High-frequency harmonic suppression method, device and equipment of vacuum circuit breaker and storage medium

Abstract

The invention provides a high-frequency harmonic suppression method, device and equipment for vacuum circuit breakers and a storage medium, each vacuum circuit breaker is provided with a high-frequency harmonic damper, and the method comprises the following steps: collecting power grid voltage and current signals of each vacuum circuit breaker in a preset area, and determining a harmonic state vector according to the power grid voltage and current signals; based on a preset multi-agent environment, calculating a first admittance value of a connected harmonic damper according to the harmonic state vector, and determining a damping action instruction according to the first admittance value; performing multi-round iteration on the damping action instruction according to a preset distributed collaborative architecture to obtain an execution instruction, the execution instruction comprising a second admittance value; modulating a driving signal of the high-frequency harmonic damper according to a second admittance value of the execution instruction so as to enable the high-frequency harmonic damper to output a high-frequency wave blocking signal; and inputting the high-frequency wave-blocking signal into the vacuum circuit breaker to reduce the high-frequency harmonic wave of the vacuum circuit breaker.

Description

Technical Field

[0001] This application relates to the field of high-frequency harmonic suppression technology, and in particular to a high-frequency harmonic suppression method, apparatus, equipment and storage medium for a vacuum circuit breaker. Background Technology

[0002] In current power distribution networks, vacuum circuit breakers, as critical switching devices, face increasingly prominent high-frequency harmonic pollution problems, especially in areas with dense power electronic loads, where they are highly susceptible to high-frequency harmonic resonance, threatening equipment insulation and system stability. Existing technologies often employ strategies such as centrally installing filters at the substation side or locally suppressing harmonics in individual vacuum circuit breakers. However, the former struggles to address the dynamic propagation and interaction of harmonics in distributed networks, while the latter, due to a lack of coordination, easily leads to each circuit breaker operating independently, resulting in limited suppression effects and potentially even new oscillation risks due to uncoordinated control actions. Therefore, effective and coordinated damping of high-frequency harmonic resonance in the regional power grid is impossible. Summary of the Invention

[0003] This application provides a method, apparatus, device, and storage medium for suppressing high-frequency harmonics in vacuum circuit breakers, which is used to organize vacuum circuit breakers scattered in a region into a unified harmonic suppression system to improve the suppression effect on high-frequency harmonics.

[0004] In a first aspect, embodiments of this application provide a method for suppressing high-frequency harmonics in a vacuum circuit breaker, wherein each vacuum circuit breaker is equipped with a high-frequency harmonic damper, and the method includes: Collect the grid voltage and current signals of each vacuum circuit breaker in the preset area, and determine the harmonic state vector based on the grid voltage and current signals; Based on a preset multi-agent environment, the first admittance value of the connected harmonic damper is calculated according to the harmonic state vector, and the damping action command is determined according to the first admittance value. The damping action command is iterated multiple times according to a preset distributed cooperative architecture to obtain an execution command, which includes a second admittance value. The drive signal of the high-frequency harmonic damper is modulated according to the second admittance value of the execution instruction, so that the high-frequency harmonic damper outputs a high-frequency blocking signal; The high-frequency blocking signal is input into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker.

[0005] Secondly, embodiments of this application provide a high-frequency harmonic suppression device for a vacuum circuit breaker, wherein each vacuum circuit breaker is equipped with a high-frequency harmonic damper, and the high-frequency harmonic suppression device for the vacuum circuit breaker includes: The harmonic analysis module is used to collect the grid voltage and current signals of each vacuum circuit breaker in a preset area, and determine the harmonic state vector based on the grid voltage and current signals; The instruction generation module is used to calculate the first admittance value of the connected harmonic damper based on the harmonic state vector according to the preset multi-agent environment, and determine the damping action instruction based on the first admittance value. The instruction iteration module is used to perform multiple iterations on the damping action instruction according to a preset distributed cooperative architecture to obtain an execution instruction, wherein the execution instruction includes a second admittance value. The instruction execution module is used to modulate the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution instruction, so that the high-frequency harmonic damper outputs a high-frequency blocking signal; The harmonic suppression module is used to input the high-frequency blocking signal into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker.

[0006] Thirdly, embodiments of this application provide an electronic device, which includes a memory and a processor; The memory is used to store computer programs; The processor is configured to execute the computer program and, in executing the computer program, implement the high-frequency harmonic suppression method for the vacuum circuit breaker as described in any of the embodiments of this application.

[0007] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to implement a high-frequency harmonic suppression method for a vacuum circuit breaker as described in any of the embodiments of this application.

