Adaptive power grid low-frequency load shedding method and device based on three-level cooperation for transmission and distribution

Abstract

The application discloses a kind of based on three-stage cooperation of transmission and distribution self-adapting power grid low-frequency load shedding method and device, the method is by constructing wide-area measurement network when detecting low-frequency overrun, collection synchronous operation data, constructs multi-machine equivalent model, calculates initial total power shortage;Draw low-frequency response curve, calculate real-time frequency change rate, according to new energy access proportion judgment threshold interval, correct or directly adopt initial shortage value;According to accurate shortage value, by equal distribution algorithm is decomposed into distribution network partition load shedding and is issued, distribution network is again decomposed to district internet of things terminal;Terminal screening cuts off the non-important load of small frequency regulation effect, feedback actual cut-off amount;Transmission network dispatching center accumulates cumulative load shedding, judges whether frequency recovers safety threshold, recovers and terminates process, does not recover and does not reach maximum round then iteration estimation correction, until up to standard or complete full round cut-off.The application solves the problems of low power shortage estimation accuracy, inaccurate load shedding and other problems in the prior art.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, specifically to an adaptive power grid low-frequency load shedding method and device based on three-level coordination of transmission, distribution and application. Background Technology

[0002] In the field of power system frequency stability control, Underfrequency Load Shedding (UFLS) has been widely used as the last line of defense to ensure system frequency security. Traditional UFLS strategies mostly adopt a round-based load shedding scheme based on fixed frequency thresholds. By pre-setting multiple load shedding thresholds and fixed load shedding ratios, loads are gradually cut off when the system frequency exceeds the limit. With the large-scale integration of high-proportion renewable energy and distributed power sources, some power grids have begun to try to introduce adaptive load shedding methods based on the rate of change of frequency (ROCOF). These methods combine single-unit equivalent (SFR) models to estimate power deficits and allocate load shedding amounts through offline-tuned load shedding strategy tables to improve load shedding response speed. At the same time, transmission and distribution network coordinated control technology is gradually developing, and some pilot projects have achieved considerable and controllable load resources on the distribution network side, providing an engineering foundation for the transition of UFLS from "centralized control of the transmission network" to "three-level coordination of transmission, distribution, and consumption".

[0003] Currently, traditional round-robin low-frequency load shedding strategies rely on fixed parameters set offline, failing to consider real-time grid operation and changes in source-load characteristics. This can easily lead to insufficient load shedding or excessive shedding in scenarios with a high proportion of renewable energy integration. ROCOF-based adaptive methods often employ single-unit equivalent models, neglecting the impact of differences in unit dynamic response, uneven load distribution, and grid topology changes. This results in insufficient accuracy in power deficit estimation, making it difficult to adapt to the actual needs of complex grids. Furthermore, existing load shedding schemes often focus on centralized decision-making on the transmission network side, lacking deep coordination with the distribution network and user substation load characteristics. This makes it difficult to accurately screen non-critical loads with weak frequency regulation effects, easily leading to erroneous shedding of critical loads and decreased power supply reliability for users. Simultaneously, load shedding initiation criteria rely solely on real-time frequency limit exceedances, lacking prediction of frequency recovery effects. The forward-looking nature and closed-loop optimization capabilities of the load shedding strategy are insufficient, making it difficult to meet the accuracy and reliability requirements of frequency stability control for grids with a high proportion of renewable energy.

[0004] Therefore, there is an urgent need for an adaptive low-frequency load shedding method for power grids based on three-level coordination of transmission, distribution and utilization, to solve the problems of low power deficit estimation accuracy, inaccurate load shedding and over / under-shearing in existing technologies. Summary of the Invention

