A power grid harmonic detection and collaborative management method and system based on harmonic feature decomposition

Abstract

The application discloses a power grid harmonic detection and collaborative management method and system based on harmonic feature decomposition, relates to the technical field of harmonic management, and comprises the following steps: collecting electrical quantities and device states and packaging the same into panoramic data frames with time and space labels; a dynamic disturbance causal network graph is generated by using a causal inference algorithm to identify disturbance sources and coupling relationships; based on the disturbance causal network graph, a minimum collaborative management total cost is calculated by using a multi-target collaborative decision model, optimal collaborative management instructions are obtained, and the optimal collaborative management instructions are sent to a static var generator controller for execution. The application also discloses a corresponding system, which comprises a data acquisition module, a data processing module, a disturbance causal network graph construction module and a collaborative management decision module. The application realizes fast sensing and accurate tracing of disturbances, and realizes multi-target dynamic collaborative optimization of reactive power compensation, harmonic management and voltage support within a limited capacity.

Description

Technical Field

[0001] This invention relates to the field of harmonic mitigation technology, and in particular to a method and system for power grid harmonic detection and collaborative mitigation based on harmonic characteristic decomposition. Background Technology

[0002] In power systems, harmonic pollution and reactive power fluctuations are the main factors affecting power quality. Currently, harmonic mitigation widely employs hybrid compensation devices consisting of a high-voltage passive filter circuit (HFC) and a static var generator (SVG). The HFC is responsible for filtering out characteristic harmonics and providing basic reactive power compensation, while the SVG dynamically and finely adjusts residual harmonics and reactive power. However, existing BW-TSVG hybrid devices still have the following technical shortcomings in actual operation.

[0003] First, the data acquisition dimensions are isolated. Existing solutions collect electrical quantities, device status, and environmental information independently, without establishing spatiotemporal correlations, making it difficult to form a complete description of disturbance events. When harmonic anomalies occur in the power grid, maintenance personnel cannot quickly trace the source, leading to difficulties in fault location.

[0004] Secondly, the analysis has inherent latency. Traditional methods rely on Fast Fourier Transform (FFT) for frequency domain analysis, but FFT requires at least one full power frequency cycle (20ms) to complete the calculation. For millisecond-level transient disturbances such as sudden changes in photovoltaic output and load switching, this method has a response lag and cannot achieve trend prediction and early warning.

[0005] Furthermore, the decision-making strategies are rigid. Existing control strategies are mostly static, predefined, fixed-ratio allocations (such as allocating SVG capacity to reactive power compensation and harmonic mitigation at a fixed percentage). When harmonic pollution is severe but the reactive power deficit is small, fixed allocations will lead to insufficient harmonic mitigation capacity; conversely, it will result in idle capacity. This allocation method, lacking dynamic game-theoretic optimization, cannot achieve optimal coordination of multiple objectives such as reactive power compensation, harmonic mitigation, and voltage support within the limited capacity of the SVG.

[0006] Furthermore, the existing solution does not consider the impact of the actual operating temperature of the device on the SVG capacity, nor does it include ambient temperature and power module temperature (such as insulated gate bipolar transistors, IGBTs) in the dynamic constraints, resulting in overload risks under high-temperature conditions.

[0007] Therefore, how to achieve rapid detection, accurate source tracing, and dynamic collaborative governance of power grid disturbances has become a pressing technical problem that needs to be solved in this field. Summary of the Invention

[0008] To address these issues, this invention provides a method and system for power grid harmonic detection and collaborative governance based on harmonic feature decomposition. This system solves the problems in existing technologies, such as isolated data acquisition leading to difficulties in tracing the source of disturbances, inherent delays in frequency domain analysis causing response lags, and the inability of static decision-making strategies to achieve multi-objective dynamic collaborative optimization within limited capacity.

[0009] To address the aforementioned technical problems, embodiments of the present invention provide a method for power grid harmonic detection and collaborative mitigation based on harmonic feature decomposition, the method comprising:

[0010] Step S1: Collect electrical quantities and device status; wherein, the electrical quantities include instantaneous values ​​of three-phase voltage and instantaneous values ​​of three-phase current, and the device status includes device spatial location information, switching status and current of each branch of the high-voltage passive filter branch, power module temperature and ambient temperature;

[0011] Step S2: Encapsulate the data collected in step S1 into panoramic data frames with spatiotemporal tags. Each panoramic data frame contains a timestamp, spatial location identifier, electrical quantity vector, device state vector, and environmental parameter vector.

