long time.

[0098] The configuration scheme generation sub-component is used to generate a configuration scheme containing capacitance sequence and switching timing after locking the optimal combination; the final verification is performed through the impedance phase calibrator, referring to the real-time energy storage state of the buffer reactor during the verification; if the verification passes, an execution instruction with a timestamp is generated; if the verification fails, the process returns to the matching interval calculation component to readjust the resonant capacity matching interval range.

[0099] The working principle and beneficial effects of the above technical solution are as follows: The information curve acquisition sub-component of this embodiment is used to call the capacitor bank topology information recorded by the basic compensation layer when constructing the variable resonant network model. This information includes the switching status of each branch, capacitance parameters, and historical operating data. Simultaneously, it acquires the impedance characteristic curve of the inductor under the current operating condition, obtains real-time data points through high-frequency sampling, and forms a continuous impedance spectrum after cubic spline interpolation. The dual-layer mapping operation sub-component is used to input the capacitor bank topology information and the inductor impedance characteristic curve into the network reconstruction engine, and processes them using a dual-layer mapping mechanism. The first layer mapping establishes an equivalent capacitive reactance network at the fundamental frequency, and automatically corrects the equivalent connection method of the capacitor bank based on the fundamental reactive component obtained from the previous harmonic characteristics separation. The second layer mapping establishes a dynamic impedance matching matrix within the candidate capacitance value range for the target harmonic frequency band. The virtual switching simulation sub-component uses cross-validation in the dynamic impedance matching process, substituting the three sets of candidate capacitance values ​​into the variable resonant network for virtual switching simulation. Each simulation includes two... The verification process consists of two parts: First, the fundamental frequency compensation effect is verified by calculating the overall power factor trend after the capacitor value is added. Second, the harmonic suppression effect is verified by calculating the residual energy percentage of the target harmonic under this configuration using inverse Fast Fourier Transform. The comprehensive evaluation sub-component receives all virtual simulation output data from the optimal capacitance value filter and applies weighted scoring to comprehensively evaluate each candidate value. The scoring criteria are the parameters in the red warning area of ​​the compensation decision matrix, with harmonic suppression accounting for 70% and fundamental frequency compensation accounting for 30%. The scoring process incorporates the aging coefficient from the capacitor bank topology information to compensate for capacity attenuation in long-term operating capacitor units. The configuration scheme generation sub-component generates a configuration scheme containing capacitance value sequences and switching timings after locking in the optimal combination. Final verification is performed using an impedance phase calibrator, referencing the real-time energy storage state of the buffer reactor. If the verification passes, an execution instruction with a timestamp is generated; if the verification fails, the process returns to the matching interval calculation component to readjust the resonant capacity matching interval range. The various modules of the above optimal capacitance value combination component work collaboratively to form a complete intelligent decision-making system for capacitor parameters. The information curve acquisition subcomponent is the basic data acquisition module, providing accurate network status information for subsequent processing. The dual-layer mapping operation subcomponent, through a hierarchical processing mechanism, achieves parameter pre-screening for both fundamental compensation and harmonic suppression objectives. The first layer of mapping ensures the basic function of reactive power compensation, while the second layer establishes a refined adjustment framework for specific harmonic issues. The virtual switching simulation subcomponent provides parameter verification functionality, simulating the operational effects of different capacitor combinations to avoid potential system shocks from direct switching. Fundamental compensation effect verification maintains the basic requirement for stable grid operation, while harmonic suppression effect verification specifically optimizes power quality. The comprehensive evaluation subcomponent introduces a weighted scoring mechanism, ensuring the decision-making process aligns with actual governance needs; a higher harmonic suppression weight reflects the system's emphasis on power quality optimization.The introduction of an aging factor enhances the adaptability of parameter adjustments and ensures long-term operational reliability. The configuration scheme generation sub-component completes final parameter confirmation, and impedance phase calibration ensures the feasibility of the scheme. A verification mechanism forms a closed-loop control, preventing invalid parameters from entering the execution phase. Feedback adjustments in case of verification failure maintain the fault tolerance of the entire system.

[0100] In summary, this embodiment provides a precise parameter optimization scheme for harmonic problems while ensuring basic reactive power compensation. It reduces the actual switching risk through simulation verification, achieves optimal decision-making through intelligent scoring, and finally outputs a safe and reliable capacitor configuration scheme, realizing efficient and comprehensive management of power quality.

[0101] Example 8: As Figure 5 As shown, based on Embodiment 1, the dynamic reactive power regulation system provided in this embodiment of the invention includes:

[0102] The voltage conversion subsystem is used to acquire real-time operating data of the stepped reactive power compensation network, including the current system reactive power deficit, voltage fluctuation range, and harmonic distortion rate; the transformer automatically adjusts the tap position according to the reactive power deficit to convert the voltage to the working voltage level suitable for the static var generator; after receiving the access command, the disconnecting switch first performs a pre-closure test; and establishes an electrical connection between the main circuit of the static var generator and the secondary side of the transformer.

[0103] The current component compensation subsystem is used by the IGBT power module drive system to acquire the voltage and current waveform data of the current cycle. After processing by the instantaneous reactive power algorithm, the reactive current component that needs to be compensated is decomposed. Based on the reactive current component, three sets of time-varying modulation signals are generated. Each set of signals contains two control layers: fundamental compensation command and harmonic suppression command.

[0104] The switching sequence forming subsystem is used for three sets of time-varying modulation signals to enter the power unit for execution. The three sets of signals are staggered by 30 degrees in the time domain by carrier phase shifting to form an interleaved parallel switching sequence. The turn-on and turn-off time of each IGBT is determined by the intersection of the corresponding modulation signal and carrier. The generated pulse group is amplified by the gate drive to form a power output with a specific phase relationship.

[0105] The depth automatic adjustment subsystem is used to synthesize the pulse voltages output by each power unit through a filter reactor to form a continuous current waveform that is opposite in phase to the reactive current component. The continuous current waveform is compared with the target compensation amount in real time through closed-loop feedback, and the difference signal of the comparison is used to automatically adjust the PWM modulation depth.