[0005] To address these issues, this invention provides an adaptive power grid low-frequency load shedding method and apparatus based on three-level coordination of transmission and distribution, which solves the problems of low power deficit estimation accuracy caused by reliance on single-machine equivalent models in traditional low-frequency load shedding, inaccurate load shedding and over / under-shearing due to insufficient transmission and distribution coordination, and lack of frequency recovery effect prediction and closed-loop adaptive control.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an adaptive power grid low-frequency load shedding method based on three-level coordination of transmission and distribution, characterized in that it includes: By deploying synchronous phasor measurement units and wide-area measurement terminals at key nodes of the power transmission network, a wide-area measurement network covering the entire power transmission network is constructed. When a low-frequency over-limit is detected in the power system, synchronous operation data of key nodes of the entire power transmission network is collected through the wide-area measurement network. Based on the synchronous operation data, a multi-machine equivalent model is constructed. Based on the multi-machine equivalent model, the system power balance equation, the rotor motion equation, and the system frequency response model considering load regulation effects, the initial total power deficit of the power system is calculated. Based on the synchronous operation data, the system's low-frequency response curve is plotted in real time; the real-time frequency change rate of the low-frequency response curve is calculated using a numerical differentiation algorithm; a threshold range for the frequency change rate is set based on the proportion of new energy access; it is determined whether the real-time frequency change rate exceeds the threshold range; if it does, a power deficit correction mechanism is activated to dynamically correct the initial total power deficit and obtain an accurate power deficit value; if it does not exceed the threshold range, the initial total power deficit is used as the accurate power deficit value. Based on the accurate power deficit value, combined with the load scale, power supply reliability level and grid security constraints of each distribution network, the total load reduction task is decomposed into the load reduction amount of each distribution network and load reduction instructions are generated through the load balancing allocation algorithm; the load reduction instructions are sent to the dispatch centers of each distribution network through the transmission and distribution network collaborative communication link; based on the load characteristics of each transformer substation in the jurisdiction, the distribution network dispatch center decomposes the load reduction amount of each region into specific load reduction instructions for each transformer substation and sends them to the IoT control terminal of the corresponding substation. Based on load identification algorithms and load importance levels, IoT control terminals screen and disconnect non-critical loads with small frequency regulation effects; the actual load disconnection amount is collected and fed back to the distribution network dispatch center and transmission network dispatch center through IoT communication links; The power grid dispatch center summarizes the actual load shedding amount of each distribution area to obtain the cumulative load reduction of the system; based on the real-time frequency data of the system in the synchronous operation data, it determines whether the system frequency has recovered to the safe threshold; if it has recovered, the load shedding process is terminated; if it is lower than the safe threshold and the maximum load shedding round has not been reached, the power deficit estimation and correction process is iteratively performed based on the slope of the current system low-frequency response curve and the remaining power deficit, until the system frequency recovers to the safe range or the full load shedding round is completed.

[0007] As a preferred embodiment of the adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution and utilization, the key nodes of the transmission network include: power plant busbars, regional tie lines and load center busbars; the synchronous operation data include: voltage, current, frequency, phase angle, unit output and total load.

[0008] As a preferred scheme for an adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization, in the process of calculating the initial total power deficit of the power system, the source-load core characteristic parameters are calculated through the system frequency response model that takes into account the load regulation effect; the source-load core characteristic parameters are integrated into the multi-machine equivalent model to obtain the system comprehensive equivalent parameters; based on the system comprehensive equivalent parameters, the observed power deficit value is calculated through the rotor motion equation of the system inertial center; the observed power deficit value is decomposed and the compensation amount is deducted through the system power balance equation to obtain the initial total power deficit of the power system. The time-domain expression of the system frequency response model that takes into account load regulation effects is as follows: ; In the formula, The system frequency deviation at time t; K is the load voltage regulation effect coefficient; K is the unit primary frequency regulation effect coefficient. This refers to the system voltage deviation. The power deficit is the observed value; D is the system damping coefficient; This is the load frequency regulation effect coefficient; ξ is the amplitude coefficient of the transient oscillation component; ξ is the damping ratio; The undamped natural angular frequency of the system; This is the damped oscillation angular frequency of the system; The initial phase angle of the transient oscillation component; The expression for the system power balance equation is: ; In the formula, These are power deficit observations; The power deficit is caused by a fault. This is for primary frequency modulation power compensation; Power compensation due to load frequency regulation effect; This is power compensation caused by the load voltage regulation effect; This represents the coupling amount between frequency and voltage; The expression for the rotor motion equation is as follows: ; In the formula, The equivalent inertial time constant of the generator; This is the system's rated frequency.