[0012] Step S3: Process the panoramic data frames obtained in Step S2 using a preset causal inference algorithm to establish the correlation between disturbance events and generate a dynamic disturbance causal network graph; wherein, the causal inference algorithm determines the causal relationship based on the temporal relationship and statistical probability of the events, specifically: when the probability of the second disturbance event occurring within a preset time window after the first disturbance event occurs is higher than the baseline probability, it is determined that there is a causal relationship between the first disturbance event and the second disturbance event; in the disturbance causal network graph, disturbance events are nodes, and the causal relationship between disturbance events is a directed edge;

[0013] Step S4: Based on the disturbance causal network diagram generated in Step S3, identify the sources of disturbance and the coupling relationship between disturbance events. Based on this, dynamically adjust the cost matrix parameters in the multi-objective collaborative decision-making model. Then, calculate the total cost of collaborative governance using the multi-objective collaborative decision-making model to obtain the minimum total cost of collaborative governance. Next, obtain the optimal collaborative governance instruction based on the minimum total cost of collaborative governance. Then, generate the corresponding compensation instruction based on the optimal collaborative governance instruction and send the compensation instruction to the static var generator controller for execution.

[0014] Preferably, in step S1, electrical quantity acquisition is completed by the following operations: installing voltage transformers and current transformers at the point of common coupling, acquiring instantaneous values ​​of three-phase voltage and three-phase current at a sampling rate higher than the power frequency cycle, and setting a preset number of sampling points in each power frequency cycle to meet the requirements of preset harmonic analysis.

[0015] Preferably, in step S4, the total cost of collaborative governance is calculated using the following formula:

[0016] ;

[0017] In the formula, For collaborative governance decision-making variable vectors; This is the cost matrix; It is a linear cost vector; This is the motion smoothing coefficient; This represents the change in the collaborative governance instruction vector. This is the matrix transpose symbol;

[0018] in, , Allocate output capacity for reactive power compensation to the static var generator. Allocate power to the static var generator for compensation of the first The effective value of the output current of the subharmonic. For harmonic order and , For natural numbers, Allocate output capacity for voltage support to the static var generator; , This is the instruction vector to be used during collaborative governance at the current moment. This is the instruction vector used during the collaborative governance in the previous time step.

[0019] Preferably, in step S4, when calculating the total cost of collaborative governance and obtaining the minimum total cost of collaborative governance based on the disturbance causal network diagram, the calculation of the total cost of collaborative governance is performed under preset constraints; wherein, the constraints are defined as follows:

[0020] Dynamic capacity constraints: ;

[0021] High-voltage passive filter branch cooperative constraints: ;

[0022] Harmonic capacity constraints: ;

[0023] Nonnegativity constraint: ;

[0024] In the formula, Rated voltage, This refers to the steady-state reactive power that the high-voltage passive filter branch can provide. For the first The maximum allowable compensation current for subharmonics. This refers to the dynamic effective rated capacity of the static var generator; among which, , The rated apparent power capacity is shown on the nameplate of the static var generator. This refers to the temperature derating factor of the PN junction in an insulated-gate bipolar transistor (IGBT) chip. This is the derating factor for ambient temperature.

[0025] Preferably, in step S3, the preset time window is less than one power frequency cycle.

[0026] This invention also provides a power grid harmonic detection and collaborative management system based on harmonic feature decomposition. This system is used to implement the aforementioned power grid harmonic detection and collaborative management method based on harmonic feature decomposition, specifically including:

[0027] The data acquisition module is used to collect electrical quantities and device status in the power grid; wherein, the electrical quantities include instantaneous values ​​of three-phase voltage and instantaneous values ​​of three-phase current, and the device status includes device spatial location information, switching status and current of each branch of the high-voltage passive filter branch, power module temperature and ambient temperature;

[0028] The data processing module is used to encapsulate the data acquired by the data acquisition module into panoramic data frames with spatiotemporal tags. Each panoramic data frame contains a timestamp, spatial location identifier, electrical quantity vector, device state vector, and environmental parameter vector.

[0029] The perturbation causal network graph construction module is used to process the panoramic data frames obtained from the data processing module using a preset causal inference algorithm, establish the correlation between perturbation events, and generate a dynamic perturbation causal network graph. Specifically, the causal inference algorithm determines causal relationships based on the temporal sequence of events and statistical probabilities. Specifically, when the probability of a second perturbation event occurring within a preset time window after the occurrence of a first perturbation event is higher than a baseline probability, a causal relationship is determined between the first and second perturbation events. In the perturbation causal network graph, perturbation events are nodes, and the causal relationships between perturbation events are directed edges.

[0030] The collaborative governance decision-making module is used to identify the coupling relationship between disturbance sources and disturbance events based on the disturbance causal network diagram generated by the disturbance causal network diagram construction module. Based on this, it dynamically adjusts the cost matrix parameters in the multi-objective collaborative decision-making model, calculates the total collaborative governance cost using the multi-objective collaborative decision-making model, obtains the minimum total collaborative governance cost, derives the optimal collaborative governance instruction based on the minimum total collaborative governance cost, generates the corresponding compensation instruction based on the optimal collaborative governance instruction, and then sends the compensation instruction to the static var generator controller for execution.

[0031] Preferably, the collaborative governance decision-making module has a built-in constraint module, which is preset with constraint conditions for constraining the multi-objective collaborative decision-making model to obtain the minimum total collaborative governance cost; the constraint conditions include dynamic capacity constraints, high-voltage passive filter branch collaborative constraints, harmonic capacity constraints, and non-negative constraints.

