Power quality treatment method and device, and electronic device

Abstract

The application discloses a power quality treatment method and device and electronic equipment. The method comprises the following steps: acquiring power quality data, and identifying a corresponding harmonic source type by using a classification model; determining marginal benefits by using a benefit model and determining marginal costs by using a cost model based on the harmonic source type, wherein the marginal benefits comprise energy-saving benefits and equipment life extension benefits corresponding to power quality treatment, and the marginal costs comprise operation costs and depreciation costs of treatment equipment; determining an optimal treatment target according to the marginal benefits and the marginal costs, wherein the optimal treatment target is used to represent a power quality index at a balance point of the benefit model and the cost model; and generating a power quality treatment instruction according to the optimal treatment target. The application solves the technical problem of low treatment efficiency caused by the fact that the power quality treatment technology in the related art usually adopts a fixed control strategy.

Description

Technical Field

[0001] This application relates to the technical field of power quality management and economic operation of power systems, and more specifically, to a power quality management method, apparatus and electronic equipment. Background Technology

[0002] With the widespread application of power electronics technology in industrial, commercial, and residential sectors, the large-scale connection of nonlinear loads (such as frequency converters, LED displays, and electric arc furnaces) has led to increasingly serious harmonic pollution in the power grid. Harmonics not only cause errors in electricity metering and interfere with communication systems, but also cause overheating of electrical equipment, insulation aging, and even fires, seriously threatening the safe and stable operation of the power grid.

[0003] Power quality control equipment in related technologies typically employs fixed control strategies, including two typical modes: First, a national standard compliance strategy. This strategy uses the limits specified by national standards (e.g., total harmonic distortion <5%, or harmonic content <3%) as the control target. Once the harmonic indicators meet the national standard requirements after treatment, the equipment stops further treatment. Second, a "the cleaner the better" all-on strategy. Some systems, in order to achieve the ultimate power quality, often operate the control equipment at maximum output to eliminate all harmonic components as much as possible. While this can achieve extremely low harmonic content, it presents a serious problem of over-treatment.

[0004] However, the above fixed strategy usually has the following drawbacks:

[0005] First, there is a lack of differentiated response capability to the characteristics of harmonic sources: different types of harmonic sources (such as frequency converters mainly generating 5th and 7th harmonics, and LED screens mainly generating 3rd harmonics) have different impact mechanisms on the power grid and are more difficult to manage. Fixed strategies usually adopt uniform management objectives and parameters, and cannot dynamically adjust the management strategy according to the type of harmonic source, resulting in low management efficiency.

[0006] Secondly, there is a lack of a mechanism for quantifying benefits: existing systems mainly focus on power quality indicators, while ignoring the energy waste of the governance process itself and the unnecessary wear and tear on equipment lifespan. They cannot calculate in real time the balance between the energy-saving benefits and lifespan extension benefits of governance and the operating costs of governance equipment, making it difficult for users to obtain the maximum economic return through governance.

[0007] Third, the control strategies are static and rigid, lacking adaptability: the control parameters of existing systems are mostly factory-set or manually fixed, and cannot be adaptively adjusted according to external electricity price fluctuations (such as peak-valley price differences), the aging of the treatment equipment itself, and sudden changes in load conditions. For example, maintaining high-intensity treatment during periods of low electricity prices may not be economical; as equipment ages, its efficiency declines, leading to increased treatment costs, and if the strategy is not adjusted in time, the treatment benefits may be negative.

[0008] There is currently no effective solution to the above problems. Summary of the Invention

[0009] This application provides a power quality management method, apparatus, and electronic device to at least solve the technical problem of low management efficiency caused by the use of fixed control strategies in power quality management technologies in related technologies.

[0010] According to one aspect of the embodiments of this application, a power quality management method is provided, comprising: acquiring power quality data and identifying corresponding harmonic source types using a classification model; determining marginal revenue corresponding to the power quality data using a benefit model based on the harmonic source types, and determining marginal cost corresponding to the power quality data using a cost model, wherein the marginal revenue includes energy-saving benefits and equipment life-extending benefits corresponding to power quality management, and the marginal cost includes operating costs and depreciation costs of the management equipment; determining an optimal management target based on the marginal revenue and marginal cost, wherein the optimal management target represents a power quality index located at the equilibrium point of the benefit model and the cost model; and generating power quality management instructions based on the optimal management target, wherein the power quality management instructions are used to control the operating state of the management equipment to minimize physical losses between the power grid and the management equipment.

[0011] Optionally, power quality data is acquired, including: collecting three-phase voltage and three-phase current at power grid monitoring points according to a preset acquisition frequency; determining the total harmonic distortion (THD), harmonic content, harmonic centroid offset, harmonic spectrum entropy, THD fluctuation, and power consumption of the control equipment based on the three-phase voltage and three-phase current; and determining power quality data based on the THD, harmonic content, harmonic centroid offset, harmonic spectrum entropy, THD fluctuation, and power consumption of the control equipment.

[0012] Optionally, a classification model is used to identify the corresponding harmonic source type, including: determining the importance scores of total harmonic distortion (THD), harmonic content, harmonic centroid shift, harmonic spectral entropy, and THD fluctuation in the power quality data, where the importance score represents the contribution of different dimensional features to distinguishing harmonic source types; generating feature vectors for harmonic source type identification based on the importance scores, and standardizing the feature vectors; analyzing the standardized feature vectors using a random forest classification model to obtain classification results corresponding to the power quality data; and determining the category with the highest posterior probability in the classification results that exceeds a preset confidence threshold as the harmonic source type corresponding to the power quality data.

[0013] Optionally, a benefit model is used to determine the marginal benefit corresponding to the power quality data, including: obtaining a first benefit model and a second benefit model corresponding to the harmonic source type, wherein the first benefit model is used to quantify the energy-saving benefit corresponding to power quality management, and the second benefit model is used to quantify the equipment life extension benefit corresponding to power quality management; the first benefit model is used to analyze the change in total harmonic distortion rate, the change rate of power loss caused by the change in unit total harmonic distortion rate, the annual equivalent operating hours, and the electricity price to obtain the energy-saving benefit; the second benefit model is used to analyze the purchase cost, design life, life extension ratio, and annual equivalent operating hours of the protected equipment to obtain the equipment life extension benefit, wherein the life extension ratio is positively correlated with the life sensitivity coefficient of the protected equipment, and the life sensitivity coefficient is used to reflect the sensitivity of the protected equipment to harmonic thermal stress; the marginal benefit corresponding to the power quality data is determined based on the energy-saving benefit and the equipment life extension benefit.

[0014] Optionally, a cost model is used to determine the marginal cost corresponding to the power quality data, including: obtaining a first cost model and a second cost model corresponding to the harmonic source type, wherein the first cost model is used to quantify the operating cost of the treatment equipment, and the second cost model is used to quantify the depreciation cost of the treatment equipment; the first cost model is used to analyze the efficiency coefficient of the treatment equipment, the compensation capacity increment required for a unit change in total harmonic distortion, the annual equivalent operating hours, and the electricity price to obtain the operating cost of the treatment equipment; the second cost model is used to analyze the purchase cost, design life, aging acceleration factor, and annual equivalent operating hours of the treatment equipment to obtain the depreciation cost of the treatment equipment; and the marginal cost corresponding to the power quality data is determined based on the operating cost and the depreciation cost.

[0015] Optionally, the optimal governance objective is determined based on marginal revenue and marginal cost, including: determining the marginal revenue curve corresponding to the marginal revenue and the marginal cost curve corresponding to the marginal cost based on the type of harmonic source; determining the minimum total harmonic distortion (THD) rate corresponding to the maximum output of the governance equipment, using the minimum THD rate as the lower limit and the THD rate in the power quality data as the upper limit to obtain the target harmonic range; and solving for the marginal revenue curve and the marginal cost curve within the target harmonic range to obtain the optimal governance objective.