[0009] As a preferred embodiment of the adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization, the initial value calculation formula for the real-time frequency change rate of the low-frequency response curve obtained through the numerical differential algorithm is as follows: ; In the formula, ROCOF is the rate of change of the initial frequency of the system; This represents the system voltage deviation.

[0010] As a preferred embodiment of the adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution and utilization, in the process of screening and cutting off non-important loads with small frequency regulation effects based on the load identification algorithm and load importance level, the magnitude of the load's frequency regulation effect is determined by calculating the load's frequency regulation effect coefficient. The formula for calculating the frequency modulation effect coefficient is as follows: ; ; In the formula, This is the load frequency regulation effect coefficient; This represents the actual power consumed by the load. Let be the actual active power of the i-th load node; The index of the set of load nodes; Let be the rated power of node i at the rated frequency; Let be the real-time power of node i; , , … The load percentage is proportional to the frequency raised to the power of 0, 1, 2...n. This is the per-unit value for frequency.

[0011] As a preferred embodiment of the adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization, the formula for calculating the cumulative load shedding of the system is as follows: ; In the formula, This represents the cumulative load reduction of the system. For accurate power deficit values; The target recovery frequency for the system; This represents the number of load nodes. Let be the rated active power of the i-th load node; The amount of active power released to compensate for other regulatory measures.

[0012] This invention also provides an adaptive power grid low-frequency load shedding device based on three-level coordination of transmission and distribution, employing the above-mentioned adaptive power grid low-frequency load shedding method based on three-level coordination of transmission and distribution, including: The initial total power deficit calculation module is used to construct a wide-area measurement network covering the entire power transmission network by deploying synchronous phasor measurement units and wide-area measurement terminals at key nodes of the power transmission network; when a low-frequency over-limit is detected in the power system, synchronous operation data of key nodes of the entire power transmission network is collected through the wide-area measurement network; based on the synchronous operation data, a multi-machine equivalent model is constructed; based on the multi-machine equivalent model, the system power balance equation, the rotor motion equation, and the system frequency response model considering load regulation effects, the initial total power deficit of the power system is calculated. The accurate power deficit value acquisition module is used to plot the system's low-frequency response curve in real time based on the synchronous operation data; calculate the real-time frequency change rate of the low-frequency response curve using a numerical differentiation algorithm; set a threshold range for the frequency change rate based on the proportion of new energy access; determine whether the real-time frequency change rate exceeds the threshold range; if it exceeds, activate the power deficit correction mechanism to dynamically correct the initial total power deficit and obtain an accurate power deficit value; if it does not exceed, use the initial total power deficit as the accurate power deficit value. The load reduction instruction generation and distribution module is used to decompose the total load reduction task into the regional load reduction amount of each distribution network based on the accurate power deficit value, combined with the load scale, power supply reliability level and grid security constraints of each distribution network, and generate load reduction instructions through a load balancing allocation algorithm; the load reduction instructions are sent to the dispatch centers of each distribution network through the transmission and distribution network collaborative communication link; based on the load characteristics of each transformer substation in the jurisdiction, the distribution network dispatch center decomposes the regional load reduction amount into the specific load reduction instructions of each transformer substation and sends them to the IoT control terminal of the corresponding transformer substation. The IoT control terminal load shedding and feedback module is used to screen and shed non-critical loads with small frequency regulation effects based on load identification algorithms and load importance levels; the actual load shedding amount is collected and fed back to the distribution network dispatch center and the transmission network dispatch center through the IoT communication link. The load iteration shedding module is used by the power grid dispatch center to summarize the actual load shedding amount of each distribution area to obtain the cumulative load reduction of the system; based on the real-time frequency data of the system in the synchronous operation data, it determines whether the system frequency has recovered to the safe threshold; if it has recovered, the load shedding process is terminated; if it is lower than the safe threshold and the maximum load shedding round has not been reached, the power deficit estimation and correction process is iteratively performed based on the slope of the current system low-frequency response curve and the remaining power deficit until the system frequency recovers to the safe range or the full load shedding round is completed.