[0032] Preferably, the dynamic capacity constraint is: ,That The cooperative constraint of the high-voltage passive filter branch is: The harmonic capacity constraint is as follows: The nonnegativity constraint is: In the formula, Allocate output capacity for reactive power compensation to the static var generator. Allocate power to the static var generator for compensation of the first The effective value of the output current of the subharmonic. Allocate output capacity for voltage support to the static var generator. Rated voltage, The steady-state reactive power that the high-voltage passive filter branch can provide. For the first The maximum allowable compensation current for subharmonics. This refers to the dynamic effective rated capacity of the static var generator. The rated apparent power capacity is shown on the nameplate of the static var generator. This refers to the temperature derating factor of the PN junction in an insulated-gate bipolar transistor (IGBT) chip. This is the derating factor for ambient temperature.

[0033] This invention also provides an electronic device, which includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the above-described method for power grid harmonic detection and collaborative management based on harmonic feature decomposition.

[0034] This invention also provides a computer storage medium storing a computer software product, the computer software product including several instructions to cause a computer device to execute the above-described method for power grid harmonic detection and collaborative management based on harmonic feature decomposition.

[0035] As can be seen from the above technical solutions, this invention application has the following beneficial effects:

[0036] First, it enables panoramic perception and precise source tracing of disturbances: by collecting electrical quantities, device status and environmental information, it encapsulates them into panoramic data frames with spatiotemporal tags, and uses causal inference algorithms to construct dynamic disturbance causal network diagrams, which can intuitively display the source, propagation path and coupling relationship of disturbance events, solving the problems of data isolation and difficulty in tracing the source of disturbances in existing technologies.

[0037] Second, it achieves millisecond-level rapid response and early warning: It uses a sampling rate higher than the power frequency (such as 12.8kHz) to capture transient disturbance signals, and the causal determination time window based on time-series probability is less than one power frequency cycle (such as less than 5ms). Compared with the delay of at least 20ms in traditional FFT analysis, it significantly improves the response speed of disturbance detection and early warning.

[0038] Third, achieve dynamic collaborative optimization of multiple objectives under limited capacity: the objectives of reactive power compensation, harmonic control, voltage support, etc. of SVG are modeled as multi-objective collaborative decision functions. Combined with dynamic capacity constraints, HFC collaborative constraints, etc., the optimal resource allocation instructions are solved in real time, which overcomes the rigidity of static fixed ratio allocation strategy and realizes adaptive optimal collaboration of multiple governance objectives within the limited capacity of SVG. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Referring to the drawings will make the features and advantages of the present invention clearer. The drawings are illustrative and should not be construed as limiting the present invention in any way. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0040] Figure 1 This is a flowchart of a power grid harmonic detection and collaborative management method based on harmonic characteristic decomposition provided by the present invention;

[0041] Figure 2 This is a block diagram of a power grid harmonic detection and collaborative management system based on harmonic feature decomposition provided by the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0043] Example 1:

[0044] To address the challenges in existing technologies, such as isolated data acquisition leading to difficulties in disturbance tracing, inherent delays in frequency domain analysis causing response lag, and the inability of static decision-making strategies to achieve multi-objective dynamic collaborative optimization within limited capacity, this embodiment proposes a power grid harmonic detection and collaborative management method based on harmonic characteristic decomposition. Figure 1 As shown, the method includes:

[0045] Step S1: Collect electrical quantities and device status; wherein, the electrical quantities include instantaneous values ​​of three-phase voltage and instantaneous values ​​of three-phase current, and the device status includes device spatial location information, switching status and current of each branch of the high-voltage passive filter branch, power module temperature and ambient temperature;

[0046] Step S2: Encapsulate the data collected in step S1 into panoramic data frames with spatiotemporal tags. Each panoramic data frame contains a timestamp, spatial location identifier, electrical quantity vector, device state vector, and environmental parameter vector.

[0047] Step S3: Process the panoramic data frames obtained in Step S2 using a preset causal inference algorithm to establish the correlation between disturbance events and generate a dynamic disturbance causal network graph; wherein, the causal inference algorithm determines the causal relationship based on the temporal relationship and statistical probability of the events, specifically: when the probability of the second disturbance event occurring within a preset time window after the first disturbance event occurs is higher than the baseline probability, it is determined that there is a causal relationship between the first disturbance event and the second disturbance event; in the disturbance causal network graph, disturbance events are nodes, and the causal relationship between disturbance events is a directed edge;

[0048] Step S4: Based on the disturbance causal network diagram generated in Step S3, identify the sources of disturbance and the coupling relationship between disturbance events. Based on this, dynamically adjust the cost matrix parameters in the multi-objective collaborative decision-making model. Then, calculate the total cost of collaborative governance using the multi-objective collaborative decision-making model to obtain the minimum total cost of collaborative governance. Next, obtain the optimal collaborative governance instruction based on the minimum total cost of collaborative governance. Then, generate the corresponding compensation instruction based on the optimal collaborative governance instruction and send the compensation instruction to the static var generator controller for execution.