[0016] Optionally, the method further includes: within the target harmonic interval, comparing the marginal revenue curve and the marginal cost curve to obtain a comparison result; if the comparison result indicates that there is an intersection point where marginal revenue equals marginal cost within the target harmonic interval, determining the total harmonic distortion rate corresponding to the intersection point as the optimal governance target; if the comparison result indicates that the marginal revenue within the target harmonic interval is greater than the marginal cost, determining the minimum total harmonic distortion rate as the optimal governance target; if the comparison result indicates that the marginal revenue within the target harmonic interval is less than the marginal cost, determining the total harmonic distortion rate in the power quality data as the optimal governance target.

[0017] Optionally, generating power quality management instructions based on the optimal management objective includes: generating a start management instruction or an enhanced management instruction when the difference between the total harmonic distortion rate in the power quality data and the optimal management objective is greater than a preset threshold, and the marginal benefit is greater than the product of the marginal cost and the first weight, wherein the first weight is the sum of 1 and a preset hysteresis coefficient; generating a stop management instruction or a weaken management instruction when the total harmonic distortion rate in the power quality data is less than or equal to the optimal management objective, or the marginal benefit is less than the product of the marginal cost and the second weight, wherein the second weight is the difference between 1 and the preset hysteresis coefficient.

[0018] Optionally, the method further includes: acquiring the grid electricity price signal and updating the optimal governance target based on the grid electricity price signal, wherein the grid electricity price signal is used to represent the change signal corresponding to the electricity price; updating the efficiency coefficient and aging acceleration factor corresponding to the governance equipment according to the operating status of the governance equipment to obtain the target efficiency coefficient and the target aging acceleration factor, and updating the optimal governance target based on the target efficiency coefficient and the target aging acceleration factor; and correcting the model weights of the benefit model and the cost model and updating the optimal governance target when the rate of change of the total harmonic distortion rate in the power quality data exceeds a preset mutation threshold.

[0019] According to another aspect of the embodiments of this application, a power quality management device is also provided, comprising: an identification module for acquiring power quality data and identifying the corresponding harmonic source type using a classification model; a first determination module for determining the marginal benefit corresponding to the power quality data based on the harmonic source type using a benefit model and determining the marginal cost corresponding to the power quality data using a cost model, wherein the marginal benefit includes energy-saving benefits and equipment life extension benefits corresponding to power quality management, and the marginal cost includes the operating cost and depreciation cost of the management equipment; a second determination module for determining the optimal management target based on the marginal benefit and marginal cost, wherein the optimal management target represents the power quality index located at the equilibrium point of the benefit model and the cost model; and a generation module for generating power quality management instructions based on the optimal management target, wherein the power quality management instructions are used to control the operating state of the management equipment to minimize the physical loss between the power grid and the management equipment.

[0020] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory and a processor, wherein the memory is used to store program instructions; and the processor is connected to the memory and used to execute the above-described power quality management method.

[0021] According to another aspect of the embodiments of this application, a non-volatile storage medium is also provided, the non-volatile storage medium including a stored computer program, wherein the device containing the non-volatile storage medium executes the above-mentioned power quality management method by running the computer program.

[0022] According to another aspect of the embodiments of this application, a computer program product is also provided, including computer instructions that, when executed by a processor, implement the above-described power quality management method.

[0023] In this embodiment, power quality data is acquired, and a classification model is used to identify the corresponding harmonic source types. Based on the harmonic source types, a benefit model is used to determine the marginal benefit corresponding to the power quality data, and a cost model is used to determine the marginal cost corresponding to the power quality data. The marginal benefit includes energy-saving benefits and equipment life-extending benefits corresponding to power quality management, and the marginal cost includes the operating cost and depreciation cost of the management equipment. An optimal management target is determined based on the marginal benefit and marginal cost, whereby the optimal management target represents the power quality index located at the equilibrium point of the benefit model and the cost model. Power quality management instructions are generated based on the optimal management target, whereby the power quality management instructions are used to control the operating state of the management equipment to minimize the physical loss between the power grid and the management equipment. This achieves the goal of suppressing grid harmonic losses while avoiding increased energy consumption and lifespan reduction of the management equipment due to overcompensation, thereby realizing a dynamic balance between the power grid operating efficiency and the physical lifespan of the management equipment. This solves the technical problem of low management efficiency caused by the fixed control strategies typically used in power quality management technologies in related technologies. Attached Figure Description

[0024] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0025] Figure 1 This is a hardware structure diagram of a computer terminal for implementing a power quality management method according to an embodiment of this application;

[0026] Figure 2 This is a flowchart of a power quality management method according to an embodiment of this application;

[0027] Figure 3 This is a structural diagram of a power quality management device according to an embodiment of this application. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] First, some nouns or terms that appear in the explanation of the embodiments of this application shall be interpreted as follows:

[0031] Power quality management refers to the process of detecting, compensating for, suppressing, or eliminating power quality problems (such as harmonics, reactive power, voltage sags, voltage fluctuations and flicker, three-phase imbalance, etc.) in a power system through physical, electronic, or control means.

[0032] Power quality management equipment refers to hardware devices or systems specifically designed to implement power quality management functions. Examples include APF (Active Power Filter), SVG (Static Var Generator), and other dynamic compensation devices.

[0033] THD (Total Harmonic Distortion): A measure of the degree of waveform distortion in a power system. It is defined as the ratio of the effective value of harmonic current (or voltage) to the effective value of the fundamental current (or voltage), usually expressed as a percentage. The higher the THD, the worse the power quality and the more severe the harmonic pollution.

[0034] HR (Harmonic Ratio / Harmonic Rate): refers to the ratio of the effective value of a first harmonic current (or voltage) to the effective value of the fundamental current (or voltage). For example, HR3 represents the third harmonic content rate, used to analyze the proportion of harmonics at a specific frequency.

[0035] Marginal benefits: In this application, it refers to the additional economic benefits that can be brought about by reducing the total harmonic distortion (THD) by one unit (e.g., 1%), mainly including energy-saving benefits (electricity cost savings due to reduced line and transformer losses) and equipment life extension benefits (potential maintenance or replacement cost savings due to reduced thermal stress and delayed equipment aging).

[0036] Marginal cost: In this application, it refers to the additional cost incurred for each additional unit (e.g., 1%) reduction in total harmonic distortion (THD). This mainly includes the additional operating electricity costs (increased power consumption) of the control equipment (e.g., APF) due to the increased output current, as well as the accelerated equipment depreciation costs.

[0037] Random Forest (RF): A machine learning algorithm based on ensemble learning that performs classification or regression prediction by constructing multiple decision trees and combining their outputs (e.g., voting or averaging). In this application, it is used to identify dominant harmonic source types based on multidimensional harmonic fingerprint features, exhibiting high accuracy, strong robustness, and anti-overfitting ability.

[0038] Gini impurity: A metric used when splitting nodes in a decision tree to evaluate the purity of the dataset. Lower Gini impurity indicates greater consistency in the class of samples within a node. In random forest training, the algorithm tends to select features that reduce Gini impurity the most for splitting.

[0039] BSA (Binary Search Algorithm): An algorithm for finding a specific element in an ordered array or interval, reducing the search range by half after each comparison. In this application, it is used to quickly approximate the optimal THD value that makes the revenue equal to the cost by repeatedly taking the midpoint and calculating the marginal revenue within the possible THD interval.

[0040] This application provides a power quality management method, which can be operated in... Figure 1 The computer terminal shown is described below.