[0008] This application provides a method, apparatus, device, and storage medium for suppressing high-frequency harmonics in a vacuum circuit breaker. Each vacuum circuit breaker is equipped with a high-frequency harmonic damper. The method includes: acquiring the grid voltage and current signals of each vacuum circuit breaker within a preset area, and determining a harmonic state vector based on the grid voltage and current signals; calculating a first admittance value of the connected harmonic damper based on the harmonic state vector according to a preset multi-agent environment, and determining a damping action command based on the first admittance value; performing multiple iterations on the damping action command according to a preset distributed cooperative architecture to obtain an execution command, the execution command including a second admittance value; modulating the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution command, so that the high-frequency harmonic damper outputs a high-frequency blocking signal; and inputting the high-frequency blocking signal into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker. In the above method, the real-time interaction of the harmonic state vector of the vacuum circuit breaker is used to achieve global perception and collaborative decision-making of the regional harmonic state. By converting the coordinated admittance command into a precise high-frequency blocking signal, the coordinated, precise and adaptive damping of high-frequency harmonic resonance in the spatiotemporal dimension effectively improves the power quality and operational stability of the regional power grid. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A schematic flowchart illustrating a high-frequency harmonic suppression method for a vacuum circuit breaker provided in this application embodiment; Figure 2 This is a schematic block diagram of a high-frequency harmonic suppression device for a vacuum circuit breaker, provided as an embodiment of this application. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described below with reference to the accompanying drawings.

[0012] The terms "first" and "second," etc., used in the specification, claims, and drawings of this application are used to distinguish different objects, not to describe a specific order. 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 includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0013] The term "embodiment" as used herein means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0014] It should be understood that in this application, "at least one (item)" means one or more, "more than one" means two or more, "at least two (items)" means two or three or more, and "and / or" is used to describe the relationship between related objects, indicating that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the related objects before and after are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0015] Please see Figure 1 , Figure 1 This is a schematic flowchart illustrating a high-frequency harmonic suppression method for a vacuum circuit breaker provided in an embodiment of this application. Figure 1 As shown, the specific steps of the high-frequency harmonic suppression method of the vacuum circuit breaker include: S101-S105.

[0016] S101. Collect the grid voltage and current signals of each vacuum circuit breaker in the preset area, and determine the harmonic state vector based on the grid voltage and current signals.

[0017] For example, a vacuum circuit breaker can be considered as a node. Adjacent vacuum circuit breakers with direct electrical connections are called adjacent circuit breakers, and adjacent circuit breakers are also called adjacent nodes. High-speed data acquisition units deployed at each node synchronously capture the instantaneous values ​​of the three-phase power grid voltage and current signals at a sampling frequency of no less than 200kHz to capture high-frequency component information, forming a raw signal sequence. Time-frequency analysis processing based on a fast S-transform algorithm is applied to this raw signal sequence. This algorithm achieves precise positioning of specific harmonics within the 2kHz to 10kHz frequency band by dynamically adjusting the width of the Gaussian window function, thereby generating a time-frequency matrix that simultaneously reflects the harmonic time-domain distribution and frequency-domain characteristics. The amplitude and phase spectra of the target high-frequency harmonics, such as the 23rd, 25th, 47th, and 49th harmonics, are extracted from the time-frequency matrix. Simultaneously, combined with the small disturbance current signal injected at the vacuum circuit breaker side, the equivalent harmonic impedance parameters at that node are obtained by solving an impedance calculation model. The aforementioned multi-dimensional characteristic parameters, such as amplitude, phase, frequency, and harmonic impedance, are encapsulated with precise timestamp information and node geographic location identifiers to construct a standardized harmonic state vector. This harmonic state vector provides comprehensive and synchronous power grid state awareness data for subsequent collaborative decision-making processes.

[0018] S102. Based on the preset multi-agent environment, calculate the first admittance value of the connected harmonic damper according to the harmonic state vector, and determine the damping action command according to the first admittance value.

[0019] For example, in a pre-built multi-agent reinforcement learning environment, each vacuum circuit breaker equipped with a high-frequency harmonic damper is defined as an agent with autonomous decision-making capabilities. Its observation state space consists of its local harmonic state vector and key state information of neighboring nodes obtained through the communication network, while its action space is set as the range of admittance values ​​that the connected harmonic damper can achieve within its safe operating range. In the multi-agent environment, a node with high computing power can be selected as the central decision-making unit to make the final decision on generating damping action commands. By adopting the Safe-MADDPG algorithm framework, a graph attention network mechanism is introduced at the input of the critic network of each agent, enabling it to assign differentiated attention weights to the state information of different neighboring nodes according to the power grid topology, thereby more accurately evaluating the long-term value generated by performing a certain admittance adjustment action under a specific power grid condition. At the same time, a safety constraint module is integrated into the output layer of the actor network. This module maps the admittance proposal value of the original output of the neural network to a feasible safety domain through a set of inequality equations describing the physical limits of the equipment and the stable operating boundary of the power grid, forming the first admittance value of the harmonic damper. Using this first admittance value as the core parameter, and combined with the collaborative control strategy that has converged after multiple iterations of training by the near-end strategy optimization algorithm, a damping action command is generated. The damping action command needs to take into account both the total harmonic distortion rate in the minimization region and the equipment operating efficiency.