[0049] As can be seen from the above technical solution, this invention proposes a method for power grid harmonic detection and collaborative management based on harmonic characteristic decomposition.

[0050] Furthermore, in step S1, this step aims to collect data using sensors installed at various data acquisition points, in order to provide a comprehensive, high-quality data foundation for collaborative power grid harmonic governance decisions. Specifically, the following data collection is performed:

[0051] (1) Electrical quantity acquisition

[0052] Voltage and current transformers are installed at the 10kV point of common coupling (PCC) to collect the instantaneous values ​​of the three-phase voltage at a sampling rate of 12.8kHz. , , and instantaneous values ​​of three-phase current , , Furthermore, 256 sampling points are set within each power frequency cycle to meet the requirements of 25th harmonic analysis. The high sampling rate (12.8kHz) enables millisecond-level capture of transient disturbance signals, effectively compensating for the inherent delay defect of traditional FFT analysis, which requires at least one power frequency cycle (20ms).

[0053] The collected electrical quantity data are used for the following purposes:

[0054] Based on the instantaneous values ​​of the three-phase currents, the amplitudes of each harmonic current are obtained through transient feature extraction and used as subsequent decision variables. The basis for target compensation.

[0055] Based on the instantaneous values ​​of the three-phase voltage, the voltage deviation is calculated in real time and used to dynamically correct the target value and weight of the voltage support decision variable u.

[0056] Based on the instantaneous values ​​of three-phase voltage and current, the reactive power deficit of the system is calculated in real time and used as the basis for compensation of the decision variable q.

[0057] At major photovoltaic power plants and nonlinear load grid connection points, it is also necessary to collect the source output current to construct a disturbance source feature library, which will support the disturbance source tracing and propagation path analysis of subsequent dynamic causal networks.

[0058] (2) Device status acquisition

[0059] The following device status data are collected in real time:

[0060] The switching status and current of each branch of the high-voltage passive filter (HFC) are used to determine the HFC cooperative constraint q. HFC The current value (i.e., the steady-state reactive power that HFC can currently provide).

[0061] Power module temperature and ambient temperature Used to dynamically correct the dynamic effective rated capacity of SVG .

[0062] In addition, it is necessary to collect the spatial location information of the device in real time, so as to add spatial location markers to subsequent panoramic data frames and realize the spatiotemporal positioning of disturbance events.

[0063] Furthermore, in step S2, all the data information collected in step S1 is encapsulated into panoramic data frames with spatiotemporal labels, which serve as the unified input format for subsequent steps. Each panoramic data frame contains a timestamp, spatial location identifier, electrical quantity vector, device state vector, and environmental parameter vector, ensuring that the data timing and spatial location are traceable, supporting disturbance tracing and game-theoretic optimization calculations.

[0064] Furthermore, in step S3, a preset causal inference algorithm is used to analyze and process the panoramic data frames obtained in step S2, automatically establishing the correlation between disturbance events and generating a dynamic "disturbance causal network diagram" to achieve disturbance source tracing and solve the problems of lagging analysis and inability to provide early warnings in traditional solutions. Specifically, this includes the following sub-steps:

[0065] S31. Disturbance event extraction and feature recognition:

[0066] Real-time monitoring of panoramic data frames identifies discrete disturbance events such as "sudden change in harmonic amplitude", "sudden increase in photovoltaic output", and "load switching". Multi-domain feature extraction is performed on the PCC point current to identify the current dominant disturbance mode.

[0067] S32. Construct a dynamic perturbation causal network diagram:

[0068] Using the perturbation events identified in step S31 as "nodes" and the causal / strong correlation relationships between events as "directed edges," a dynamic "perturbation causal network graph" is constructed in real time. This graph visually illustrates the process.

[0069] 1. Source of disturbance: Which nodes are "root nodes" that only have outgoing edges and no incoming edges;

[0070] 2. Propagation path: How does the disturbance energy propagate through other nodes in the network and converge at the PCC point?

[0071] 3. Coupling relationship: Whether there is a mutual excitation coupling relationship between harmonics of different orders, and between harmonics and reactive power fluctuations.

[0072] The criteria for determining causality are as follows: If event A occurs after... The probability of event B occurring within the time interval is significantly higher than the baseline probability, and Then a directed edge A→B is established. It is much shorter than the power frequency cycle (20ms), enabling millisecond-level rapid cause-effect determination and disturbance early warning.

[0073] The causal inference algorithm determines the causal relationship based on the temporal relationship and statistical probability of the events. Specifically, when the probability of the second disturbance event occurring is higher than the baseline probability within a preset time window (less than one power frequency cycle, preferably less than 5ms) after the first disturbance event occurs, it is determined that there is a causal relationship between the first disturbance event and the second disturbance event.