[0041] The power quality management method embodiments provided in this application can be executed on mobile terminals, computer terminals, or similar computing devices. Figure 1 A hardware block diagram of a computer terminal for implementing power quality management methods is shown. Figure 1 As shown, the computer terminal 10 may include one or more processors (shown as 102a, 102b, ..., 102n in the figure) (the processor may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission module 106 for communication functions connected via wired and / or wireless networks. In addition, it may also include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the I / O interface), a network interface, and a BUS bus. Those skilled in the art will understand that... Figure 1The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0042] It should be noted that the aforementioned one or more processors and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10. As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0043] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the power quality management method in this embodiment. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned power quality management method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0044] The transmission module 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission module 106 includes a network interface controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission module 106 may be a radio frequency (RF) module, used for wireless communication with the Internet.

[0045] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10.

[0046] It should be noted here that, in some optional embodiments, the above... Figure 1The computer terminal shown may include hardware elements (including circuitry), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware and software elements. It should be noted that... Figure 1 This is only one instance of a specific particular instance, and is intended to illustrate the types of components that may exist in the aforementioned computer terminal.

[0047] In the above operating environment, this application provides an embodiment of a power quality management method. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Also, although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than that shown here.

[0048] Figure 2 This is a flowchart of a power quality management method according to an embodiment of this application, such as... Figure 2 As shown, the method includes the following steps:

[0049] Step S202: Obtain power quality data and use a classification model to identify the corresponding harmonic source type.

[0050] Step S204: Based on the type of harmonic source, a benefit model is used to determine the marginal benefit corresponding to the power quality data, and a cost model is used to determine the marginal cost corresponding to the power quality data. The marginal benefit includes the energy-saving benefit and the equipment life extension benefit corresponding to power quality management, and the marginal cost includes the operating cost and depreciation cost of the management equipment.

[0051] Step S206: Determine the optimal governance objective based on marginal revenue and marginal cost, wherein the optimal governance objective is used to represent the power quality index located at the equilibrium point of the benefit model and the cost model.

[0052] Step S208: Generate power quality management instructions based on the optimal management objective. The power quality management instructions are used to control the operating status of the management equipment to minimize the physical losses between the power grid and the management equipment.

[0053] Through steps S202 to S208, the goal of suppressing grid harmonic losses while avoiding increased energy consumption and lifespan reduction of the power quality management equipment due to overcompensation is achieved. This realizes a dynamic balance between grid operating efficiency and the physical lifespan of the power quality management equipment, thereby solving the technical problem of low efficiency caused by the fixed control strategies typically used in related power quality management technologies. The following is a detailed explanation.

[0054] In this embodiment, the power quality management system includes the following functional units:

[0055] 1. Power Quality Monitoring Unit: Deployed on the incoming line side or upstream of critical loads, with a power quality level not lower than 12.8. The sampling rate is used to synchronously acquire the instantaneous waveforms of three-phase voltage and current in real time, providing high-precision raw data for subsequent processing.

[0056] 2. Equipment Identification Engine: Based on real-time waveform extraction, a harmonic fingerprint vector containing 7-dimensional features such as total harmonic distortion rate, harmonic centroid shift, and spectral entropy is extracted. A pre-trained random forest classifier is used to identify the dominant harmonic source type, providing a basis for differentiated governance.

[0057] 3. Benefit Model Library: Stores marginal benefit characteristic curves and parameters corresponding to different harmonic source types (such as logarithmically decreasing type, piecewise function type, etc.), and supports dynamic calibration of model parameters through online data identification to adapt to equipment aging or changes in operating conditions.

[0058] 4. Marginal Analysis Engine: Calculates the marginal benefits (including energy-saving benefits and lifespan extension benefits) and marginal costs (including operating costs and depreciation costs of governance equipment) of the current governance level in real time, quantifying the physical and economic value of governance activities.

[0059] 5. Optimize the decision engine: Based on the marginal analysis results, run the optimal stopping algorithm to determine the optimal governance target value that minimizes the total physical loss of the system, and generate power quality governance instructions.

[0060] 6. Governance Equipment Interface: Through communication protocols such as Modbus TCP and CANopen, the power quality governance commands generated by the optimization decision engine are sent to governance equipment such as APF and SVG, and the equipment status feedback is received to realize closed-loop control.

[0061] In step S202 above, acquiring power quality data includes: collecting the three-phase voltage and three-phase current at the power grid monitoring point according to a preset acquisition frequency; determining the total harmonic distortion (THD), harmonic content, harmonic centroid offset, harmonic spectrum entropy, THD fluctuation, and power consumption of the control equipment based on the THD and three-phase current; and determining power quality data based on the THD, harmonic content, harmonic centroid offset, harmonic spectrum entropy, THD fluctuation, and power consumption of the control equipment.

[0062] Specifically, the power quality monitoring unit first synchronously samples the grid monitoring points (usually located on the user's incoming line side or at the front end of key nonlinear loads) at a preset sampling frequency (e.g., 12.8 kHz, 256 points per power frequency cycle) to collect the three-phase voltage in real time. and three-phase current The raw waveform data is obtained. Then, the data is processed using a sliding window with a fixed time window (e.g., every 10 power frequency cycles, i.e., 200ms as a data window), and key power quality indicators are calculated according to the following formulas:

[0063] 1. Total Harmonic Distortion (THD):

[0064] The formula used to reflect the overall level of harmonic content in the power grid voltage or current is as follows:

[0065]

[0066] in, For the first RMS value of subharmonic current This represents the effective value of the fundamental current.

[0067] 2. Harmonic content ( ):

[0068] This reflects the proportion of a specific harmonic in the fundamental frequency and is used to locate the main harmonic source. The calculation formula is:

[0069]

[0070] 3. Harmonic centroid offset ( ):

[0071] This reflects the centroid position of harmonic energy in the spectral distribution. A lower centroid value indicates that lower-order harmonics (such as the 3rd and 5th) are dominant; a higher centroid value indicates that higher-order harmonics account for a larger proportion. The calculation formula is:

[0072]

[0073] 4. Harmonic spectral entropy ( ):

[0074] This reflects the flatness or complexity of the harmonic spectrum. A higher entropy value indicates more dispersed harmonic components and a more complex spectrum (e.g., under an electric arc furnace load); a lower entropy value indicates more concentrated harmonic components and a simpler spectrum (e.g., under an LED screen load). The calculation formula is:

[0075]

[0076]

[0077] in, For the first The proportion of subharmonic current to total harmonic current energy. For the first The effective value of the subharmonic current.

[0078] 5. THD volatility ( ):

[0079] It reflects the stability of the load operation. It is characterized by calculating the standard deviation of the THD value within a sliding window; the greater the fluctuation, the more drastic the load changes (e.g., in an electric arc furnace). The calculation formula is:

[0080]

[0081] in, The number of sampling points within the sliding window (e.g.) ,correspond ), For the first THD values ​​at each sampling point This represents the average THD value within the sliding window.

[0082] 6. Power consumption of the treatment equipment itself:

[0083] Real-time active power consumption of the power management equipment is obtained through internal sensors of the APF or an external power meter. This serves as the basis for subsequent calculations of marginal cost.

[0084] Finally, by combining the various power quality indicators obtained from the above calculations, the current power quality data is determined.

[0085] Through the above steps, a comprehensive and multi-dimensional quantitative description of the power quality status of the power grid is achieved, providing an accurate data foundation for subsequent intelligent governance decisions.

[0086] Furthermore, a classification model is used to identify the corresponding harmonic source types, including: determining the importance scores of total harmonic distortion (THD), harmonic content, harmonic centroid shift, harmonic spectral entropy, and THD fluctuation in the power quality data, where the importance score represents the contribution of different dimensional features to distinguishing harmonic source types; generating feature vectors for harmonic source type identification based on the importance scores, and standardizing the feature vectors; analyzing the standardized feature vectors using a random forest classification model to obtain classification results corresponding to the power quality data; and identifying the category with the highest posterior probability in the classification results that exceeds a preset confidence threshold as the harmonic source type corresponding to the power quality data.