[0020] S103. The damping action command is iterated multiple times according to the preset distributed cooperative architecture to obtain the execution command, which includes the second admittance value.

[0021] For example, the damping action command set generated in the aforementioned steps is transmitted to each vacuum circuit breaker via a fiber optic communication ring network that communicates according to a preset communication protocol, such as IEC61850. This communication network provides a highly reliable and low-latency data exchange channel. After receiving the damping action command issued by the central decision-making unit, each node uses it as the reference value for its local command, and simultaneously exchanges its initial command value and locally measured equivalent harmonic impedance parameters with its directly electrically connected neighboring nodes. Based on the exchanged network state information, each node uses a distributed weighted consensus algorithm to calculate the electrical coupling strength coefficient between itself and each neighboring node, and assigns corresponding weight factors to the commands of neighboring nodes according to this coefficient. Through multiple rounds of iterative local negotiation, each node continuously performs weighted average calculation and dynamic correction on its local command estimate, so that the admittance adjustment commands of all nodes gradually become consistent and smoothly transition. When the difference between the command estimates of all nodes is less than a preset convergence threshold, a set of mutually coordinated and globally optimized execution commands is generated, which includes a second admittance value after collaborative optimization.

[0022] S104. Modulate the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution instruction, so that the high-frequency harmonic damper outputs a high-frequency blocking signal.

[0023] For example, based on the second admittance value specified by the execution command, the reference trajectory of the output current and the reference trajectory of the internal capacitor voltage required for the high-frequency harmonic damper to achieve the target admittance characteristic are derived through a real-time calculation model. The output current and capacitor voltage values ​​monitored by sensors during the actual operation of the high-frequency harmonic damper are continuously compared with the aforementioned reference trajectories to obtain the state tracking error signal. An adaptive sliding mode control strategy is adopted to handle this state tracking error. By designing an integral sliding mode surface and constructing a composite control law containing equivalent control and switching control based on Lyapunov stability theory, this control law can effectively overcome the adverse effects of grid parameter fluctuations on the control effect. The synthesized control signal is input to the space vector pulse width modulation module to generate a set of high-precision pulse width modulation drive signals with specific duty cycles and phases. After power amplification, these drive signals directly control the switching on and off of the fully controlled power electronic devices within the high-frequency harmonic damper, causing it to output a high-frequency blocking signal with the same amplitude but opposite phase as the original high-frequency harmonics in the grid.

[0024] S105. Input the high-frequency blocking signal into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker.

[0025] For example, a high-frequency blocking signal generated by a high-frequency harmonic damper is injected into the power grid branch where the vacuum circuit breaker is located via a magnetic coupling interface or direct electrical connection. This high-frequency blocking signal creates a specific voltage drop across the power grid impedance, thereby altering the harmonic voltage distribution at the vacuum circuit breaker terminals. The injected high-frequency blocking signal and the naturally occurring high-frequency harmonic signals in the power grid exhibit waveform superposition at the vacuum circuit breaker terminals. Because the two signals are similar in amplitude and maintain an approximately 180-degree anti-phase relationship, they significantly weaken the amplitude of the high-frequency harmonic current flowing through the vacuum circuit breaker's contact points through destructive interference. This process alters the equivalent harmonic impedance characteristics of the local power grid where the vacuum circuit breaker is located, particularly producing a significant damping effect on harmonic resonance phenomena in the 1500Hz to 5000Hz frequency band. This reduces the target high-frequency harmonic content at the vacuum circuit breaker terminals to below the limits specified by international standards, improving the operating environment of the vacuum circuit breaker and enhancing its breaking performance and service life.