[0074] Furthermore, in step S4, given that traditional fixed strategies cannot cope with dynamically changing disturbance scenarios, the limited capacity of the SVG needs to be dynamically allocated among multiple objectives such as reactive power compensation, harmonic mitigation, and voltage support. This step, based on dynamic game theory and convex optimization methods, constructs a multi-objective collaborative decision-making model to solve for the optimal collaborative governance command within milliseconds. Specifically, it includes the following sub-steps:

[0075] S41. Construction of a multi-objective collaborative decision-making model:

[0076] In traditional solutions, the capacity of the SVG is allocated in a fixed ratio (e.g., 60% reactive power + 40% harmonics), which cannot adapt to real-time changes in disturbances. When harmonic pollution is severe and the reactive power deficit is small, the fixed ratio leads to insufficient harmonic mitigation capacity; conversely, it leads to idle capacity. This invention models the finite compensation capacity of the SVG as a "resource" and multiple mitigation objectives (reactive power compensation, harmonic mitigation, and voltage support) as "participants" in a game. By minimizing the total cost function, the optimal resource allocation for each mitigation objective is solved, and the globally optimal coordinated control command is output.

[0077] S41-1, Define the decision variable vector:

[0078] Transforming the multi-objective resource allocation problem of SVG into a vector-based mathematical expression makes subsequent optimization possible. The decision variable vector fully describes the resource allocation scheme of SVG at any given time, and is defined as follows:

[0079] ;

[0080] In the formula, T is the matrix transpose symbol; The SVG resource allocation decision vector includes the allocation of three types of governance objectives: reactive power compensation capacity, harmonic compensation current, and voltage support capacity. Allocate output capacity for reactive power compensation to the SVG, in Mvar; Assign an effective value of the output current, in amperes (A), to the SVG to compensate for the h-th harmonic; where Represents the harmonic order; the harmonic order selected in this scheme is [number missing]. =5, 7, 11, 13, 17, 19, 23, 25. This sequence originates from the harmonic generation theory of three-phase bridge rectifier circuits. For a three-phase bridge rectifier, the characteristic harmonic orders it generates satisfy: , It is a natural number; Allocate the output capacity for voltage support to the SVG, in Mvar.

[0081] S41-2. Calculate the total cost of collaborative governance:

[0082] The complex multi-objective resource allocation problem is transformed into a standard quadratic programming problem using the total cost function of collaborative governance. The total cost function of collaborative governance is defined as follows:

[0083] ;

[0084] In the formula: The smaller the value, the better the allocation scheme;

[0085] For collaborative governance decision-making variable vectors;

[0086] For the cost matrix, Its diagonal elements For the first The unit uncompensated cost of each governance objective, with off-diagonal elements representing the coupling cost between different governance objectives, is used to quantify the coupling relationship between different governance objectives.

[0087] This is a linear cost vector, specifically the fixed cost coefficient for each governance objective. ,in To utilize Fixed cost coefficient for harmonic coordinated mitigation;

[0088] The motion smoothing coefficient is a constant greater than 0, used to adjust the intensity of instruction smoothing penalty. Its typical value range is... It can be dynamically adjusted according to the system response speed and smoothing requirements;

[0089] This refers to the change in the command vector for collaborative governance, used to characterize the fluctuation range of the commands. ,in This is the instruction vector to be used during collaborative governance at the current moment. This is the instruction vector used during the previous collaborative governance process;

[0090] This is the matrix transpose symbol.

[0091] The meanings of each term in the above cost function are as follows:

[0092] The term is quadratic, representing the coupling cost and increasing marginal cost effect between various governance objectives;

[0093] This is a linear cost term, representing the fixed cost of each governance objective, used to balance the overall optimization cost;

[0094] A penalty term for smooth operation is included to suppress drastic changes in commands, ensuring smooth output, reducing switching losses, and improving system stability. This penalty term avoids additional switching losses and system oscillations caused by high-frequency operation of power devices, achieving a smooth and stable transition of output current, effectively protecting the IGBT power module, reducing equipment noise, and significantly improving the device's operational stability, robustness, and lifespan.

[0095] Step S42, Constraints and Feasibility Domain Construction:

[0096] Since unconstrained optimization may lead to solutions exceeding the actual physical capabilities of the Static Var Generator (SVG) (e.g., total output exceeding rated capacity) or violating the collaborative logic between the High Voltage Passive Filter Branch (HFC) and the SVG, this embodiment introduces constraints when calculating the total collaborative governance cost to ensure that the solution for the total collaborative governance cost is always within the actual capabilities of the SVG and HFC, and conforms to the engineering principle of "HFC as the main variable, SVG for fine-tuning." The values ​​of are constrained.

[0097] The constraints are defined as follows:

[0098] Dynamic capacity constraints:

[0099] ;

[0100] High-voltage passive filter branch cooperative constraints:

[0101] ;

[0102] Harmonic capacity constraints:

[0103] ;

[0104] Nonnegativity constraint:

[0105] ;

[0106] In the formula:

[0107] Rated voltage (e.g., 10kV);

[0108] The steady-state reactive power that the high-voltage passive filter branch can currently provide is determined by the capacity of the filter branch currently switched by the HFC, and the unit is Mvar;

[0109] For the first The maximum allowable compensation current for subharmonics is limited by the SVG switching frequency and filter design, and is expressed in amperes (A).