[0087] Traditional methods rely solely on THD amplitude, failing to differentiate between the characteristics of various harmonic sources. Therefore, this application introduces a harmonic source type identification method based on harmonic fingerprints. This method constructs a multi-dimensional harmonic fingerprint vector and employs a supervised learning classifier (such as a random forest classification model) to achieve accurate identification. The specific implementation process is as follows:

[0088] 1. Feature Importance Assessment:

[0089] First, the system needs to determine the importance scores of each dimension in the power quality data in order to filter out the features that contribute the most to distinguishing the types of harmonic sources, thereby optimizing the model input and reducing computational complexity.

[0090] Specifically, during the model training phase, historical labeled data can be used to evaluate the importance of each feature dimension, such as using the reduction in Gini impurity as an importance metric. For each decision tree in the random forest, the reduction in Gini impurity for each feature at node splits is calculated, and the average reduction for the same feature across all decision trees is taken to obtain the feature's importance score. Based on extensive experimental data, the importance ranking of typical features is as follows: .

[0091] 2. Feature vector construction and standardization:

[0092] Based on importance scores, a 7-dimensional harmonic fingerprint feature vector is constructed. .

[0093] If the computing power of the embedded processor on site is limited, the features can be reduced in dimensionality based on the importance score (e.g., only the top 5 important features are retained). However, in this preferred embodiment, a full 7-dimensional feature vector is used to ensure recognition accuracy, and each component is calculated in real time in step S202.

[0094]

[0095] Among them, the physical meaning of each dimension is clear, and it has significant distinguishability for different harmonic sources, for example:

[0096] Frequency converter: , and High, medium;

[0097] LED display screen: , Extremely high (can reach) above), Lower;

[0098] UPS: , and Dominant Smaller;

[0099] Electric arc furnace: Wider Larger (rich spectrum) great;

[0100] Photovoltaic inverters: lower THD ( ), mainly high-frequency harmonics, Relatively large.

[0101] Because the dimensions and numerical ranges of the various characteristics differ greatly (for example, THD is usually between 0 and 50%), Typically between 2 and 30, directly inputting these features into the model can lead to the model's decisions being dominated by features with larger numerical values. Therefore, the feature vectors need to be standardized using Z-score, as shown in the following expression:

[0102]

[0103] in, These are the original eigenvalues. These are the standardized feature values. and The first The mean and standard deviation of each feature across all training samples.

[0104] Standardized target feature vector Recorded as: .

[0105] 3. Random Forest Classification and Obtaining Classification Results:

[0106] The standardized target feature vector Inputting a pre-trained classification model, in this embodiment of the application, a random forest is selected as the classification model, as detailed below:

[0107] (1) Model structure:

[0108] This random forest classification model is derived from The decision tree is integrated into a single tree (preferred in this embodiment). Each decision tree uses a bootstrap sampling method to generate a subset during training, and randomly selects a subset of features (preferred) when splitting nodes. Optimal splitting of features is performed to increase model diversity.

[0109] (2) Reasoning process:

[0110] Target feature vector The data is simultaneously input into these 100 decision trees. Each decision tree divides the samples into leaf nodes according to the internally stored splitting rules (threshold and feature index). The distribution of training sample categories stored in the leaf nodes determines the voting weight of the tree for each category.

[0111] (3) Calculation of posterior probability:

[0112] By summarizing the voting results of all decision trees, we can determine which decision tree the current power quality data belongs to. Posterior probability of a harmonic source:

[0113]

[0114] Among them, table For indicator functions, if the first... The predicted result for the number of trees is If the value is 1, then the value is 1; otherwise, it is 0.

[0115] Final classification results The category with the highest posterior probability:

[0116]

[0117] 4. Confidence level determination and type confirmation:

[0118] In this embodiment of the application, in order to avoid misjudgment caused by instantaneous noise or transient disturbances, a timing smoothing and confidence threshold mechanism is also introduced to determine the final harmonic source type.

[0119] (1) Confidence check:

[0120] Obtain the maximum delay probability At the same time, a confidence threshold is set. .

[0121] like This indicates that the model has a high degree of confidence in the recognition result, and the current dominant type is recorded as […]. ;

[0122] like If the current working conditions are complex or the features are ambiguous, the identification result is considered invalid. The system will retain the previous confirmation type or mark it as "unknown / mixed" and wait for subsequent data.

[0123] (2) Time series smoothing:

[0124] Even if the confidence level of a single recognition result meets the standard, the system still maintains a length of [length missing]. A sliding window records the dominant category of the three most recent identifications. (Choose the category with the highest probability).

[0125] A change in harmonic source type is only formally confirmed when the dominant category is the same for three consecutive identifications, and the confidence level of each identification exceeds the threshold. Upon confirmation of the switch, the benefit model library is triggered, the corresponding new type of marginal benefit curve parameters are loaded, and a new optimal governance target calculation is initiated to ensure that the control strategy matches the current harmonic source characteristics in real time.

[0126] In this embodiment, to adapt to long-term drift such as equipment aging and changes in power grid conditions, the system also supports an online self-learning mode. For example, when the identification results within a certain period are manually confirmed to be consistent with the actual equipment operation records, the feature data of that period can be labeled and added to the training set. The model can be retrained periodically (e.g., monthly) in non-real-time tasks and updated to the online engine through a security mechanism to continuously improve the identification accuracy.

[0127] In the above process, by constructing standardized feature vectors based on importance scores and combining them with the posterior probability confidence judgment mechanism of random forest, the effective fusion of features of different dimensions and the noise-resistant smoothing of the recognition results are achieved, thereby significantly improving the accuracy, robustness and real-time stability of harmonic source type recognition.

[0128] In step S204 above, the aim is to quantify the energy-saving and life-extending benefits of governance by combining the types of harmonic sources, as well as the operating and depreciation costs of the governance equipment itself, so as to achieve real-time and accurate quantification of the economic benefits of power quality governance, and provide a data basis for finding the optimal governance balance point where marginal benefits equal marginal costs.

[0129] Optionally, a benefit model is used to determine the marginal benefit corresponding to power quality data, including: obtaining a first benefit model and a second benefit model corresponding to the harmonic source type, wherein the first benefit model is used to quantify the energy-saving benefit corresponding to power quality management, and the second benefit model is used to quantify the equipment life extension benefit corresponding to power quality management; the first benefit model is used to analyze the change in total harmonic distortion rate, the change rate of power loss caused by the change in unit total harmonic distortion rate, the annual equivalent operating hours, and the electricity price to obtain the energy-saving benefit; the second benefit model is used to analyze the purchase cost, design life, life extension ratio, and annual equivalent operating hours of the protected equipment to obtain the equipment life extension benefit, wherein the life extension ratio is positively correlated with the life sensitivity coefficient of the protected equipment, and the life sensitivity coefficient is used to reflect the sensitivity of the protected equipment to harmonic thermal stress; the marginal benefit corresponding to power quality data is determined based on the energy-saving benefit and the equipment life extension benefit. The specific process analysis is as follows:

[0130] 1. Energy-saving benefit model (first benefit model):

[0131] The core logic of the energy-saving benefit model is that reducing losses equals saving money. It calculates the reduction in additional heat losses from power grid lines, transformers, and other equipment due to the weakening of the skin effect and eddy current effect after the harmonic current is reduced. This reduction in losses, multiplied by the electricity price and operating time, represents the direct savings in electricity costs. The specific expression is as follows:

[0132]

[0133] in, For annual energy-saving benefits, The change in THD The rate of change in power loss per unit change in THD. Annual equivalent operating hours (can be taken as...) ), Price of electricity (real-time electricity price, yuan / kWh).

[0134] It should be noted that, (in, For THD, Adding power loss to harmonics, , For the first RMS value of subharmonic current For the first The equivalent resistance under subharmonics can theoretically be calculated from system parameters. However, in complex industrial environments, user-side equipment is diverse (mixed loads), line parameters are unknown, and harmonic sources change dynamically. Accurately obtaining the equivalent resistance of each device is challenging. The parameters are quite difficult to determine.