[0026] This application provides a method, apparatus, device, and storage medium for suppressing high-frequency harmonics in a vacuum circuit breaker. Each vacuum circuit breaker is equipped with a high-frequency harmonic damper. The method includes: acquiring the grid voltage and current signals of each vacuum circuit breaker within a preset area, and determining a harmonic state vector based on the grid voltage and current signals; calculating a first admittance value of the connected harmonic damper based on the harmonic state vector according to a preset multi-agent environment, and determining a damping action command based on the first admittance value; performing multiple iterations on the damping action command according to a preset distributed cooperative architecture to obtain an execution command, the execution command including a second admittance value; modulating the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution command, so that the high-frequency harmonic damper outputs a high-frequency blocking signal; and inputting the high-frequency blocking signal into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker. In the above method, the real-time interaction of the harmonic state vector of the vacuum circuit breaker is used to achieve global perception and collaborative decision-making of the regional harmonic state. By converting the coordinated admittance command into a precise high-frequency blocking signal, the coordinated, precise and adaptive damping of high-frequency harmonic resonance in the spatiotemporal dimension effectively improves the power quality and operational stability of the regional power grid.

[0027] To more clearly illustrate the technical solution of this application, the technical solution of this application will be described below through specific embodiments. It should be noted that the specific embodiments are used to expand the description of the technical solution of this application, and are not intended to limit this application.

[0028] In some embodiments, acquiring the grid voltage and current signals of each vacuum circuit breaker within a preset area, and determining the harmonic state vector based on the grid voltage and current signals, includes: synchronously sampling the three-phase voltage and current signals of each node within the preset area according to a preset high-frequency sampling frequency to obtain a high-frequency sampling sequence; applying a preset fast S-transform algorithm to calculate the time-frequency matrix of the high-frequency sampling sequence within a preset frequency band; locating and extracting the amplitude spectrum and phase spectrum of the preset sub-high frequency harmonic from the time-frequency matrix; injecting a micro-changing current into the vacuum circuit breaker, and determining the equivalent harmonic impedance of the vacuum circuit breaker based on the amplitude spectrum, phase spectrum, and micro-changing current; and encapsulating the amplitude spectrum, phase spectrum, and equivalent harmonic impedance into a harmonic state vector.

[0029] For example, the process of acquiring the grid voltage and current signals of each vacuum circuit breaker within a preset area and determining the harmonic state vector based on these signals is achieved through precision measuring devices deployed at key nodes of the distribution network. These devices synchronously acquire the instantaneous values ​​of the three-phase grid voltage and current signals at a sampling frequency of no less than 200kHz, forming a high-frequency sampling sequence. This high-frequency sampling sequence is processed using a fast S-transform algorithm, which achieves accurate analysis of signals in different frequency bands by dynamically adjusting the width of the Gaussian window function, with a particular focus on analyzing harmonic components in the range of 2kHz to 10kHz, generating a time-frequency matrix that simultaneously reflects the signal's time and frequency domain characteristics. From this time-frequency matrix, the amplitude and phase spectrum information of specific high-frequency harmonics, such as the 23rd, 25th, 47th, and 49th harmonics, are accurately located and extracted. Simultaneously, combined with a micro-changing current signal with an amplitude not exceeding 5% of the rated current injected at the vacuum circuit breaker side, the equivalent harmonic impedance of the node is calculated based on the ratio of harmonic voltage change to current change. The extracted amplitude spectrum, phase spectrum, frequency parameters, and calculated equivalent harmonic impedance data are encapsulated together with precise timestamp information and node identifiers into a structured harmonic state vector. This harmonic state vector provides accurate and complete power grid harmonic state information for subsequent collaborative decision-making.

[0030] In some embodiments, based on a preset multi-agent algorithm, a first admittance value of the connected harmonic damper is calculated according to the harmonic state vector, and a damping action command is determined based on the first admittance value. This includes: constructing each vacuum circuit breaker as an agent based on a cooperative algorithm of Safe-MADDPG, embedding the agents into a critic network and an actor network, defining the state space of the agent as the harmonic state vector and the key state information of neighboring agents, and defining the action space as the range of admittance value variation of the harmonic damper; the agents are connected through the critic network; and setting a graph attention network layer before the input layer of the critic network, so that each agent, when evaluating the action value, ... It can assign different attention weights to the state information of neighboring agents based on electrical distance; a safety layer is set after the output layer of the actor network, which is constructed through an inequality constraint function to map the action output of the agent to the feasible region, which is bounded by the safe operating area of ​​the harmonic damper and the voltage stability boundary of the power grid; a reward function is designed with the core objective of minimizing the total harmonic distortion rate of the region, and the reward function also includes voltage deviation penalty of the vacuum circuit breaker and energy consumption penalty of the harmonic damper; the actor network and the critic network are alternately iteratively trained using a near-end policy optimization method until the policy converges and the damping action command is generated.