[0110] This is the dynamic effective rated capacity of the SVG, measured in MVA, and is obtained from the fixed rated capacity of the equipment after temperature correction.

[0111] in, In the formula The rated apparent power capacity is specified on the SVG device nameplate. The temperature derating factor of the PN junction in an Insulated Gate Bipolar Transistor (IGBT) chip. This is the derating factor for ambient temperature.

[0112] The function and logical explanation of each constraint are as follows:

[0113] Dynamic capacity constraint: Ensure that the total apparent power output of the SVG (including the power converted from reactive power compensation, harmonic compensation current, and voltage support capacity) does not exceed the temperature-corrected dynamic effective rated capacity, prevent overload damage to the equipment, and improve the operational safety under high-temperature conditions.

[0114] High-voltage passive filter branch cooperative constraint ( This constraint reflects the optimization principle of "HFC handling the main load and SVG handling fine-tuning". As the basic compensation device, HFC undertakes most of the steady-state reactive power compensation and characteristic subharmonic filtering tasks, providing steady-state reactive power... This constitutes the main body of reactive power compensation. As a dynamic adjustment device, the SVG only provides fine-tuning when the HFC capacity is insufficient to meet the system's reactive power demand, or absorbs excess reactive power when the HFC is overcompensated. Therefore, the output capacity q allocated by the SVG for reactive power compensation should not exceed the steady-state reactive power that the HFC can currently provide. This ensures that the SVG only plays an auxiliary regulation role, avoiding its long-term operation in a high-power output state, effectively reducing switching losses and extending equipment life.

[0115] Harmonic capacity constraint: Limits the compensation current of a single harmonic to prevent local overload of the SVG due to excessive compensation current of a certain harmonic, thus protecting the power device.

[0116] Non-negativity constraint: Ensures that all allocated capacities are non-negative, which conforms to physical reality.

[0117] S43. Result Output and Execution:

[0118] After solving the optimal collaborative governance command in the multi-objective collaborative decision-making model based on the dynamic game optimization model, the generated compensation command is efficiently output to the SVG controller, and the visualization of the collaborative governance effect, data traceability and system self-iteration are realized simultaneously, forming a closed-loop collaborative governance process from perception, decision-making to execution and evaluation.

[0119] Specifically: the optimal decision vector obtained by solving... The current and reactive power commands are converted into executable SVG commands and sent to the SVG controller via a high-speed communication interface. At the same time, based on the current switching status of HFC and the characteristics of harmonic distribution, adaptive switching commands for the LC filter branch are generated to achieve millisecond-level coordinated response between SVG and HFC, ensuring the integrated execution of multiple objectives such as reactive power compensation, harmonic control, and voltage support.

[0120] In addition, as an optional optimization step, before performing dimensionality reduction and smoothing on the data collected in step S1, the data collected in step S1 can be decoupled based on the fundamental frequency and harmonic frequency domain, and then the decoupled data can be filtered to improve the data quality.

[0121] This concludes the implementation of the power grid harmonic detection and collaborative management method based on harmonic feature decomposition described in this embodiment.

[0122] Example 2:

[0123] This embodiment provides a power grid harmonic detection and collaborative management system based on harmonic feature decomposition. This system is used to implement the power grid harmonic detection and collaborative management method based on harmonic feature decomposition described in Embodiment 1 above, such as... Figure 2 As shown, it specifically includes:

[0124] The data acquisition module is used to collect electrical quantities and device status in the power grid. The electrical quantities include instantaneous values ​​of three-phase voltage and three-phase current. The device status includes device spatial location information, the switching status and current of each branch of the high-voltage passive filter, power module temperature, and ambient temperature. The data acquisition module installs voltage transformers and current transformers at the 10kV point of common coupling and collects the instantaneous values ​​of three-phase voltage at a sampling rate of 12.8kHz. , , and instantaneous values ​​of three-phase current , , Furthermore, 256 sampling points are set within each power frequency cycle.

[0125] The data processing module is used to encapsulate the data collected by the data acquisition module into panoramic data frames with spatiotemporal labels. Each panoramic data frame contains a timestamp, spatial location identifier, electrical quantity vector, device status vector, and environmental parameter vector.

[0126] The perturbation causal network graph construction module is used to process the panoramic data frames obtained from the data processing module using a preset causal inference algorithm, establish the correlation between perturbation events, and generate a dynamic perturbation causal network graph. Specifically, the causal inference algorithm determines causal relationships based on the temporal sequence of events and statistical probabilities. Specifically, when the probability of a second perturbation event occurring within a preset time window (less than one power frequency cycle, preferably less than 5ms) after the occurrence of the first perturbation event is higher than a baseline probability, a causal relationship is determined between the first and second perturbation events. In the perturbation causal network graph, perturbation events are nodes, and the causal relationships between perturbation events are directed edges.