[0106] The working principle and beneficial effects of the above technical solution are as follows: The voltage conversion subsystem of this embodiment is used to acquire real-time operating data of the stepped reactive power compensation network, including the current system reactive power deficit, voltage fluctuation range, and harmonic distortion rate; the transformer automatically adjusts the tap position according to the reactive power deficit to convert the voltage to a working voltage level suitable for the static var generator; after receiving the access command, the disconnecting switch first performs a pre-closure detection; an electrical connection is established between the main circuit of the static var generator and the secondary side of the transformer; the current component compensation subsystem is used by the IGBT power module drive system to acquire the voltage and current waveform data of the current cycle, and after processing by the instantaneous reactive power algorithm, decomposes the reactive current component that needs to be compensated; three sets of time-varying modulation are generated based on the reactive current component. The system comprises two control layers: a fundamental compensation command and a harmonic suppression command. A switching sequence formation subsystem handles the execution of three time-varying modulation signals by shifting the carrier phase, creating an interleaved parallel switching sequence. The on / off timing of each IGBT is determined by the intersection of the corresponding modulation signal and the carrier. The resulting pulse group is amplified by the gate drive to form a power output with a specific phase relationship. An automatic depth adjustment subsystem synthesizes the pulse voltages output from each power unit through a filter reactor, forming a continuous current waveform with a phase opposite to the reactive current component. This continuous current waveform is compared in real-time with the target compensation amount via closed-loop feedback, and the difference signal automatically adjusts the PWM modulation depth. The various modules of this dynamic reactive power regulation system work collaboratively to form a complete reactive power compensation and harmonic suppression solution. The voltage conversion subsystem ensures the safe and reliable grid connection of the static var generator (SVM) through transformer tap adjustment and isolation detection, providing stable operating conditions for subsequent compensation. The current component compensation subsystem accurately decomposes the reactive current component of the system and generates dual control commands that include fundamental compensation and harmonic suppression, enabling the compensation current to simultaneously meet the requirements of reactive power balance and harmonic filtering. The switching sequence formation subsystem employs a specific phase-interleaved parallel switching strategy to optimize the switching timing of power devices, reduce output harmonic content, and improve the system's dynamic response. The deep automatic adjustment subsystem uses a closed-loop feedback mechanism to correct the compensation current waveform in real time, ensuring that the output always accurately tracks the target compensation amount, achieving continuously optimized dynamic adjustment.

[0107] In summary, this embodiment accurately identifies reactive power demand, generates efficient compensation strategies, and achieves dynamic reactive power compensation and harmonic suppression through intelligent closed-loop regulation, thereby improving power quality and operational stability while ensuring safe equipment access.

[0108] Example 9: Based on Example 8, the current component compensation subsystem provided in this embodiment of the invention includes:

[0109] The vector decomposition component is used to input synchronized voltage and current data into instantaneous power calculation and establish a dynamic reference coordinate system using the orthogonal decomposition principle. The rotation angle of the reference coordinate system is determined in real time by the positive sequence component of the voltage fundamental wave, forming an adaptive rotation calculation reference. The current vector is decomposed into two orthogonal projection components in the reference coordinate system, of which the component perpendicular to the voltage vector is identified as the reactive current component to be compensated.

[0110] The harmonic feature extraction component is used to extract the reactive current component, and then perform a spectrum scan on the reactive current. The scan result generates a harmonic energy distribution map, in which the fundamental frequency region corresponds to the traditional reactive power and the high frequency region reflects the harmonic distortion component.

[0111] The dynamic gain adjustment component is used to control the command generator to receive the decomposed reactive current components and harmonic distortion distribution information; to dynamically adjust the gain of the fundamental reactive component, the gain coefficient of which is determined by the ratio of the system reactive power deficit to the preset threshold, and to generate the basic compensation amplitude reference; at the same time, the harmonic components are processed by frequency domain weighting to suppress the preset characteristic harmonics.

[0112] The harmonic suppression component uses a layered coding structure for the three sets of time-varying modulated signals for output. The bottom layer of each set of three time-varying modulated signals is the fundamental compensation waveform, whose amplitude changes with the reference value after gain adjustment. The upper layer is superimposed with a phase-modulated harmonic suppression waveform, and the envelope of the suppression waveform matches the energy distribution of the characteristic harmonics. The three sets of time-varying modulated signals maintain a phase interlock relationship.

[0113] The working principle and beneficial effects of the above technical solution are as follows: The vector decomposition component in this embodiment is used to input the synchronized voltage and current data into instantaneous power calculation, and to establish a dynamic reference coordinate system using the orthogonal decomposition principle; the rotation angle of the reference coordinate system is determined in real time by the positive sequence component of the fundamental voltage wave, forming an adaptive rotation calculation benchmark; the current vector is decomposed into two orthogonal projection components in the reference coordinate system, of which the component perpendicular to the voltage vector is identified as the reactive current component to be compensated; the harmonic feature extraction component is used for the reactive current component to then enter the harmonic feature extraction, and to perform a spectrum scan on the reactive current; the scan result generates a harmonic energy distribution map, in which the fundamental frequency region corresponds to the traditional reactive power, and the high-frequency region reflects the harmonic distortion component; dynamic The dynamic gain adjustment component is used to control the command generator to receive the decomposed reactive current components and harmonic distortion distribution information; it dynamically adjusts the gain of the fundamental reactive component, with the gain coefficient determined by the ratio of the system's reactive power deficit to a preset threshold, generating a basic compensation amplitude reference; simultaneously, the harmonic components undergo frequency domain weighting processing, focusing on suppressing pre-set characteristic harmonics; the harmonic suppression component outputs three sets of time-varying modulation signals using a layered coding structure. The bottom layer of each set of three time-varying modulation signals is the fundamental compensation waveform, whose amplitude changes with the reference value after gain adjustment; the upper layer is superimposed with a phase-modulated harmonic suppression waveform, whose envelope matches the energy distribution of the characteristic harmonics; the three sets of time-varying modulation signals maintain a phase interlocking relationship. The current component compensation subsystem of the above scheme achieves precise power quality management through modular collaboration: the dynamic reference coordinate system constructed by the vector decomposition component realizes real-time separation of current components, providing an adaptive calculation reference for the system, ensuring accurate extraction of reactive components under different operating conditions. The harmonic feature extraction component establishes an energy distribution map through spectrum analysis, visualizing the requirements of traditional reactive power compensation and harmonic mitigation, providing data support for subsequent targeted compensation. The dynamic gain adjustment component intelligently adjusts the compensation amount, dynamically correcting the fundamental wave compensation intensity according to the actual system needs, while highlighting key harmonic mitigation priorities through frequency domain weighting. The harmonic suppression component adopts a hierarchical coding structure, ensuring the continuity of fundamental wave compensation while achieving targeted suppression of characteristic harmonics through envelope matching of phase modulation. Phase interlocking of the three-phase signals ensures system stability.