[0031] For example, based on a preset multi-agent environment, the first admittance value of the connected harmonic damper is calculated according to the harmonic state vector, and the implementation process of the damping action command is determined according to the first admittance value, thus constructing a complete distributed decision-making framework. The core of this framework is the Safe-MADDPG collaborative algorithm. In this algorithm architecture, each vacuum circuit breaker equipped with a high-frequency harmonic damper is constructed as an agent with autonomous learning and decision-making capabilities. These agents make collaborative decisions through a complex neural network structure. The state space of each agent is composed of its local harmonic state vector and key state information of neighboring agents obtained through the communication network. This key state information includes at least the harmonic amplitude, phase, and harmonic impedance parameters of neighboring nodes. The action space of each agent is strictly limited to the range of admittance value variation allowed by the physical device of the connected harmonic damper. This range is usually set between -0.01 and 0.01 Siemens to ensure safe operation of the equipment.

[0032] The graph attention network layer preceding the commentator network employs an attention allocation mechanism based on the power grid topology. This mechanism dynamically adjusts the weighting coefficients of the state information of neighboring agents by calculating the electrical distance between nodes. The calculation of electrical distance comprehensively considers multiple factors such as line impedance, transformer turns ratio, and network topology relationships, enabling each agent to prioritize the states of electrically connected neighboring nodes when evaluating the value of actions. This design significantly improves the accuracy and efficiency of collaborative decision-making. The output of the graph attention network layer is a weighted and fused global state feature vector, which accurately reflects the overall harmonic state of the regional power grid, providing high-quality input data for subsequent value assessment.

[0033] A safety layer, placed after the actor network's output layer, ensures that the agent's output actions always remain within the feasible region through a set of strict inequality constraint functions. These inequality constraint functions detail physical limitations such as the maximum current withstand capability of the harmonic damper, the maximum voltage operating range, and the grid voltage stability boundary. The current constraint is typically set to 120% of the device's rated current, while the voltage constraint refers to the harmonic voltage distortion rate limits specified in the IEEE 519 standard. The safety layer works by mapping the actor network's original output action suggestions to the feasible region defined by these constraints in real time. When the original output exceeds the feasible region, the safety layer projects it to the nearest feasible point. This mechanism effectively prevents control actions that could damage equipment or affect grid stability.

[0034] The designed reward function employs a multi-objective optimization structure, with minimizing the regional total harmonic distortion (THD) as the core objective, typically weighted at 0.6. The reward function also includes a penalty term for the vacuum circuit breaker node voltage deviation and a penalty term for the harmonic damper's own energy consumption, with weights set to 0.2 and 0.2 respectively. The voltage deviation penalty is calculated based on the squared deviation of the node voltage from the rated voltage, while the energy consumption penalty is determined by the product of the harmonic damper's output current and terminal voltage. This reward function design enables the trained control strategy to ensure harmonic suppression while also considering grid voltage stability and equipment operating economy.

[0035] The process of iteratively training the actor and critic networks using a proximal policy optimization method employs a distributed parallel computing architecture. This training process comprises three main stages: experience collection, gradient calculation, and parameter update. In the experience collection stage, each agent interacts with the environment to collect a large amount of state-action-reward data and stores it in a shared experience replay pool. In the gradient calculation stage, small batches of data are randomly sampled from the experience replay pool to calculate the parameter gradients of the actor and critic networks respectively. The actor network update uses an importance-weighted policy gradient, while the critic network update minimizes the temporal difference error. In the parameter update stage, the network parameters are iteratively optimized using the Adam optimizer, while a soft update strategy ensures the stability of the training process. When the value function estimate and actual cumulative reward of the policy network reach a stable state, the training process is considered converged, at which point a set of coordinated and effective damped action instructions can be generated.

[0036] In some embodiments, vacuum circuit breakers are distributed and coordinated through an optical fiber communication ring network, which communicates according to a preset communication protocol.

[0037] For example, a distributed collaborative communication architecture is constructed using a fiber optic communication ring network that communicates with vacuum circuit breakers according to a preset communication protocol, such as IEC 61850. This fiber optic communication ring network employs a bidirectional redundant topology to ensure uninterrupted communication even under any single-point failure, with transmission latency strictly controlled to within 4 milliseconds to meet the stringent real-time requirements of collaborative control. The communication protocol fully complies with the modeling specifications and service mapping mechanism of the IEC 61850 standard, abstracting physical devices such as vacuum circuit breakers and harmonic dampers into standardized objects conforming to a logical node model, enabling interoperability and seamless integration of different types of devices. Based on manufacturing message specifications and sampled value transmission services, the network can simultaneously transmit manufacturing message information such as harmonic state vector data and damping action commands, as well as real-time acquired current and voltage sampled value data, providing comprehensive data support for distributed collaborative control. Through multicast communication based on a publisher / subscriber mechanism, each node can efficiently exchange status information and control commands, creating the necessary technical conditions for subsequent distributed collaborative decision-making.