[0013] As a preferred embodiment of an adaptive power grid low-frequency load shedding device based on three-level coordination of transmission, distribution, and utilization, the key nodes of the transmission network in the initial total power deficit calculation module include: power plant busbars, regional tie lines, and load center busbars; the synchronous operation data include: voltage, current, frequency, phase angle, unit output, and total load.

[0014] As a preferred embodiment of an adaptive power grid low-frequency load shedding device based on three-level coordination of transmission, distribution, and utilization, the initial total power deficit calculation module, in the process of calculating the initial total power deficit of the power system, calculates the core characteristic parameters of the source and load through the system frequency response model that takes into account the load regulation effect; integrates the core characteristic parameters of the source and load into the multi-machine equivalent model to obtain the comprehensive equivalent parameters of the system; based on the comprehensive equivalent parameters of the system, calculates the observed power deficit value through the rotor motion equation of the system inertial center; and decomposes and deducts the compensation amount from the observed power deficit value through the system power balance equation to obtain the initial total power deficit of the power system.

[0015] The time-domain expression of the system frequency response model that takes into account load regulation effects is as follows: ; In the formula, The system frequency deviation at time t; K is the load voltage regulation effect coefficient; K is the unit primary frequency regulation effect coefficient. This refers to the system voltage deviation. The power deficit is the observed value; D is the system damping coefficient; This is the load frequency regulation effect coefficient; ξ is the amplitude coefficient of the transient oscillation component; ξ is the damping ratio; The undamped natural angular frequency of the system; This is the damped oscillation angular frequency of the system; The initial phase angle of the transient oscillation component; The expression for the system power balance equation is: ; In the formula, These are power deficit observations; The power deficit is caused by a fault. This is for primary frequency modulation power compensation; Power compensation due to load frequency regulation effect; This is power compensation caused by the load voltage regulation effect; This represents the coupling amount between frequency and voltage; The expression for the rotor motion equation is as follows: ; In the formula, The equivalent inertial time constant of the generator; This is the system's rated frequency.

[0016] As a preferred embodiment of an adaptive power grid low-frequency load shedding device based on three-level coordination of transmission, distribution, and utilization, in the accurate power deficit value acquisition module, during the process of calculating the real-time frequency change rate of the low-frequency response curve using the numerical differential algorithm, the initial value calculation formula for the real-time frequency change rate is as follows: ; In the formula, ROCOF is the rate of change of the initial frequency of the system; This represents the system voltage deviation.

[0017] As a preferred embodiment of the adaptive power grid low-frequency load shedding device based on three-level coordination of transmission, distribution and use, in the load shedding and feedback module of the Internet of Things control terminal, during the process of screening and shedding non-important loads with small frequency regulation effects based on the load identification algorithm and load importance level, the magnitude of the load's frequency regulation effect is determined by calculating the load's frequency regulation effect coefficient.

[0018] The formula for calculating the frequency modulation effect coefficient is as follows: ; ; In the formula, This is the load frequency regulation effect coefficient; This represents the actual power consumed by the load. Let be the actual active power of the i-th load node; The index of the set of load nodes; Let be the rated power of node i at the rated frequency; Let be the real-time power of node i; , , … The load percentage is proportional to the frequency raised to the power of 0, 1, 2...n. This is the per-unit value for frequency.

[0019] As a preferred embodiment of the adaptive power grid low-frequency load shedding device based on three-level coordination of transmission, distribution, and utilization, the calculation formula for the cumulative load shedding in the load iteration cut-off module is as follows: ; In the formula, This represents the cumulative load reduction of the system. For accurate power deficit values; M represents the system target recovery frequency; M represents the number of load nodes. Let be the rated active power of the i-th load node; The amount of active power released to compensate for other regulatory measures.

[0020] The present invention has the following advantages: First, improve the adaptability of frequency stability control: by building a multi-machine equivalent model based on wide-area measurement data, it can better reflect the actual dynamic characteristics of the power grid under the high proportion of new energy access, and make the power deficit estimation more in line with the system operating status.