[0114] In summary, this embodiment first deconstructs the characteristics of the current components, then designs differentiated compensation strategies for different frequency bands, and finally outputs a composite modulation waveform, achieving the dual objectives of dynamic compensation of the fundamental reactive power and directional suppression of characteristic harmonics. The system tracks voltage and current changes in real time, ensuring that the compensation strategy remains synchronized with the actual needs of the power grid.

[0115] Example 10: Based on Example 9, the dynamic gain adjustment component provided in this embodiment of the invention includes:

[0116] The mode switching sub-component is used to acquire the separated orthogonal reactive current components, compare the current reactive component amplitude with the preset reactive deficit threshold, establish a dynamic adjustment coefficient calculation program, and adopt a nonlinear mapping relationship. When the deviation between the measured value and the threshold is within a small range, a smooth adjustment characteristic is adopted, while in the large deviation range, it automatically switches to a fast response mode.

[0117] The coefficient generation sub-component is used by the main channel to generate a coarse compensation coefficient based on the comparison results. The coarse compensation coefficient has a piecewise linear relationship with the deviation. The auxiliary channel introduces an inertia factor formed by historical compensation data to smooth the output of the main channel. The outputs of the two channels are weighted and superimposed at the fusion node. The weight value depends on the dynamic response requirements of the current system.

[0118] The dynamic correction sub-component is used to apply the coarse gain coefficient after channel fusion to anti-saturation limiting, and has an adaptive limiting boundary. The boundary value of the adaptive limiting boundary is dynamically corrected by the spectral purity index provided by the harmonic feature extraction component to obtain the dynamic gain coefficient. When the harmonic content of the system is high, the adjustment range is automatically reduced to prevent the risk of resonance caused by gain adjustment.

[0119] The coupling operation sub-component generates dynamic gain coefficients which are then fed into the reference generation unit. The dynamic gain coefficients are coupled with the system's nominal compensation capacity. The coupling process takes into account the influence of voltage fluctuation coefficients, which are derived from the voltage positive sequence component amplitude information in the vector decomposition component. The output base compensation amplitude reference value has time-varying characteristics, and its rate of change is constrained by the system's transient response characteristics.

[0120] The working principle and beneficial effects of the above technical solution are as follows: The mode switching subcomponent of this embodiment is used to acquire the separated orthogonal reactive current components, compare the current reactive component amplitude with the preset reactive deficit threshold, establish a dynamic adjustment coefficient calculation program, and adopt a nonlinear mapping relationship. When the deviation between the measured value and the threshold is within a small range, a smooth adjustment characteristic is adopted, while in the large deviation range, it automatically switches to a fast response mode; the coefficient generation subcomponent is used by the main channel to generate a coarse compensation coefficient based on the comparison result. The coarse compensation coefficient has a piecewise linear relationship with the deviation; the auxiliary channel introduces an inertia factor formed by historical compensation data to smooth the output of the main channel; the outputs of the two channels are weighted and superimposed at the fusion node, and the weight value depends on the dynamic response requirements of the current system. The dynamic correction subcomponent is used to input the coarse gain coefficient after channel fusion processing into the anti-saturation limiting, with an adaptive limiting boundary. The boundary value of the adaptive limiting boundary is dynamically corrected by the spectral purity index provided by the harmonic feature extraction component to obtain the dynamic gain coefficient. When the system harmonic content is high, the adjustment range is automatically narrowed to prevent the risk of resonance caused by gain adjustment. The coupling operation subcomponent is used to send the generated dynamic gain coefficient to the reference generation unit, and the dynamic gain coefficient is coupled with the nominal compensation capacity of the system. The coupling process considers the influence of the voltage fluctuation coefficient, which comes from the voltage positive sequence component amplitude information in the vector decomposition component. The output basic compensation amplitude reference value has time-varying characteristics, and its rate of change is constrained by the transient response characteristics of the system. The dynamic gain adjustment component of the above scheme achieves precise adjustment of reactive power compensation through multi-level collaborative control: the mode switching subcomponent establishes a nonlinear response mechanism, automatically adjusts the adjustment characteristics according to the reactive power deficit deviation, maintains system stability in the small deviation range, and responds quickly in the large deviation range to ensure the dynamic adaptability of compensation. The coefficient generation subcomponent combines primary and secondary dual-channel regulation. The primary channel generates preliminary compensation coefficients based on real-time deviation, while the secondary channel incorporates historical data for inertial correction. Weighted fusion balances dynamic response and smoothness, preventing regulation oscillations. The dynamic correction subcomponent suppresses over-adjustment risk through adaptive limiting and dynamically adjusts the gain range using harmonic spectrum purity indicators, achieving synergistic optimization of harmonic suppression and reactive power compensation, preventing resonance problems caused by gain adjustment. The coupling operation subcomponent integrates dynamic gain coefficients and voltage fluctuation factors, combining them with nominal compensation capacity to generate time-varying reference values. This ensures that compensation commands match the actual system operating conditions while constraining the rate of change to maintain transient stability.

[0121] In summary, this embodiment progressively addresses the deviation detection and reference output, ensuring that the reactive power compensation can quickly respond to system demands while also considering operational stability and harmonic adaptability. The final output reference value accurately matches the current system state, providing a reliable compensation basis for subsequent harmonic suppression components.