[0127] The collaborative governance decision-making module is used to identify the coupling relationship between disturbance sources and disturbance events based on the disturbance causal network diagram generated by the disturbance causal network diagram construction module. Based on this, it dynamically adjusts the cost matrix parameters in the multi-objective collaborative decision-making model, calculates the total collaborative governance cost using the multi-objective collaborative decision-making model, obtains the minimum total collaborative governance cost, derives the optimal collaborative governance instruction based on the minimum total collaborative governance cost, generates the corresponding compensation instruction based on the optimal collaborative governance instruction, and then sends the compensation instruction to the static var generator controller for execution.

[0128] The collaborative governance decision-making module includes a built-in constraint module, which contains preset constraints to ensure the multi-objective collaborative decision-making model achieves the minimum total collaborative governance cost. These constraints include dynamic capacity constraints, high-voltage passive filter branch collaborative constraints, harmonic capacity constraints, and non-negativity constraints. Specifically:

[0129] The dynamic capacity constraint is: ,in ;

[0130] The cooperative constraint of the high-voltage passive filter branch is: ;

[0131] The harmonic capacity constraint is: ;

[0132] The nonnegativity constraint is: .

[0133] In the formula, Allocate output capacity for reactive power compensation to the static var generator. Allocate power to the static var generator for compensation of the first The effective value of the output current of the subharmonic. Allocate output capacity for voltage support to the static var generator. Rated voltage, This refers to the steady-state reactive power that the high-voltage passive filter branch can provide. For the first The maximum allowable compensation current for subharmonics. This refers to the dynamic effective rated capacity of the static var generator. The rated apparent power capacity is shown on the nameplate of the static var generator. This refers to the temperature derating factor of the PN junction in an insulated-gate bipolar transistor (IGBT) chip. This is the derating factor for ambient temperature.

[0134] The power grid harmonic detection and collaborative governance system based on harmonic feature decomposition described in this embodiment is used to implement the power grid harmonic detection and collaborative governance method based on harmonic feature decomposition in the aforementioned embodiment 1. Therefore, the specific implementation methods of each module in the system can be found in the description of the corresponding steps in embodiment 1. To avoid redundancy, they will not be repeated here.

[0135] Example 3:

[0136] This invention provides an electronic device, which includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the above-mentioned method for power grid harmonic detection and collaborative management based on harmonic feature decomposition.

[0137] Example 4:

[0138] This invention provides a computer storage medium storing a computer software product, which includes several instructions to cause a computer device to execute the above-described method for power grid harmonic detection and collaborative management based on harmonic feature decomposition.

[0139] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0142] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for power grid harmonic detection and collaborative management based on harmonic characteristic decomposition, characterized in that, include: Step S1: Collect electrical quantities and device status; wherein, the electrical quantities include instantaneous values ​​of three-phase voltage and instantaneous values ​​of three-phase current, and the device status includes device spatial location information, switching status and current of each branch of the high-voltage passive filter branch, power module temperature and ambient temperature; Step S2: Encapsulate the data collected in step S1 into panoramic data frames with spatiotemporal tags. Each panoramic data frame contains a timestamp, spatial location identifier, electrical quantity vector, device state vector, and environmental parameter vector. Step S3: Process the panoramic data frames obtained in Step S2 using a preset causal inference algorithm to establish the correlation between disturbance events and generate a dynamic disturbance causal network graph; wherein, the causal inference algorithm determines the causal relationship based on the temporal relationship and statistical probability of the events, specifically: when the probability of the second disturbance event occurring within a preset time window after the first disturbance event occurs is higher than the baseline probability, it is determined that there is a causal relationship between the first disturbance event and the second disturbance event; in the disturbance causal network graph, disturbance events are nodes, and the causal relationship between disturbance events is a directed edge; Step S4: Based on the disturbance causal network diagram generated in Step S3, identify the sources of disturbance and the coupling relationship between disturbance events. Based on this, dynamically adjust the cost matrix parameters in the multi-objective collaborative decision-making model. Then, calculate the total cost of collaborative governance using the multi-objective collaborative decision-making model to obtain the minimum total cost of collaborative governance. Next, obtain the optimal collaborative governance instruction based on the minimum total cost of collaborative governance. Then, generate the corresponding compensation instruction based on the optimal collaborative governance instruction and send the compensation instruction to the static var generator controller for execution.

2. The power grid harmonic detection and collaborative management method based on harmonic feature decomposition according to claim 1, characterized in that, In step S1, electrical quantity acquisition is completed through the following operations: voltage transformers and current transformers are installed at the point of common coupling, and the instantaneous values ​​of three-phase voltage and three-phase current are acquired at a sampling rate higher than the power frequency cycle. A preset number of sampling points are set in each power frequency cycle to meet the requirements of preset harmonic analysis.

3. The power grid harmonic detection and collaborative management method based on harmonic feature decomposition according to claim 1, characterized in that, In step S4, the total cost of collaborative governance is calculated using the following formula: ； In the formula, For collaborative governance decision-making variable vectors; This is the cost matrix; It is a linear cost vector; This is the motion smoothing coefficient; This represents the change in the collaborative governance instruction vector. This is the matrix transpose symbol; in, , Allocate output capacity for reactive power compensation to the static var generator. Allocate power to the static var generator for compensation of the first The effective value of the output current of the subharmonic. For harmonic order and , For natural numbers, Allocate output capacity for voltage support to the static var generator; , This is the instruction vector to be used during collaborative governance at the current moment. This is the instruction vector used during the collaborative governance in the previous time step.