[0135] Therefore, this application provides an online parameter identification method that can be used to directly obtain... This includes the following two methods:

[0136] Method 1: Online parameter identification.

[0137] During periods of change in the output compensation current of the treatment equipment (such as APF) (e.g., from...) Rated compensation increased to ), synchronously record the changes in THD and the change in total active power input to the system .

[0138] Assuming the load's own power remains stable for a short period of time, then It is approximately equal to the change in harmonic-induced losses. Calculated by averaging multiple measurements:

[0139]

[0140] To eliminate the influence of load power fluctuations, the following measures can be taken to improve identification accuracy:

[0141] (1) Identify during periods of relatively stable load (such as at night or during production breaks);

[0142] (2) To and Perform low-pass filtering to remove high-frequency noise;

[0143] (3) Perform two disturbances in the positive and negative directions (increase compensation → decrease compensation), and take the average of the absolute values ​​to offset the load trend term.

[0144] Method 2: Experience-based table lookup based on equipment type.

[0145] When the system lacks online identification conditions (such as APF running at full capacity for extended periods, or severe load fluctuations preventing the identification of a stable window), the following empirical values ​​should be looked up in a table based on the harmonic source type output by the device identification engine. Initial valuation:

[0146] Table 1 Empirical Table of Power Loss Change Rate

[0147]

[0148] in, The rated power of the main harmonic load equipment currently being treated (or the total load of the branch), and the coefficient. The calculation results of typical parameters are derived from the aforementioned theoretical mechanisms and corrected through a large amount of field measurement data. During system operation, if reliable online identification results are subsequently obtained, the system will automatically update and overwrite the lookup table values, thereby achieving continuous optimization of the model.

[0149] 2. Equipment life extension benefit model (second benefit model):

[0150] The core logic of the equipment lifespan extension benefit model is that reducing heat generation equates to extending lifespan. It is based on the Arrhenius aging principle, which posits that the additional temperature rise caused by harmonics accelerates the aging of insulation materials. By reducing the effective value of harmonic current, the equipment operating temperature decreases, thereby extending its actual service life. The specific expression is as follows:

[0151]

[0152] in, Extending the annual lifespan of equipment will bring benefits. The cost of purchasing the protected equipment. The design life of the protected equipment;

[0153] Percentage increase in the lifespan of the protected equipment:

[0154]

[0155] in, This is the sensitivity coefficient for equipment lifespan.

[0156] The actual expected lifespan of the protected equipment under current operating conditions:

[0157]

[0158] in, For standard reference life, For the equipment operating temperature, For standard reference temperature, Activation energy (a physical constant reflecting the temperature sensitivity of a material's aging process) The larger the value, the more sensitive the material's aging process is to temperature changes. For power electronic devices, this is approximately... ), is the Boltzmann constant.

[0159] 3. Marginal revenue:

[0160] The total marginal benefit resulting from a unit change in THD is calculated by combining the aforementioned energy-saving benefits and equipment life-extending benefits. for:

[0161]

[0162] It should be noted that the mitigation equipment refers to equipment that actively applies mitigation measures, such as APF or SVG, and is the bearer of the costs, with its own power consumption, depreciation, and maintenance costs constituting the marginal cost; while the protected equipment refers to loads or infrastructure in the power grid affected by harmonic pollution, such as transformers, motors, cables in factories or LED displays in commercial complexes, and is the beneficiary of the benefits, with the electricity cost savings (energy saving benefits) and the extended lifespan value (lifespan benefits) due to the reduction of harmonics constituting the marginal benefits.

[0163] In the above process, by constructing a comprehensive marginal benefit quantification model that includes energy-saving benefits and life extension benefits, a comprehensive assessment of the economic value of power quality governance is achieved. This avoids over- or under-governance caused by considering only a single indicator, and ensures that the system makes adaptive decisions under the premise of maximizing economic benefits.

[0164] Optionally, a cost model is used to determine the marginal cost corresponding to the power quality data, including: obtaining a first cost model and a second cost model corresponding to the harmonic source type, wherein the first cost model is used to quantify the operating cost of the treatment equipment, and the second cost model is used to quantify the depreciation cost of the treatment equipment; the first cost model is used to analyze the efficiency coefficient, the compensation capacity increment required for a unit change in total harmonic distortion, the annual equivalent operating hours, and the electricity price of the treatment equipment to obtain the operating cost of the treatment equipment; the second cost model is used to analyze the purchase cost, design life, aging acceleration factor, and annual equivalent operating hours of the treatment equipment to obtain the depreciation cost of the treatment equipment; and the marginal cost corresponding to the power quality data is determined based on the operating cost and the depreciation cost. The specific process analysis is as follows:

[0165] 1. Operating cost model (first cost model):

[0166] The operating cost model, by linking the APF efficiency coefficient, the incremental compensation capacity, electricity price, and operating time, quantifies in real time the additional electricity costs directly resulting from increased harmonic mitigation output, reflecting the immediate energy economic costs of mitigation efforts. The specific expression is as follows:

[0167]

[0168] in, Annual operating costs;

[0169] APF power consumption increment:

[0170]

[0171] in, No-load loss (approximately) (rated capacity). The efficiency coefficient (approximately) The amount of compensation capacity required to reduce THD per unit It can be obtained through online measurement.

[0172] 2. Depreciation Cost Model (Second Cost Model):

[0173] The depreciation cost model is based on Miner's linear cumulative damage theory. Combining purchase cost, design life, and aging acceleration factors, it quantifies the incremental depreciation over the entire equipment lifecycle due to accelerated wear and tear on core components such as IGBTs and capacitors caused by high-frequency switching and ripple current. This reflects the economic depreciation of hardware lifespan caused by management actions. The specific expression is as follows:

[0174]

[0175] in, Annual depreciation cost The cost of purchasing APF For the design life of the APF;

[0176] Cumulative damage increment over APF lifetime:

[0177]

[0178] in, This represents the effective value increment of the ripple current. This is the rated current of the APF. Aging Accelerator (typical value) ).

[0179] 3. Marginal cost:

[0180] Calculate the total marginal revenue resulting from a unit change in THD by combining the above operating costs and depreciation costs. for:

[0181]

[0182] In the above process, by constructing a comprehensive cost model that covers operating electricity costs and equipment depreciation, the cost of power quality management equipment investment is refined and dynamically quantified. This avoids the drawbacks of traditional methods that only consider fixed costs or ignore the economic impact of equipment aging, thus ensuring that the power quality management strategy achieves true economic optimization based on the consideration of the entire life cycle cost of equipment.

[0183] In step S206 above, the optimal governance objective is determined based on marginal revenue and marginal cost, including: determining the marginal revenue curve corresponding to the marginal revenue and the marginal cost curve corresponding to the marginal cost based on the type of harmonic source; determining the minimum total harmonic distortion rate corresponding to the maximum output of the governance equipment, using the minimum total harmonic distortion rate as the lower limit and the total harmonic distortion rate in the power quality data as the upper limit to obtain the target harmonic range; and solving the marginal revenue curve and the marginal cost curve within the target harmonic range to obtain the optimal governance objective.

[0184] In this embodiment of the application, depending on the type of harmonic source, the corresponding marginal characteristic curve structure (such as logarithmically decreasing type, piecewise function type, etc.) and initial parameters can be obtained from the benefit model library, as shown in Table 2, thereby generating the marginal revenue curve and marginal cost curve corresponding to the marginal revenue and marginal cost.

[0185] Table 2 Examples of marginal revenue curve structures for different harmonic source types

[0186]

[0187] in, For marginal revenue curves / functions, For the revenue intercept parameter, This is the profit decay coefficient. To maximize marginal revenue, For constant return values, The slope coefficient, (THD) represents the current total harmonic distortion rate. This is the theoretical zero point or reference threshold.