[0122] Example 11: Based on Example 10, the dynamic correction sub-component provided in this embodiment of the invention includes:

[0123] The initial amplitude limiting boundary value mapping module is used to acquire spectral purity index data from the harmonic energy distribution map, characterizing the current system's harmonic pollution level, including two key parameters: fundamental reactive power ratio and harmonic distortion rate. Boundary calculations are performed on these two key parameters, establishing a boundary amplitude benchmark value in the fundamental power ratio dimension and a boundary contraction coefficient in the harmonic distortion rate dimension. The fundamental reactive power ratio parameter in the spectral purity index is normalized and mapped to the initial amplitude limiting boundary value. Simultaneously, the harmonic distortion rate parameter undergoes logarithmic transformation to generate a boundary contraction factor. An adjustable curvature coefficient is introduced during the transformation process to make the contraction effect more significant at high harmonic content.

[0124] The product operation module is used to multiply the initial amplitude limit boundary value with the boundary contraction factor to obtain the preliminary dynamic boundary; the boundary smoothing module combines historical boundary data to form an inertial adjustment effect; the adjustment trend correction of the preliminary dynamic boundary acceptance coefficient generation sub-component; when the compensation coefficient after the fusion of the main and auxiliary channels shows a continuous growth trend, the boundary value is preventively contracted; otherwise, the boundary constraints are appropriately relaxed; the magnitude of the correction is proportional to the duration and intensity of the trend.

[0125] The gain correction module generates dynamic limiting boundary values ​​which are then fed into the gain correction module. Coarse gain coefficients exceeding the boundary range are constrained to the boundary values, and a boundary over-limit flag is triggered. Information from the boundary over-limit flag is fed back to the mode switching subcomponent, prompting the system to adjust its response mode, thus forming a cross-component coordinated control mechanism.

[0126] The working principle and beneficial effects of the above technical solution are as follows: The initial amplitude limiting boundary value mapping module of this embodiment is used to obtain the spectral purity index data in the harmonic energy distribution spectrum, which characterizes the degree of harmonic pollution of the current system, including two key parameters: fundamental reactive power ratio and harmonic distortion rate; the two key parameters are calculated on the boundary, and a boundary amplitude benchmark value is established in the dimension of fundamental power ratio, and a boundary contraction coefficient is established in the dimension of harmonic distortion rate; the fundamental reactive power ratio parameter in the spectral purity index is normalized and mapped to the initial amplitude limiting boundary value; at the same time, the harmonic distortion rate parameter is logarithmically transformed to generate a boundary contraction factor, and an adjustable curvature coefficient is introduced in the transformation process to make the contraction effect more significant when the harmonic content is high; the product operation module is used to convert the initial... The initial dynamic boundary is obtained by multiplying the limiting boundary value and the boundary contraction factor. Boundary smoothing, combined with historical boundary data, forms an inertial adjustment effect. The initial dynamic boundary acceptor coefficient generation subcomponent corrects the adjustment trend. When the compensation coefficient after the fusion of the main and auxiliary channels shows a continuous increasing trend, the boundary value is preventively contracted; otherwise, the boundary constraints are appropriately relaxed. The magnitude of the correction is proportional to the duration and intensity of the trend. The dynamically limiting boundary value generated by the gain correction module is fed into the gain correction module. Coarse gain coefficients exceeding the boundary range are constrained to the boundary value, simultaneously triggering a boundary over-limit flag. The boundary over-limit flag information is fed back to the mode switching subcomponent, prompting the system to adjust its response mode, forming a cross-component coordinated control mechanism. The dynamic correction subcomponent of the above scheme achieves safe boundary management of gain adjustment through multi-level coordinated control: the initial limiting boundary value mapping module establishes a basic boundary constraint mechanism based on harmonic feature extraction results through two-parameter coupling calculations: the fundamental reactive power ratio determines the boundary reference value, and the harmonic distortion rate controls the degree of boundary contraction, forming initial boundary conditions reflecting the harmonic characteristics of the current system. The product operation module transforms static boundaries into dynamic constraints, smooths boundary continuity using historical data, and achieves forward-looking correction by adjusting the trend of compensation coefficients, ensuring that boundary values ​​both track system state changes and avoid drastic fluctuations. The gain correction module executes the final boundary constraints, safely truncates excessive gains, and achieves cross-component coordination through flag feedback, ensuring that the overall control strategy is adjusted when the system approaches the stability boundary.

[0127] In summary, this embodiment achieves adaptive gain adjustment under harmonic conditions, effectively suppressing resonance risks while ensuring compensation effectiveness and maintaining stable system operation. The entire process establishes a complete control chain from characteristic analysis to safety constraints, enabling dynamic gain adjustment to combine fast response and operational reliability.

[0128] Example 12: Based on Example 11, the initial amplitude limiting boundary value mapping module provided in this embodiment of the invention includes:

[0129] The conversion curve generation submodule receives the original fundamental reactive power ratio parameter from the harmonic feature extraction component and inputs it into the boundary reference generator, which has a configurable nonlinear conversion curve set inside.

[0130] The raw boundary value output submodule is used to input the standardized raw fundamental reactive power ratio data into the dynamic slope adjustment unit and synchronously receive the load rate; it dynamically corrects the slope characteristics of the conversion curve according to the load rate: under high load conditions, the slope is appropriately reduced to prevent abrupt boundary value changes; under low load conditions, the slope sensitivity is increased; the corrected conversion curve outputs the uncalibrated raw boundary value.

[0131] The boundary smoothing submodule is used to determine the reasonableness of the original boundary value entering the boundary. It calls the boundary records from the three furthest periods in the historical database to calculate the deviation between the current boundary value and the historical average. When the deviation exceeds the preset threshold, the boundary smoothing program is started, and a weighted average is used to bring the current boundary value closer to the historical median. Finally, the initial amplitude limit boundary value is output.