[0038] In some embodiments, the damping action command is iterated multiple times according to a preset distributed cooperative architecture to obtain an execution command. The execution command includes a second admittance value. The process includes: encapsulating the damping action command into a control message and broadcasting it to all vacuum circuit breakers through an optical fiber ring network, so that the vacuum circuit breakers reset their initial values ​​according to the corresponding command in the control message and exchange their initial values ​​and equivalent harmonic impedances with adjacent circuit breakers to calculate the electrical coupling strength. The adjacent circuit breakers are vacuum circuit breakers with direct electrical connections. Each vacuum circuit breaker generates a command weighting coefficient based on the electrical coupling strength, and the command weighting coefficient is assigned to the directly electrically connected vacuum circuit breakers. Each vacuum circuit breaker performs a fusion calculation on the damping action command of the vacuum circuit breaker and the damping action commands of the adjacent circuit breakers based on the weighting coefficient to generate a command estimate, until the difference between all command estimates is less than a preset threshold, and then outputs the execution command.

[0039] For example, the damping action command generated by the central decision-making unit is encapsulated into a control message according to the IEC 61850-90-5 protocol and broadcast to all relevant nodes in the region via a fiber optic communication ring network. Upon receiving the control message, each node parses the corresponding command value and sets it as the initial value of its local command. Simultaneously, it exchanges its initial command values ​​and locally stored equivalent harmonic impedance parameters with adjacent circuit breaker nodes that have direct electrical connections. Based on the exchanged network state information, each node uses an admittance-weighted average algorithm to calculate the electrical coupling strength between itself and its adjacent nodes. This strength value reflects the degree of harmonic mutual influence between nodes. According to the calculated electrical coupling strength value, a corresponding command weight coefficient is assigned to the command of each adjacent node; the node with the greater coupling strength has a correspondingly larger command weight coefficient. Based on the assigned weight coefficients, each node performs a weighted fusion calculation on its local damping action command and the received damping action commands from adjacent circuit breakers to generate a new round of command estimates. Through multiple rounds of iterative local negotiation, the instruction estimates of each node gradually become consistent. When the maximum difference between the instruction estimates of all nodes is less than the preset Siemens convergence threshold of 0.01, the negotiation process terminates and outputs a set of coordinated execution instructions, which includes the second admittance value after distributed optimization.

[0040] In some embodiments, the calculation of the time-frequency matrix of a high-frequency sampling sequence within a preset frequency band using a preset fast S-transform algorithm includes: performing Gaussian window processing on the high-frequency sampling sequence based on multiple preset analysis frequency points to obtain a first time-frequency signal; performing fast Fourier transform and frequency domain convolution on the first time-frequency signal to obtain a frequency domain filtered signal; performing inverse fast Fourier transform on the frequency domain filtered signal to obtain a second time-frequency signal for each preset analysis frequency point; and determining the time-frequency matrix based on the second time-frequency signal.

[0041] For example, an adaptive Gaussian window is applied to the input high-frequency sampling sequence based on multiple preset analysis frequencies. The width of the Gaussian window function is inversely proportional to the frequency value of the current analysis frequency, allowing a wider time window to achieve higher frequency resolution when analyzing low-frequency components and a narrower time window to achieve more accurate time positioning when analyzing high-frequency components. After windowing, a first time-frequency signal is obtained. A Fast Fourier Transform (FFT) is performed on the first time-frequency signal to convert it from the time domain to the frequency domain. Then, a convolution operation is performed in the frequency domain with a kernel function designed according to the characteristics of the analysis frequency, achieving frequency domain filtering for a specific frequency band, resulting in a frequency-domain filtered signal. An Inverse Fast Fourier Transform (IFFT) is then performed on the frequency-domain filtered signal to convert it back to the time domain, obtaining a second time-frequency signal corresponding to each preset analysis frequency. This signal reflects the time-varying characteristics of the original signal at each specific frequency. All these second time-frequency signals are arranged and combined in ascending order of frequency into a two-dimensional complex matrix, i.e., the required time-frequency matrix. This time-frequency matrix provides a comprehensive time-frequency analysis foundation for subsequent harmonic feature extraction.

[0042] In some embodiments, modulating the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution instruction includes: calculating the output current reference value and the capacitor voltage reference value of the damper according to the second admittance value to obtain a state reference trajectory; comparing the actual output state quantity of the high-frequency harmonic damper with the state reference trajectory to determine the state tracking error; and generating a pulse width modulation signal according to the state tracking error as the drive signal.