[0188] Marginal cost is usually approximately constant (power consumption is proportional to compensation current) and does not change with... The (THD) value changes significantly with changes in output, or increases slightly with increasing output, and can be simplified as follows: (constant). The specific value is calculated based on the APF's rated parameters, electricity price, and depreciation period, with a typical range of values. Thousand yuan / % / year.

[0189] Furthermore, within the target harmonic range, the marginal revenue curve and the marginal cost curve are compared to obtain the comparison results. If the comparison results indicate that there is an intersection point within the target harmonic range where marginal revenue equals marginal cost, the total harmonic distortion rate corresponding to this intersection point is determined as the optimal control target. If the comparison results indicate that the marginal revenue within the target harmonic range is greater than the marginal cost, the minimum total harmonic distortion rate is determined as the optimal control target. If the comparison results indicate that the marginal revenue within the target harmonic range is less than the marginal cost, the total harmonic distortion rate in the power quality data is determined as the optimal control target. The specific expression is as follows:

[0190]

[0191] like For all If it is true, then take (Minimum THD corresponding to maximum APF output); if For all If it is true, then take (Current THD).

[0192] In this embodiment, the decision engine can be further optimized to initiate a binary search algorithm within the target harmonic range. Inner iteration, taking the midpoint each time. calculate and Compare until the interval length is reached. .

[0193] Specifically, take the midpoint each time. Substitute into the marginal revenue curve Calculate the current revenue and compare it with the constant marginal cost. Compare, if, if This indicates that the benefit exceeds the cost, and the optimal solution lies in the left half of the interval. Therefore, the upper limit is updated to... Conversely, the lower limit will be updated to... By continuously narrowing the search interval until its length is less than 0.1%, the optimal governance objective that makes marginal revenue equal to marginal cost can be approximated quickly and with high precision. .

[0194] In the above process, by constructing a dynamic balance model of marginal revenue and marginal cost curves and using numerical search algorithms to accurately locate the economically optimal governance point, a leap from blindly meeting standards to governance on demand has been achieved. This effectively avoids resource waste caused by over-governance and safety hazards caused by under-governance, and significantly improves the economic benefits of the power quality governance system.

[0195] In step S208 above, generating power quality management instructions based on the optimal management target includes: generating a start management instruction or an enhanced management instruction when the difference between the total harmonic distortion rate in the power quality data and the optimal management target is greater than a preset threshold, and the marginal benefit is greater than the product of the marginal cost and the first weight, wherein the first weight is the sum of 1 and the preset hysteresis coefficient; generating a stop management instruction or a weaken management instruction when the total harmonic distortion rate in the power quality data is less than or equal to the optimal management target, or the marginal benefit is less than the product of the marginal cost and the second weight, wherein the second weight is the difference between 1 and the preset hysteresis coefficient.

[0196] Specifically, the autonomous start / stop criteria in this application embodiment are as follows:

[0197] Triggering Criterion: If and If so, a start governance command or an enhanced governance command is generated to control the APF to start or increase its output.

[0198] Stop criterion: If or If so, a stop or reduce control command is generated to control the APF to stop or reduce its output.

[0199] in, A preset sensitivity threshold is used to prevent THD from affecting the sensitivity threshold. Minor fluctuations in the vicinity can cause the APF to frequently start and stop or significantly adjust its output. The hysteresis coefficient is used to start the system by expanding the range where the benefits exceed the costs, and to stop the system by narrowing the range where the benefits exceed the costs, thus forming a hysteresis loop and improving system stability.

[0200] In this embodiment of the application, the optimization decision engine sends power quality management instructions to the APF via a communication interface. The instruction types include:

[0201] Target THD directive: e.g., "target THD=4.2%";

[0202] Harmonic limit instructions: such as "3rd harmonic ≤ 3.5%, 5th harmonic ≤ 2.0%";

[0203] Start and stop commands: such as "start", "stop", "standby".

[0204] After receiving the command, the APF local controller adjusts the compensation current through internal closed-loop regulation to bring the THD close to the target value. Simultaneously, the system continuously monitors the actual THD; if the deviation exceeds 0.5% and remains there for 5 seconds... If so, the marginal analysis will be retried for fine-tuning.

[0205] In this embodiment, a multi-factor dynamic response mechanism is also introduced, including: acquiring the grid electricity price signal and updating the optimal governance target based on the grid electricity price signal, wherein the grid electricity price signal is used to represent the change signal corresponding to the electricity price; updating the efficiency coefficient and aging acceleration factor corresponding to the governance equipment according to the operating status of the governance equipment to obtain the target efficiency coefficient and target aging acceleration factor, and updating the optimal governance target based on the target efficiency coefficient and target aging acceleration factor; and correcting the model weights of the benefit model and cost model and updating the optimal governance target when the rate of change of total harmonic distortion in the power quality data exceeds a preset mutation threshold. The specific analysis is as follows:

[0206] 1. Electricity Price Response:

[0207] Real-time electricity price in the marginal revenue formula Updated in real time according to peak and off-peak electricity prices.

[0208] The system can receive electricity price signals or adjust according to a preset time schedule. For example, during peak electricity prices... Increase Automatic reduction (due to strengthened governance); conversely, reduction during off-peak electricity pricing.

[0209] 2. Equipment aging response:

[0210] The marginal cost of an APF increases with operating time (inefficiency decreases, depreciation accelerates).

[0211] The system can adjust the monthly update based on the APF's operating status (such as cumulative operating hours and number of failures). Efficiency coefficient in and aging acceleration factors And adjust accordingly .

[0212] 3. Response to sudden changes in operating conditions:

[0213] When a step change (such as rate of change) is detected in the load power or THD. Upon this, a complete fingerprint recognition and model calibration is immediately triggered, and the calculation is recalculated. .

[0214] This multi-factor dynamic response mechanism dynamically adjusts the marginal revenue and cost model parameters by integrating electricity price signals, equipment health status, and changes in grid operating conditions in real time. This transforms the governance strategy from static and fixed to dynamic and adaptive, ensuring that the economically optimal solution of "marginal revenue = marginal cost" can be achieved under different economic environments and equipment conditions. It also effectively adapts to sudden disturbances in the grid, improving the robustness of the system and the overall operational efficiency.

[0215] In this embodiment of the application, investment advice reports can also be generated offline using historical data from long-term monitoring.

[0216] Calculate the payback period for installing an APF in the current scenario. This transforms the abstract concept of power quality management into a concrete financial indicator (investment payback period), where the numerator is the initial investment cost and the denominator is the net annual return after deducting maintenance costs, thus providing a direct assessment of the economic feasibility of installing APF. The specific expression is as follows:

[0217]

[0218] in, For the expected annual return, Annual maintenance cost.

[0219] Meanwhile, competing solutions (on-load tap changer, DVR, dual-circuit power supply) are introduced for economic comparison. By comparing the input-output ratio of each solution, the technical path with the best cost performance in a specific scenario is identified.

[0220] Based on the above calculations and comparisons, the system automatically generates ranking suggestions to help users select the most economically efficient solution from numerous governance options, thus achieving a closed-loop decision-making process from technical compliance to economic optimization.

[0221] Overall, in this embodiment, multi-dimensional harmonic fingerprinting technology is used to accurately distinguish harmonic source types to match differentiated benefit models. A real-time quantification mechanism for marginal revenue and marginal cost is introduced. The optimal stopping theory is used to dynamically solve the economically optimal governance objective. Combined with a dynamic response mechanism for multiple factors such as electricity price, equipment aging, and sudden changes in operating conditions, the intelligent leap from blindly operating all the equipment to precise governance on demand is realized. Under the premise of ensuring that power quality meets the standards, the economically optimal goals of energy saving, consumption reduction, equipment life extension, and return on investment are maximized.

[0222] According to embodiments of this application, a power quality management device is provided. It should be noted that the power quality management device of this application embodiment can be used to execute the power quality management method provided in this application embodiment. The power quality management device provided in the embodiments of this application is described below.