[0132] The working principle and beneficial effects of the above technical solution are as follows: The conversion curve generation submodule of this embodiment is used to receive the original fundamental reactive power ratio parameter from the harmonic feature extraction component and input it into the boundary reference generator. The generator is internally set with a configurable nonlinear conversion curve. The original boundary value output submodule is used to input the standardized original fundamental reactive power ratio data into the dynamic slope adjustment unit and synchronously receive the load rate. The slope characteristics of the conversion curve are dynamically corrected according to the load rate: the slope is appropriately reduced under high load conditions to prevent abrupt boundary value changes; the slope sensitivity is increased under low load conditions. The corrected conversion curve outputs the uncalibrated original boundary value. The boundary smoothing processing submodule is used for the boundary rationality judgment of the original boundary value, calls the boundary records within the three furthest periods in the historical database, and calculates the deviation between the current boundary value and the historical average. When the deviation exceeds the preset threshold, the boundary smoothing program is started, and the weighted average is used to bring the current boundary value closer to the historical median, and finally outputs the initial limiting boundary value. The initial limiting boundary value mapping module of the above scheme ensures the accuracy and stability of boundary values ​​through multi-level processing: the conversion curve generation submodule establishes the basic mapping relationship, and initially transforms the original fundamental reactive power ratio parameter into a boundary reference value through a nonlinear conversion curve, providing standardized input for subsequent processing. The original boundary value output submodule introduces adaptive adjustment based on operating conditions, dynamically adjusting the slope characteristics of the conversion curve according to the real-time load rate, so that the boundary reference value can adapt to changes in system characteristics under different load conditions, avoiding abrupt boundary value changes or sluggish response caused by a single conversion strategy. The boundary smoothing processing submodule implements stability assurance, and smooths and corrects abnormally abrupt boundary values ​​through historical data comparison and deviation detection, ensuring the continuity and rationality of the output value and preventing boundary value fluctuations caused by instantaneous interference.

[0133] In summary, the initial limiting boundary values ​​generated in this embodiment not only reflect the fundamental reactive power characteristics of the current system but also maintain consistency with historical operating states, providing a reliable benchmark for subsequent dynamic boundary calculations. The entire process achieves accurate conversion from original parameters to stable boundary values, ensuring that the system obtains reasonable initial constraint boundaries under different operating conditions.

[0134] Example 13: Based on Example 12, the boundary smoothing processing submodule provided in this embodiment of the invention includes:

[0135] A historical reference benchmark unit is established to call the boundary records of the most recent three operating cycles in the historical database and extract boundary value samples under the same operating conditions in each cycle; a time decay weighted calculation is performed on the boundary value samples, and the cycle data closer to the current time is assigned a higher weight coefficient to generate a historical reference median with time-sensitive characteristics; the median reflects the boundary benchmark level of the system under normal conditions;

[0136] The deviation measurement evaluation unit is used to compare the difference between the uncalibrated original boundary value and the obtained historical reference median. The difference comparison adopts the relative deviation measurement method to calculate the percentage deviation of the current original boundary value from the historical median. When the percentage exceeds the dynamic threshold range, the smoothing mechanism is triggered. The dynamic threshold range is automatically adjusted according to the system operation stage. The threshold is tightened during steady-state operation and appropriately relaxed during transient process.

[0137] A progressive correction unit is implemented to perform multi-level weighted fusion processing on the original boundary value that triggers smoothing; the algebraic average of the current original boundary value and the historical reference median is taken as the intermediate value; the intermediate value is then weighted again with the historical reference median, and the weight coefficient is dynamically adjusted according to the magnitude of the previous deviation, with the greater the deviation, the higher the weight of the historical data; a boundary change rate limit is applied to the result of the second weighting, so that the corrected original boundary value will not exceed the preset maximum adjustment range per step; after these three layers of processing, the initial amplitude limit boundary value of the final output is obtained.

[0138] The working principle and beneficial effects of the above technical solution are as follows: The historical reference benchmark establishment unit in this embodiment is used to call the boundary records of the three most recent operating cycles in the historical database and extract boundary value samples under the same operating conditions within each cycle; time decay weighted calculation is performed on the boundary value samples, assigning higher weight coefficients to cycle data closer to the current time, generating a historical reference median with time-sensitive characteristics; the median reflects the boundary benchmark level of the system under normal conditions; the deviation measurement and evaluation unit is used to compare the uncalibrated original boundary values ​​with the obtained historical reference median; the difference comparison adopts a relative deviation measurement method to calculate the percentage deviation of the current original boundary value from the historical median; when the percentage deviation is 100%, the deviation is calculated. When the amplitude of the fraction exceeds the dynamic threshold range, a smoothing mechanism is triggered. The dynamic threshold range is automatically adjusted according to the system operation stage, tightening the threshold during steady-state operation and appropriately relaxing it during transient processes. A progressive correction unit is implemented to perform multi-order weighted fusion processing on the original boundary value that triggers smoothing. The algebraic average of the current original boundary value and the historical reference median is taken as the intermediate value. The intermediate value is then weighted again with the historical reference median, and the weight coefficient is dynamically adjusted according to the magnitude of the previous deviation. The greater the deviation, the higher the weight of the historical data. A boundary change rate limit is applied to the result of the second weighting, so that the corrected original boundary value will not exceed the preset maximum adjustment amplitude per step. After these three layers of processing, the initial amplitude limit boundary value of the final output is obtained. The boundary smoothing submodule of the above scheme achieves stable optimization of boundary values ​​through a triple collaborative mechanism: First, a dynamic reference benchmark based on time decay weighting is established to ensure the timely weight distribution of historical data, forming a reference standard reflecting the normal operating characteristics of the system; this benchmark provides an accurate comparison basis for subsequent deviation assessment. Second, a deviation assessment system employing an adaptive threshold strategy is adopted, using dynamically adjusted threshold ranges to differentiate between steady-state and transient operating conditions; this mechanism effectively distinguishes between normal fluctuations and abnormal jumps, avoiding excessive intervention or insufficient response. Third, through the dual constraints of multi-level weighted fusion and rate of change limitation, a smooth transition of boundary values ​​is achieved while preserving the characteristics of the current operating condition. The correction process considers both the stabilizing effect of historical data and prevents numerical mutations through step-by-step processing. The final output initial limit boundary value combines operating condition adaptability and operational stability, providing the system with boundary constraints that reflect both real-time status and long-term operating patterns.

[0139] Example 14: Based on Example 13, the progressive correction unit provided in this embodiment of the invention includes:

[0140] A dynamic weighted benchmark subunit is established to construct a nonlinear weight mapping based on percentage amplitude differences. The previous deviation value is input into the mapping function to generate an initial weight allocation ratio. The generated initial weight allocation ratio is input into the weight compensation regulator and fine-tuned according to the characteristics of the current system operation phase. If the system is in a steady-state operation phase, the weight compensation regulator applies positive compensation to the weights of historical data. If the system is in a transient transition phase, the compensation intensity is maintained or appropriately reduced depending on the transient type. The final weight coefficients after compensation adjustment ensure that the influence of historical reference data matches the dynamic characteristics of the system.