4. The power grid harmonic detection and collaborative management method based on harmonic characteristic decomposition according to claim 3, characterized in that, In step S4, when calculating the total cost of collaborative governance and obtaining the minimum total cost of collaborative governance based on the disturbance causal network diagram, the total cost of collaborative governance is calculated under preset constraints; wherein, the constraints are defined as follows: Dynamic capacity constraints: ; High-voltage passive filter branch cooperative constraints: ; Harmonic capacity constraints: ; Nonnegativity constraint: ; In the formula, Rated voltage, This refers to the steady-state reactive power that the high-voltage passive filter branch can provide. For the first The maximum allowable compensation current for subharmonics. This refers to the dynamic effective rated capacity of the static var generator; among which, , The rated apparent power capacity is shown on the nameplate of the static var generator. This refers to the temperature derating factor of the PN junction in an insulated-gate bipolar transistor (IGBT) chip. This is the derating factor for ambient temperature.

5. The power grid harmonic detection and collaborative management method based on harmonic characteristic decomposition according to claim 1, characterized in that, In step S3, the preset time window is less than one power frequency cycle.

6. A power grid harmonic detection and collaborative management system based on harmonic characteristic decomposition, characterized in that, The system is used to execute the power grid harmonic detection and collaborative management method based on harmonic feature decomposition as described in any one of claims 1 to 5, specifically including: The data acquisition module is used to collect electrical quantities and device status in the power grid; wherein, the electrical quantities include instantaneous values ​​of three-phase voltage and instantaneous values ​​of three-phase current, and the device status includes device spatial location information, switching status and current of each branch of the high-voltage passive filter branch, power module temperature and ambient temperature; The data processing module is used to encapsulate the data acquired by the data acquisition module into panoramic data frames with spatiotemporal tags. Each panoramic data frame contains a timestamp, spatial location identifier, electrical quantity vector, device state vector, and environmental parameter vector. The perturbation causal network graph construction module is used to process the panoramic data frames obtained from the data processing module using a preset causal inference algorithm, establish the correlation between perturbation events, and generate a dynamic perturbation causal network graph. Specifically, the causal inference algorithm determines causal relationships based on the temporal sequence of events and statistical probabilities. Specifically, when the probability of a second perturbation event occurring within a preset time window after the occurrence of a first perturbation event is higher than a baseline probability, a causal relationship is determined between the first and second perturbation events. In the perturbation causal network graph, perturbation events are nodes, and the causal relationships between perturbation events are directed edges. The collaborative governance decision-making module is used to identify the coupling relationship between disturbance sources and disturbance events based on the disturbance causal network diagram generated by the disturbance causal network diagram construction module. Based on this, it dynamically adjusts the cost matrix parameters in the multi-objective collaborative decision-making model, calculates the total collaborative governance cost using the multi-objective collaborative decision-making model, obtains the minimum total collaborative governance cost, derives the optimal collaborative governance instruction based on the minimum total collaborative governance cost, generates the corresponding compensation instruction based on the optimal collaborative governance instruction, and then sends the compensation instruction to the static var generator controller for execution.

7. The power grid harmonic detection and collaborative management system based on harmonic characteristic decomposition according to claim 6, characterized in that, The collaborative governance decision-making module has a built-in constraint module, which is preset with constraint conditions for the multi-objective collaborative decision-making model to obtain the minimum total cost of collaborative governance. The constraint conditions include dynamic capacity constraints, high-voltage passive filter branch collaborative constraints, harmonic capacity constraints, and non-negative constraints.

8. The power grid harmonic detection and collaborative management system based on harmonic characteristic decomposition according to claim 7, characterized in that, The dynamic capacity constraint is: ,That The cooperative constraint of the high-voltage passive filter branch is: The harmonic capacity constraint is as follows: The nonnegativity constraint is: In the formula, Allocate output capacity for reactive power compensation to the static var generator. Allocate power to the static var generator for compensation of the first The effective value of the output current of the subharmonic. Allocate output capacity for voltage support to the static var generator. Rated voltage, The steady-state reactive power that the high-voltage passive filter branch can provide. For the first The maximum allowable compensation current for subharmonics. This refers to the dynamic effective rated capacity of the static var generator. The rated apparent power capacity is shown on the nameplate of the static var generator. This refers to the temperature derating factor of the PN junction in an insulated-gate bipolar transistor (IGBT) chip. This is the derating factor for ambient temperature.

9. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a bus system. The processor and the memory are connected through the bus system. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory to implement the power grid harmonic detection and collaborative management method based on harmonic feature decomposition as described in any one of claims 1 to 5.

10. A computer storage medium, characterized in that, The computer storage medium stores a computer software product, which includes several instructions to cause a computer device to execute the power grid harmonic detection and collaborative management method based on harmonic feature decomposition as described in any one of claims 1 to 5.

Images