[0223] Figure 3This is a structural diagram of a power quality management device according to an embodiment of this application. Figure 3 As shown, the device includes:

[0224] The identification module 30 is used to acquire power quality data and identify the corresponding harmonic source type using a classification model;

[0225] The first determining module 32 is used to determine the marginal benefit corresponding to the power quality data based on the harmonic source type using a benefit model and to determine the marginal cost corresponding to the power quality data using a cost model. The marginal benefit includes the energy-saving benefit and equipment life extension benefit corresponding to power quality management, and the marginal cost includes the operating cost and depreciation cost of the management equipment.

[0226] The second determining module 34 is used to determine the optimal governance objective based on marginal revenue and marginal cost, wherein the optimal governance objective is used to represent the power quality index located at the equilibrium point of the benefit model and the cost model;

[0227] The generation module 36 is used to generate power quality management instructions based on the optimal management objective. The power quality management instructions are used to control the operating status of the management equipment to minimize the physical losses between the power grid and the management equipment.

[0228] By using the identification module, the first determination module, the second determination module, and the generation module in the aforementioned power quality management device, the goal of suppressing grid harmonic losses is achieved while avoiding increased energy consumption and lifespan reduction of the management equipment due to overcompensation. This realizes a dynamic balance between grid operating efficiency and the physical lifespan of the management equipment, thereby solving the technical problem of low management efficiency caused by the fixed control strategies typically used in related power quality management technologies.

[0229] In the power quality management device provided in this application embodiment, the identification module is further used to collect the three-phase voltage and three-phase current at the power grid monitoring point according to a preset acquisition frequency; and to determine the total harmonic distortion rate, harmonic content, harmonic centroid offset, harmonic spectrum entropy, total harmonic distortion rate fluctuation, and power consumption of the management equipment based on the three-phase voltage and three-phase current; and to determine power quality data based on the total harmonic distortion rate, harmonic content, harmonic centroid offset, harmonic spectrum entropy, total harmonic distortion rate fluctuation, and power consumption of the management equipment.

[0230] In the power quality management device provided in this application embodiment, the identification module is further used to determine the importance scores of total harmonic distortion (THD), harmonic content, harmonic centroid shift, harmonic spectral entropy, and THD fluctuation in power quality data. The importance score is used to represent the contribution of different dimensional features to distinguishing harmonic source types. Based on the importance score, a feature vector for harmonic source type identification is generated, and the feature vector is standardized. The standardized feature vector is analyzed using a random forest classification model to obtain the classification result corresponding to the power quality data. The category with the highest posterior probability in the classification result and exceeding a preset confidence threshold is determined as the harmonic source type corresponding to the power quality data.

[0231] In the power quality management device provided in this application embodiment, the first determining module is further used to obtain a first benefit model and a second benefit model corresponding to the type of harmonic source. The first benefit model is used to quantify the energy-saving benefits corresponding to power quality management, and the second benefit model is used to quantify the equipment life extension benefits corresponding to power quality management. The energy-saving benefits are obtained by analyzing the change in total harmonic distortion rate, the change rate of power loss caused by the change in unit total harmonic distortion rate, the annual equivalent operating hours, and the electricity price through the first benefit model. The equipment life extension benefits are obtained by analyzing the purchase cost, design life, life extension ratio, and annual equivalent operating hours of the protected equipment through the second benefit model. The life extension ratio is positively correlated with the life sensitivity coefficient of the protected equipment, and the life sensitivity coefficient is used to reflect the sensitivity of the protected equipment to harmonic thermal stress. The marginal benefits corresponding to the power quality data are determined based on the energy-saving benefits and the equipment life extension benefits.

[0232] In the power quality management device provided in this application embodiment, the first determining module is further used to obtain a first cost model and a second cost model corresponding to the type of harmonic source. The first cost model is used to quantify the operating cost of the management equipment, and the second cost model is used to quantify the depreciation cost of the management equipment. The first cost model is used to analyze the efficiency coefficient, the compensation capacity increment required for the change of unit total harmonic distortion rate, the annual equivalent operating hours, and the electricity price of the management equipment to obtain the operating cost of the management equipment. The second cost model is used to analyze the purchase cost, design life, aging acceleration factor, and annual equivalent operating hours of the management equipment to obtain the depreciation cost of the management equipment. The marginal cost corresponding to the power quality data is determined based on the operating cost and the depreciation cost.

[0233] In the power quality management device provided in this application embodiment, the second determining module is further configured to determine the marginal revenue curve corresponding to the marginal revenue and the marginal cost curve corresponding to the marginal cost based on the type of harmonic source; determine the minimum total harmonic distortion rate corresponding to the maximum output of the management equipment, take the minimum total harmonic distortion rate as the lower limit and the total harmonic distortion rate in the power quality data as the upper limit to obtain the target harmonic range; and solve the marginal revenue curve and the marginal cost curve within the target harmonic range to obtain the optimal management target.

[0234] In the power quality management device provided in this application embodiment, the second determining module is further configured to compare the marginal revenue curve and the marginal cost curve within the target harmonic interval to obtain a comparison result; if the comparison result indicates that there is an intersection point in the target harmonic interval where the marginal revenue equals the marginal cost, the total harmonic distortion rate corresponding to the intersection point is determined as the optimal management target; if the comparison result indicates that the marginal revenue within the target harmonic interval is greater than the marginal cost, the minimum total harmonic distortion rate is determined as the optimal management target; if the comparison result indicates that the marginal revenue within the target harmonic interval is less than the marginal cost, the total harmonic distortion rate in the power quality data is determined as the optimal management target.

[0235] In the power quality management device provided in this application embodiment, the generation module is further configured to generate a start management command or an enhanced management command when the difference between the total harmonic distortion rate in the power quality data and the optimal management target is greater than a preset threshold, and the marginal benefit is greater than the product of the marginal cost and the first weight, wherein the first weight is the sum of 1 and a preset hysteresis coefficient; and to generate a stop management command or a weaken management command when the total harmonic distortion rate in the power quality data is less than or equal to the optimal management target, or the marginal benefit is less than the product of the marginal cost and the second weight, wherein the second weight is the difference between 1 and the preset hysteresis coefficient.

[0236] The power quality management device provided in this application embodiment further includes an update module 38, which is used to acquire the grid electricity price signal and update the optimal management target based on the grid electricity price signal, wherein the grid electricity price signal is used to represent the change signal corresponding to the electricity price; and to update the efficiency coefficient and aging acceleration factor corresponding to the management equipment according to the operating status of the management equipment, to obtain the target efficiency coefficient and the target aging acceleration factor, and update the optimal management target based on the target efficiency coefficient and the target aging acceleration factor; and when the rate of change of the total harmonic distortion rate in the power quality data exceeds a preset mutation threshold, to correct the model weights of the benefit model and the cost model, and update the optimal management target.

[0237] This application also provides an electronic device, including: a memory and a processor, wherein the memory is used to store program instructions; and the processor is connected to the memory and used to execute the above-described power quality management method.

[0238] It should be noted that the aforementioned electronic equipment is used to perform Figure 2 The power quality management method shown above is also applicable to this electronic device, and will not be repeated here.

[0239] This application also provides a non-volatile storage medium including a stored computer program, wherein the device containing the non-volatile storage medium executes the above-described power quality management method by running the computer program.

[0240] It should be noted that the aforementioned non-volatile storage media is used for execution. Figure 2 The power quality management method shown above is also applicable to this non-volatile storage medium, and will not be repeated here.

[0241] This application also provides a computer program product, including computer instructions that, when executed by a processor, implement the above-described power quality management method.

[0242] It should be noted that the above-mentioned computer program product is used to execute Figure 2 The power quality management method shown above is also applicable to this computer program product, and will not be repeated here.