[0141] A bidirectional weighted fusion subunit is executed, using the intermediate value obtained from the first round of processing as the current operating condition characteristic quantity, and the historical reference median value after compensation and adjustment as the stability benchmark quantity; the two are asymmetricly fused according to the final weight coefficient, and the weight ratio of historical data is controlled in the range of 40%-80%; the fusion result forms a secondary weighted value with dynamic balance characteristics;

[0142] An incremental safety constraint subunit is applied to input the secondary weighted value into the boundary change rate limiter and compare it with the final boundary value of the previous cycle. If the boundary change exceeds the maximum adjustment range in a single step, the adjustment amount is scaled up by a preset ratio. The value after constraint processing is output as the final correction result to the boundary value processing pipeline.

[0143] The working principle and beneficial effects of the above technical solution are as follows: In this embodiment, the dynamic weight benchmark subunit is used to construct a nonlinear weight mapping based on the percentage amplitude difference; the previous deviation value is input into the mapping function to generate an initial weight allocation ratio; the generated initial weight allocation ratio is input into the weight compensation regulator and fine-tuned based on the characteristics of the current system operating phase; if in a steady-state operating phase, the weight compensation regulator applies positive compensation to the historical data weights; if in a transient transition phase, the compensation intensity is maintained or appropriately reduced according to the transient type; the final weight coefficient after compensation adjustment ensures that the influence of historical reference data matches the dynamic characteristics of the system; The bidirectional weighted fusion subunit uses the intermediate value obtained from the first round of processing as the current operating condition feature quantity, and the compensated and adjusted historical reference median as the stability benchmark quantity; it performs asymmetric fusion calculation on the two according to the final weight coefficient, with the weight ratio of historical data controlled within the range of 40%-80%; the fusion result forms a secondary weighted value with dynamic equilibrium characteristics; the incremental safety constraint application subunit uses the secondary weighted value to input the boundary change rate limiter and compare it with the final boundary value of the previous cycle; if the boundary change amplitude exceeds the maximum adjustment amplitude in a single step, the adjustment amount is scaled according to a preset ratio; the value after constraint processing is output as the final correction result to the boundary value processing pipeline. The above scheme has a progressive processing chain of dynamic weighting-feature fusion-safety constraint: the compensated weight coefficient output by the dynamic weight benchmark subunit is used as the input parameter for executing the bidirectional weighted fusion subunit; the secondary weighted value generated by fusion enters the incremental safety constraint application subunit for final normalization processing. The final output correction results inherit the essence of historical operating patterns, integrate the key characteristics of the current operating conditions, and ensure output stability through strict safety constraints, providing the system with high-quality boundary values ​​that are both dynamically adaptable and maintain long-term consistency.

[0144] Example 15: As Figure 6 As shown, based on Embodiments 1-14, the power quality comprehensive management device provided in this embodiment of the invention includes: a 10kV busbar 1, a first current transformer 2, a first signal relay 3, a first voltage relay 4, a second signal relay 5, a first disconnecting switch 6, a first contactor 7, a second current transformer 8, a first inductor 9, a second voltage relay 10, a first voltage transformer 11, a first capacitor 12, a second contactor 13, a third current transformer 14, a second inductor 15, a third voltage relay 16, a second voltage transformer 17, a second capacitor 18, a transformer 19, a second disconnecting switch 20, and a static var generator 21;

[0145] Specifically, the 10kV busbar 1 is connected to one end of the first current transformer 2 via a switch. The other end of the first current transformer 2 is connected to one end of the first signal relay 3, one end of the first voltage relay 4, one end of the switch, and one end of the protective grounding terminal. The other ends of the first signal relay 3, the first voltage relay 4, and the switch are grounded. The other end of the protective grounding terminal is connected to one end of the second signal relay 5 and the moving contact of the first disconnecting switch 6. The other end of the second signal relay 5 is grounded. One end of the stationary contact of the first disconnecting switch 6 is grounded, and the other end is connected to one end of the transformer 19 of the first contactor 7 and the second contactor 13. The other end of the first contactor 7 is connected to one end of the second current transformer 8, and the other end of the second current transformer 8 is connected to one end of the first inductor 9. The other end of the first inductor 9 is connected to one end of the second voltage relay 10, the first voltage transformer 11, and the first capacitor 12. The other end of the second voltage relay 10 is grounded. The first voltage transformer 11 and the first capacitor 12 are connected in parallel. The other end of the second contactor 13 is connected to one end of the third current transformer 14. The other end of the third current transformer 14 is connected to one end of the second inductor 15. The other end of the second inductor 15 is connected to one end of the third voltage relay 16, the second voltage transformer 17, and the second capacitor 18. The other end of the third voltage relay 16 is grounded. The second voltage transformer 17 and the second capacitor 18 are connected in parallel. The other end of the transformer 19 is connected to the static contact of the second disconnector 20. The moving contact of the second disconnector 20 is connected to the static var generator 21.

[0146] The working principle and beneficial effects of the above technical solution are as follows: The static compensation section (TBB) of this embodiment consists of a first capacitor 12 and a second capacitor 18 connected in parallel, a first inductor 9 and a second inductor 15, and a first contactor 7 and a second contactor 13 of the contactor switching unit. The 10kV bus 1 monitors the system current through the first current transformer 2. The first signal relay 3 and the second signal relay 5, as well as the first voltage relay 4, the second voltage relay 10, and the third voltage relay 16 detect the voltage / current signals to determine whether compensation is needed. After the first disconnecting switch 6 is closed, the system can select to switch the first contactor 7 or the second contactor 13 to connect the first capacitor 12 or the second capacitor 18 respectively. The first inductor 9 and the second inductor 15 are used to suppress harmonics and prevent capacitor resonance. Static compensation is suitable for stable reactive loads and is mainly used for basic reactive power compensation to improve the power factor.

[0147] The dynamic compensation section of the SVG consists of a static var generator 21, a transformer 19, and a second disconnector 20. The working process is as follows: Transformer 19 is connected to the system and connected to the static var generator 21 via the second disconnector 20. Based on IGBT power electronic devices, it detects the system's reactive power demand in real time and dynamically adjusts the output; it is suitable for rapidly changing reactive loads, such as motor starting, electric arc furnaces, and other impact loads, with a millisecond-level response.