[0043] For example, based on the second admittance value included in the execution instruction, the output current reference value and capacitor voltage reference value required for the high-frequency harmonic damper to achieve the target admittance characteristic are calculated through a preset state-space equation. These two reference values ​​together constitute the state reference trajectory of the control system. The actual output current value and capacitor voltage value of the high-frequency harmonic damper are monitored in real time by high-precision sensors. These actual output state quantities are continuously compared with the calculated state reference trajectory to obtain a precise state tracking error signal. Based on this state tracking error signal, a corresponding pulse width modulation signal is generated using sliding mode variable structure control theory. The parameters of the sliding mode controller are adjusted in real time according to the grid operating state to ensure good control performance under different operating conditions. The pulse width modulation signal serves as a drive signal to directly control the switching action of the fully controlled power electronic devices within the high-frequency harmonic damper, thereby achieving precise control of the output impedance of the high-frequency harmonic damper.

[0044] Please see Figure 2 , Figure 2This is a schematic block diagram of a high-frequency harmonic suppression device for a vacuum circuit breaker, provided in an embodiment of this application. The high-frequency harmonic suppression device 200 is used to perform the aforementioned high-frequency harmonic suppression method for a vacuum circuit breaker. The high-frequency harmonic suppression device 200 can be configured in a server.

[0045] The server can be a standalone server, a server cluster, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0046] like Figure 2 As shown, the high-frequency harmonic suppression device 200 of the vacuum circuit breaker includes: a harmonic analysis module 201, an instruction generation module 202, an instruction iteration module 203, an instruction execution module 204, and a harmonic suppression module 205.

[0047] The harmonic analysis module 201 is used to collect the grid voltage and current signals of each vacuum circuit breaker in the preset area and determine the harmonic state vector based on the grid voltage and current signals.

[0048] The instruction generation module 202 is used to calculate the first admittance value of the connected harmonic damper based on the harmonic state vector according to the preset multi-agent environment, and determine the damping action instruction based on the first admittance value.

[0049] The instruction iteration module 203 is used to perform multiple iterations on the damping action instruction according to the preset distributed cooperative architecture to obtain the execution instruction, which includes the second admittance value.

[0050] The instruction execution module 204 is used to modulate the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution instruction, so that the high-frequency harmonic damper outputs a high-frequency blocking signal.

[0051] The harmonic suppression module 205 is used to input high-frequency blocking signals into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker.

[0052] This application provides an electronic device, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and, when executing the computer program, implement a high-frequency harmonic suppression method for a vacuum circuit breaker as described in any of the embodiments of this application.

[0053] This application provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program causes the processor to implement a high-frequency harmonic suppression method for a vacuum circuit breaker as described in any of the embodiments of this application.

[0054] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for suppressing high-frequency harmonics in a vacuum circuit breaker, characterized in that, Each vacuum circuit breaker is equipped with a high-frequency harmonic damper, and the method includes: Collect the grid voltage and current signals of each vacuum circuit breaker in the preset area, and determine the harmonic state vector based on the grid voltage and current signals; Based on a preset multi-agent environment, the first admittance value of the connected harmonic damper is calculated according to the harmonic state vector, and the damping action command is determined according to the first admittance value. The damping action command is iterated multiple times according to a preset distributed cooperative architecture to obtain an execution command, which includes a second admittance value. The drive signal of the high-frequency harmonic damper is modulated according to the second admittance value of the execution instruction, so that the high-frequency harmonic damper outputs a high-frequency blocking signal; The high-frequency blocking signal is input into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker.

2. The method as described in claim 1, characterized in that, The process of collecting grid voltage and current signals from each vacuum circuit breaker within a preset area, and determining harmonic state vectors based on the grid voltage and current signals, includes: The three-phase voltage and current signals of each node in the preset area are synchronously sampled according to the preset high-frequency sampling frequency to obtain a high-frequency sampling sequence; The time-frequency matrix of the high-frequency sampling sequence within a preset frequency band is calculated using a preset fast S-transform algorithm. From the time-frequency matrix, locate and extract the amplitude spectrum and phase spectrum of the preset sub-high frequency harmonic; A micro-current is injected into the vacuum circuit breaker, and the equivalent harmonic impedance of the vacuum circuit breaker is determined based on the amplitude spectrum, the phase spectrum, and the micro-current. The amplitude spectrum, phase spectrum, and equivalent harmonic impedance are encapsulated as a harmonic state vector.