[0243] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0244] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0245] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0246] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0247] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0248] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0249] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for improving power quality, characterized in that, include: Acquire power quality data and use a classification model to identify the corresponding harmonic source types; Based on the harmonic source type, a benefit model is used to determine the marginal benefit corresponding to the power quality data, and a cost model is used to determine the marginal cost corresponding to the power quality data. The marginal benefit includes energy-saving benefits and equipment life extension benefits corresponding to power quality management, and the marginal cost includes the operating cost and depreciation cost of the management equipment. The optimal governance objective is determined based on the marginal benefit and the marginal cost, wherein the optimal governance objective is used to represent the power quality index located at the equilibrium point of the benefit model and the cost model; Power quality management instructions are generated based on the optimal management objective, wherein the power quality management instructions are used to control the operating state of the management equipment to minimize the physical losses between the power grid and the management equipment.

2. The method according to claim 1, characterized in that, Obtain power quality data, including: Based on the preset acquisition frequency, the three-phase voltage and three-phase current at the power grid monitoring points are collected respectively; Based on the three-phase voltage and the three-phase current, the total harmonic distortion rate, the content of each harmonic, the centroid offset of the harmonics, the harmonic spectrum entropy, the fluctuation of the total harmonic distortion rate, and the power consumption of the treatment equipment are determined respectively. The power quality data is determined based on the total harmonic distortion rate, the content of each harmonic, the centroid offset of the harmonics, the harmonic spectrum entropy, the fluctuation of the total harmonic distortion rate, and the power consumption of the treatment equipment.

3. The method according to claim 2, characterized in that, A classification model is used to identify the corresponding harmonic source types, including: The importance scores of the total harmonic distortion rate, the harmonic content, the harmonic centroid offset, the harmonic spectral entropy, and the total harmonic distortion rate fluctuation in the power quality data are determined, wherein the importance scores are used to represent the degree of contribution of different dimensional features to distinguishing harmonic source types. Based on the importance score, a feature vector is generated for harmonic source type identification, and the feature vector is then standardized. The standardized feature vectors are analyzed using a random forest classification model to obtain the classification results corresponding to the power quality data. The category with the highest posterior probability in the classification results that exceeds a preset confidence threshold is determined as the harmonic source type corresponding to the power quality data.

4. The method according to claim 1, characterized in that, The marginal revenue corresponding to the power quality data is determined using a benefit model, including: Obtain a first benefit model and a second benefit model corresponding to the type of harmonic source, wherein the first benefit model is used to quantify the energy-saving benefits corresponding to power quality management, and the second benefit model is used to quantify the equipment life extension benefits corresponding to power quality management; The energy-saving benefits are obtained by analyzing the change in total harmonic distortion rate, the change rate of power loss caused by the change in unit total harmonic distortion rate, the annual equivalent operating hours, and the electricity price through the first benefit model. The second benefit model is used to analyze the purchase cost, design life, life extension ratio, and annual equivalent operating hours of the protected equipment to obtain the equipment life extension benefit. The life extension ratio is positively correlated with the life sensitivity coefficient of the protected equipment, which reflects the sensitivity of the protected equipment to harmonic thermal stress. The marginal benefit corresponding to the power quality data is determined based on the energy-saving benefits and the equipment life extension benefits.

5. The method according to claim 4, characterized in that, The marginal cost corresponding to the power quality data is determined using a cost model, including: Obtain a first cost model and a second cost model corresponding to the type of harmonic source, wherein the first cost model is used to quantify the operating cost of the treatment equipment, and the second cost model is used to quantify the depreciation cost of the treatment equipment; The operating cost of the treatment equipment is obtained by analyzing the efficiency coefficient, the compensation capacity increment required for a unit change in total harmonic distortion, the annual equivalent operating hours, and the electricity price of the treatment equipment using the first cost model. The second cost model is used to analyze the purchase cost, design life, aging acceleration factor, and annual equivalent operating hours of the treatment equipment to obtain the depreciation cost of the treatment equipment. The marginal cost corresponding to the power quality data is determined based on the operating cost and the depreciation cost.

6. The method according to claim 1, characterized in that, Determining the optimal governance objective based on the marginal benefits and marginal costs includes: Based on the harmonic source type, determine the marginal revenue curve corresponding to the marginal revenue, and determine the marginal cost curve corresponding to the marginal cost; Determine the minimum total harmonic distortion rate corresponding to the maximum output of the treatment equipment, use the minimum total harmonic distortion rate as the lower limit, and use the total harmonic distortion rate in the power quality data as the upper limit to obtain the target harmonic range; The marginal revenue curve and the marginal cost curve are solved within the target harmonic range to obtain the optimal governance target.

7. The method according to claim 6, characterized in that, The method further includes: Within the target harmonic range, the marginal revenue curve and the marginal cost curve are compared to obtain a comparison result; If the comparison result indicates that there is an intersection point where marginal revenue equals marginal cost within the target harmonic interval, the total harmonic distortion rate corresponding to the intersection point is determined as the optimal governance target. If the comparison results indicate that the marginal benefits within the target harmonic interval are all greater than the marginal costs, the minimum total harmonic distortion rate is determined as the optimal governance target. If the comparison results indicate that the marginal benefit within the target harmonic interval is less than the marginal cost, the total harmonic distortion rate in the power quality data is determined as the optimal governance target.

8. The method according to claim 7, characterized in that, Based on the optimal governance objective, power quality governance instructions are generated, including: If the difference between the total harmonic distortion rate in the power quality data and the optimal governance target is greater than a preset threshold, and the marginal benefit is greater than the product of the marginal cost and the first weight, a governance start command or a governance enhancement command is generated, wherein the first weight is the sum of 1 and a preset hysteresis coefficient. If the total harmonic distortion rate in the power quality data is less than or equal to the optimal governance target, or if the marginal benefit is less than the product of the marginal cost and the second weight, a stop governance command or a weaken governance command is generated, wherein the second weight is the difference between 1 and the preset hysteresis coefficient.

9. The method according to claim 1, characterized in that, The method further includes: Acquire the grid electricity price signal and update the optimal governance objective based on the grid electricity price signal, wherein the grid electricity price signal is used to represent a change signal corresponding to the electricity price; and Based on the operating status of the treatment equipment, the efficiency coefficient and aging acceleration factor corresponding to the treatment equipment are updated respectively to obtain the target efficiency coefficient and the target aging acceleration factor, and the optimal treatment target is updated based on the target efficiency coefficient and the target aging acceleration factor; and If the rate of change of total harmonic distortion in the power quality data exceeds a preset mutation threshold, the model weights of the benefit model and the cost model are corrected, and the optimal governance objective is updated.

10. A power quality management device, characterized in that, include: The identification module is used to acquire power quality data and use a classification model to identify the corresponding harmonic source types; The first determining module is used to determine the marginal benefit corresponding to the power quality data based on the harmonic source type using a benefit model, and to determine the marginal cost corresponding to the power quality data using a cost model. The marginal benefit includes energy-saving benefits and equipment life extension benefits corresponding to power quality management, and the marginal cost includes the operating cost and depreciation cost of the management equipment. The second determining module is used to determine the optimal governance objective based on the marginal benefit and the marginal cost, wherein the optimal governance objective is used to represent the power quality index located at the equilibrium point of the benefit model and the cost model; The generation module is used to generate power quality management instructions based on the optimal management objective, wherein the power quality management instructions are used to control the operating state of the management equipment to minimize the physical losses between the power grid and the management equipment.

11. An electronic device, characterized in that, include: A memory and a processor, wherein the memory is used to store program instructions; The processor, connected to the memory, is used to execute the power quality management method according to any one of claims 1 to 9.

12. A non-volatile storage medium, characterized in that, The non-volatile storage medium includes a stored computer program, wherein the device containing the non-volatile storage medium executes the power quality management method according to any one of claims 1 to 9 by running the computer program.

13. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the power quality management method according to any one of claims 1 to 9.

Images