[0148] In this embodiment, static compensation is responsible for basic reactive power compensation and is suitable for stable loads; dynamic compensation is responsible for rapid tracking compensation and is suitable for fluctuating loads; the combination of the two enables fine-tuning compensation, improving compensation accuracy and dynamic response capability. Typical application scenarios include industrial power grids (such as steel plants and rolling mills), new energy power plants (such as photovoltaic and wind power grid connection points), and large data centers (ensuring voltage stability).

[0149] This embodiment of the power quality comprehensive management device adopts a high-voltage reactive power dynamic hybrid compensation scheme, specifically combining the static reactive power compensation of a high-voltage parallel capacitor compensation device with the dynamic reactive power compensation of a static var generator (SVG). The static reactive power compensation consists of 1 to N vacuum contactor-switched reactor capacitor banks. The SVG dynamic reactive power compensation uses fully controlled IGBT devices as the core component, employs a modular design, and provides stepless compensation. Static reactive power compensation is used to compensate for the relatively stable basic load with infrequent reactive power changes in the system. When the load changes rapidly, the SVG dynamic compensation reacts quickly, tracking and compensating according to the load changes. Throughout the load compensation process, static and dynamic compensation work together. Static reactive power compensation provides minimal reactive power, while the SVG dynamic compensation provides a rapid response. The combined switching allows for fine-tuning, ensuring both compensation accuracy and rapid dynamic tracking compensation, achieving excellent compensation results.

[0150] This embodiment employs a high-voltage reactive power dynamic hybrid compensation device combining a complete set of parallel capacitor compensation devices (TBB) and a static var generator (SVG). This avoids the disadvantages of the TBB and adds the advantages of the SVG. The SVG's dynamic compensation has a fast response speed, capable of compensating for loads with rapidly changing reactive power, while the static reactive power compensation of the TBB is used to compensate for loads with relatively stable reactive power and infrequent changes. The combined compensation allows for fine-tuning, high accuracy, and good results. This embodiment achieves fine-tuning, high accuracy, and good results; it improves the power factor through reactive power compensation, reduces the total supply current, enhances power supply safety, and improves power quality; it effectively supports the load-side voltage and strengthens system voltage stability; it saves energy and reduces consumption; after the compensation device is put into operation, it increases the transformer's operating margin, reduces losses, and improves equipment utilization.

[0151] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of equivalents of this invention, this invention is also intended to include these modifications and variations.

Claims

1. A power quality comprehensive treatment device, characterized in that, Include: The dynamic reactive power regulation system is used to generate a compensation current with a phase opposite to the reactive power component of the system by means of a step-type reactive power compensation network based on real-time compensation needs, after voltage matching through a transformer, and then connected to a static var generator through an isolating switch. The system uses the PWM control technology of the IGBT power module to generate a compensation current with a phase opposite to the reactive power component of the system, thus forming a dynamic compensation output with continuous adjustment capability. Dynamic reactive power regulation system, including: The voltage conversion subsystem is used to acquire real-time operating data of the stepped reactive power compensation network. The transformer automatically adjusts the tap position according to the reactive power deficit and converts the voltage to the working voltage level that is compatible with the static var generator. After receiving the access command, the disconnecting switch first performs a pre-closure detection; The current component compensation subsystem is used by the IGBT power module drive system to obtain the voltage and current waveform data of the current cycle, and after processing by the instantaneous reactive power algorithm, decompose the reactive current component that needs to be compensated. Three sets of time-varying modulation signals are generated based on the reactive current component. Each set of signals contains two control layers: fundamental wave compensation command and harmonic suppression command. The switching sequence forming subsystem is used for three sets of time-varying modulation signals to enter the power unit for execution. The three sets of signals are staggered by 30 degrees in the time domain by carrier phase shifting to form an interleaved parallel switching sequence. The turn-on and turn-off time of each IGBT is determined by the intersection of the corresponding modulation signal and carrier. The generated pulse group is amplified by the gate drive to form a power output with phase relationship. The deep automatic adjustment subsystem is used to synthesize the pulse voltages output by each power unit through a filter reactor to form a continuous current waveform that is opposite in phase to the reactive current component. The continuous current waveform is compared with the target compensation amount in real time through closed-loop feedback, and the difference signal between the comparisons is automatically adjusted to adjust the PWM modulation depth.

2. The power quality comprehensive management device of claim 1, wherein, The current component compensation subsystem includes: The vector decomposition component is used to input synchronized voltage and current waveform data into instantaneous power calculation and establish a dynamic reference coordinate system using the orthogonal decomposition principle. The rotation angle of the reference coordinate system is determined in real time by the positive sequence component of the voltage fundamental wave, forming an adaptive rotation calculation reference. The current vector is decomposed into two orthogonal projection components in the reference coordinate system; The harmonic feature extraction component is used to extract the reactive current component, which then performs a spectral scan on the reactive current component. The scanning results generate a harmonic energy distribution spectrum; The dynamic gain adjustment component is used to control the command generator to receive the decomposed reactive current components and harmonic distortion distribution information. The fundamental reactive component is dynamically gain-adjusted. The gain coefficient is determined by the ratio of the system reactive power deficit to a preset threshold, generating a basic compensation amplitude reference. At the same time, the harmonic components are subjected to frequency domain weighting to suppress the preset characteristic subharmonics. The harmonic suppression component uses a layered coding structure for the three sets of time-varying modulation signals for output. The bottom layer of each set of time-varying modulation signals is the fundamental compensation waveform, whose amplitude changes with the reference value after gain adjustment. The upper layer is superimposed with a phase-modulated harmonic suppression waveform, and the envelope of the harmonic suppression waveform matches the energy distribution of the characteristic subharmonic. The three sets of time-varying modulation signals maintain a phase interlock relationship.