3. The method as described in claim 1, characterized in that, The multi-agent system based on a preset algorithm calculates the first admittance value of the connected harmonic damper according to the harmonic state vector, and determines the damping action command according to the first admittance value, including: Based on the Safe-MADDPG collaborative algorithm, each vacuum circuit breaker is constructed as an agent, and the agent is embedded in the critic network and the actor network. The state space of the agent is defined as the harmonic state vector and the key state information of the neighboring agents, and the action space is defined as the range of the admittance value of the harmonic damper. The agents are connected through the critic network. A graph attention network layer is set before the input layer of the critic network, so that each agent can assign different attention weights to the state information of neighboring agents based on electrical distance when evaluating the value of an action. A safety layer is set after the output layer of the actor network. The safety layer is constructed by inequality constraint functions and is used to map the action output of the agent to a feasible region. The feasible region is enclosed by the safe operating area of ​​the harmonic damper and the grid voltage stability boundary. The reward function is designed with the core objective of minimizing the total harmonic distortion rate of the region. The reward function also includes: voltage deviation penalty of the vacuum circuit breaker and energy consumption penalty of the harmonic damper. Using a proximal policy optimization method, the actor network and the critic network are trained alternately and iteratively until the policy converges, generating damped action instructions.

4. The method as described in claim 2, characterized in that, The vacuum circuit breakers are distributed and coordinated through an optical fiber communication ring network, which communicates according to a preset communication protocol.

5. The method as described in claim 4, characterized in that, The damping action command is iterated multiple times according to a preset distributed cooperative architecture to obtain an execution command, the execution command including a second admittance value, including: The damping action command is encapsulated into a control message and broadcast to all the vacuum circuit breakers through the optical fiber ring network, so that the vacuum circuit breakers reset their initial values ​​according to the corresponding command in the control message and exchange the initial values ​​and the equivalent harmonic impedance with the adjacent circuit breakers to calculate the electrical coupling strength. The adjacent circuit breakers are vacuum circuit breakers with direct electrical connection. Each of the vacuum circuit breakers generates a command weighting coefficient based on the electrical coupling strength, and the command weighting coefficient is assigned to the vacuum circuit breakers that are directly electrically connected. Based on the weighting coefficient, each vacuum circuit breaker performs a fusion calculation on the damping action command of the vacuum circuit breaker and the damping action command of the adjacent circuit breaker to generate a command estimate. The execution command is then output when the difference between all the command estimates is less than a preset threshold.

6. The method as described in claim 2, characterized in that, The application of a preset fast S-transform algorithm calculates the time-frequency matrix of the high-frequency sampling sequence within a preset frequency band, including: Based on multiple preset analysis frequency points, the high-frequency sampling sequence is processed by a Gaussian window to obtain a first time-frequency signal; The first time-frequency signal is subjected to a fast Fourier transform and frequency domain convolution to obtain a frequency domain filtered signal; The frequency domain filtered signal is subjected to inverse fast Fourier transform to obtain the second time-frequency signal for each preset analysis frequency point, and the time-frequency matrix is ​​determined based on the second time-frequency signal.

7. The method as described in claim 1, characterized in that, The modulation of the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution instruction includes: Based on the second admittance value, the output current reference value and capacitor voltage reference value of the damper are calculated to obtain the state reference trajectory; The actual output state quantity of the high-frequency harmonic damper is compared with the state reference trajectory to determine the state tracking error. A pulse width modulation signal is generated based on the state tracking error and used as a driving signal.

8. A high-frequency harmonic suppression device for a vacuum circuit breaker, characterized in that, The high-frequency harmonic suppression device of the vacuum circuit breaker is used to perform the high-frequency harmonic suppression method of the vacuum circuit breaker as described in any one of claims 1-7, and the high-frequency harmonic suppression device of the vacuum circuit breaker includes: The harmonic analysis module is used to collect the grid voltage and current signals of each vacuum circuit breaker in a preset area, and determine the harmonic state vector based on the grid voltage and current signals; The instruction generation module is used to calculate the first admittance value of the connected harmonic damper based on the harmonic state vector according to the preset multi-agent environment, and determine the damping action instruction based on the first admittance value. The instruction iteration module is used to perform multiple iterations on the damping action instruction according to a preset distributed cooperative architecture to obtain an execution instruction, wherein the execution instruction includes a second admittance value. The instruction execution module is used to modulate the drive signal of the high-frequency harmonic damper according to the second admittance value of the execution instruction, so that the high-frequency harmonic damper outputs a high-frequency blocking signal; The harmonic suppression module is used to input the high-frequency blocking signal into the vacuum circuit breaker to reduce the high-frequency harmonics of the vacuum circuit breaker.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor; The memory is used to store computer programs; The processor is configured to execute the computer program and, in executing the computer program, implement the high-frequency harmonic suppression method for the vacuum circuit breaker as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, causes the processor to implement the high-frequency harmonic suppression method for a vacuum circuit breaker as described in any one of claims 1 to 7.

Images