3. The power quality comprehensive management device of claim 2, wherein, The dynamic gain adjustment component includes: The mode switching sub-component is used to obtain the separated quadrature reactive current components, compare the current quadrature reactive current component amplitude with a preset threshold, and establish a dynamic adjustment coefficient calculation program. The coefficient generation sub-component is used by the main channel to generate a coarse compensation coefficient based on the comparison results. The coarse compensation coefficient has a piecewise linear relationship with the deviation. The auxiliary channel introduces an inertia factor formed by historical compensation data to smooth the output of the main channel; the outputs of the two channels are weighted and superimposed at the fusion node, and the weight value depends on the dynamic response requirements of the current system. The dynamic correction sub-component is used to apply the coarse gain coefficient after channel fusion to anti-saturation limiting, and has an adaptive limiting boundary. The boundary value of the adaptive amplitude limiting boundary is dynamically corrected by the spectral purity index provided by the harmonic feature extraction component to obtain the dynamic gain coefficient; The coupling operation sub-component generates dynamic gain coefficients which are then fed into the reference generation unit. The dynamic gain coefficients are coupled with the system's nominal compensation capacity. The output base compensation amplitude reference value has time-varying characteristics, and its rate of change is constrained by the system's transient response characteristics.

4. The power quality comprehensive management device of claim 3, wherein, The dynamic correction subcomponent includes: The initial amplitude limiting boundary value mapping module is used to acquire spectral purity index data from the harmonic energy distribution map, characterizing the current system's harmonic pollution level, including two key parameters: fundamental reactive power ratio and harmonic distortion rate. Boundary calculations are performed on these two key parameters, establishing a boundary amplitude benchmark value in the fundamental reactive power ratio dimension and a boundary contraction coefficient in the harmonic distortion rate dimension. The fundamental reactive power ratio parameter in the spectral purity index is normalized and mapped to the initial amplitude limiting boundary value; simultaneously, the harmonic distortion rate parameter is logarithmically transformed to generate a boundary contraction factor. The product operation module is used to multiply the initial amplitude limit boundary value with the boundary contraction factor to obtain the preliminary dynamic boundary; the boundary smoothing module combines historical boundary data to form an inertial adjustment effect; the adjustment trend correction of the preliminary dynamic boundary acceptance coefficient generation sub-component; when the compensation coefficient after the fusion of the main and auxiliary channels shows a continuous growth trend, the boundary value is preventively contracted. The gain correction module generates dynamic limiting boundary values ​​which are then fed into the gain correction module. Coarse gain coefficients exceeding the boundary range are constrained to the boundary values, and a boundary over-limit flag is triggered. Information from the boundary over-limit flag is fed back to the mode switching subcomponent.

5. The power quality comprehensive management device of claim 4, wherein, The initial amplitude limiting boundary value mapping module includes: The conversion curve generation submodule is used to receive the original fundamental reactive power ratio parameter from the harmonic feature extraction component and input it into the boundary reference generator. The boundary reference generator is internally set with a configurable nonlinear conversion curve. The original boundary value output submodule is used to input the standardized original fundamental reactive power ratio parameter into the dynamic slope adjustment unit and synchronously receive the load rate; dynamically correct the slope characteristics of the nonlinear conversion curve according to the load rate; and output the uncalibrated original boundary value of the corrected nonlinear conversion curve. The boundary smoothing submodule is used to determine the reasonableness of the boundary when the uncalibrated original boundary value enters the boundary. It calls the boundary records from the three furthest periods in the historical database and calculates the deviation between the uncalibrated original boundary value and the historical average. When the deviation exceeds the preset threshold, the boundary smoothing program is started, and a weighted average is used to bring the uncalibrated original boundary value closer to the historical average, and finally outputs the initial amplitude limit boundary value.

6. The power quality comprehensive management device of claim 5, wherein, The boundary smoothing submodule includes: A historical reference benchmark unit is established to call the boundary records of the three furthest operating cycles in the historical database and extract boundary value samples under the same operating conditions in each cycle; time decay weighted calculation is performed on the boundary value samples to generate historical averages with time-sensitive characteristics; The deviation measurement evaluation unit is used to compare the uncalibrated original boundary value with the obtained historical mean. The difference comparison adopts the relative deviation measurement method to calculate the percentage deviation of the uncalibrated original boundary value from the historical mean. When the percentage deviation exceeds the dynamic threshold range, the smoothing mechanism is triggered. A progressive correction unit is implemented to perform multi-level weighted fusion processing on the uncalibrated original boundary values ​​that trigger smoothing; the algebraic average of the uncalibrated original boundary values ​​and the historical mean is taken as the intermediate value; the intermediate value is then weighted again with the historical mean, and the weight coefficients are dynamically adjusted according to the magnitude of the previous deviation; a boundary change rate limit is applied to the result of the second weighting, so that the corrected original boundary value will not exceed the preset maximum adjustment range per step; after processing, the initial amplitude limit boundary value of the final output is obtained.

7. The power quality comprehensive management device of claim 6, wherein, Implementing a progressive correction unit includes: A dynamic weighted benchmark sub-unit is established to construct a nonlinear weight mapping based on the percentage magnitude difference; the previous deviation value is input into the mapping function to generate the initial weight allocation ratio; The generated initial weight allocation ratio is input into the weight compensation adjuster and fine-tuned based on the characteristics of the current system operation phase; The bidirectional weighted fusion subunit is executed, and the intermediate value obtained from the first round of processing is used as the current operating condition characteristic quantity, while the historical average value after compensation and adjustment is used as the stability benchmark quantity. The two are asymmetrically fused according to the final weighting coefficients; The fusion result forms a quadratic weighted value with dynamic equilibrium characteristics; An incremental safety constraint sub-unit is applied to input the quadratic weighted value into the boundary rate of change limiter and compare it with the final boundary value of the previous period.

8. The power quality comprehensive management device of claim 1, wherein, It also includes: an energy monitoring system, which uses current transformers to sample the current of the 10kV bus, voltage relays to collect voltage signals, and signal relays to monitor protection signals, forming comprehensive operating parameters that include system voltage, current and harmonic characteristics.

9. The power quality comprehensive management device of claim 1, wherein, It also includes: a controllable static compensation system, which, based on comprehensive operating parameters, uses intelligent switching control of contactors to drive capacitors to form capacitive compensation branches, and combines them with series resonant suppression circuits composed of inductors to establish a stepped reactive power compensation network with harmonic suppression function.

Images