[0021] Second, enhance the rationality of the load reduction strategy: by combining the frequency change rate with the proportion of new energy sources for dynamic correction, the load reduction strategy can be adjusted according to the severity of the disturbance, avoiding deviations caused by fixed-cycle load reduction.

[0022] Third, optimize the targeting of load shedding: select shedding targets based on the load frequency regulation effect coefficient, and prioritize shedding loads with weak frequency support, which helps to reduce the impact on important loads and improve power supply reliability.

[0023] Fourth, improve the closed-loop nature of the control process: evaluate the load reduction effect by predicting the steady-state offset value of the frequency, realize the closed-loop control of "prediction-load reduction-evaluation-optimization", and make the load reduction strategy more forward-looking.

[0024] Fifth, strengthen the three-level coordination capability of transmission, distribution and consumption: extend load reduction decisions to distribution network and consumer areas, realize the combination of centralized decision-making and distributed control, and make fuller use of controllable load resources on the distribution network side. Attached Figure Description

[0025] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0026] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0027] Figure 1 This is a flowchart illustrating the adaptive low-frequency load shedding method for power grids based on three-level coordination of transmission, distribution, and utilization provided in Embodiment 1 of the present invention. Figure 2 This is a schematic diagram of the equivalent system response model of a single machine in the adaptive power grid low-frequency load reduction method based on three-level coordination of transmission, distribution and utilization provided in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the system frequency response model considering the load regulation effect in the adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution and utilization provided in Embodiment 1 of the present invention. Figure 4 This is a schematic diagram of the frequency variation of systems with different proportions of new energy under the same power deficit in the adaptive grid low-frequency load reduction method based on three-level coordination of transmission, distribution and utilization provided in Embodiment 1 of the present invention. Figure 5 This is a schematic diagram of the ROCOF correction strategy under high renewable energy ratio in the adaptive power grid low-frequency load reduction method based on three-level coordination of transmission, distribution and utilization provided in Embodiment 1 of the present invention. Figure 6 This is a schematic diagram of the primary frequency regulation and active power characteristic curves of the system in the adaptive power grid low-frequency load reduction method based on three-level coordination of transmission, distribution and application provided in Embodiment 1 of the present invention. Figure 7 This is a schematic diagram of the precise load shedding process in the adaptive power grid low-frequency load reduction method based on three-level coordination of transmission, distribution and utilization provided in Embodiment 1 of the present invention; Figure 8 This is a schematic diagram of the architecture of the adaptive power grid low-frequency load shedding device based on three-level coordination of transmission and distribution provided in Embodiment 2 of the present invention. Detailed Implementation

[0028] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Example 1 See Figure 1 Embodiment 1 of the present invention provides an adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution and utilization, comprising the following steps: S1. By deploying synchronous phasor measurement units and wide-area measurement terminals at key nodes of the power transmission network, a wide-area measurement network covering the entire power transmission network is constructed. When a low-frequency over-limit is detected in the power system, synchronous operation data of key nodes of the entire power transmission network is collected through the wide-area measurement network. Based on the synchronous operation data, a multi-machine equivalent model is constructed. Based on the multi-machine equivalent model, the system power balance equation, the rotor motion equation, and the system frequency response model considering load regulation effects, the initial total power deficit of the power system is calculated. S2. Based on the synchronous operation data, the system low-frequency response curve is plotted in real time; the real-time frequency change rate of the low-frequency response curve is calculated using a numerical differentiation algorithm; a threshold range for the frequency change rate is set based on the proportion of new energy access; it is determined whether the real-time frequency change rate exceeds the threshold range; if it exceeds, a power deficit correction mechanism is activated to dynamically correct the initial total power deficit and obtain an accurate power deficit value; if it does not exceed, the initial total power deficit is used as the accurate power deficit value. S3. Based on the accurate power deficit value, combined with the load scale, power supply reliability level and grid security constraints of each distribution network, the total load reduction task is decomposed into the load reduction amount of each distribution network and load reduction instructions are generated through the load balancing allocation algorithm; the load reduction instructions are sent to the dispatch centers of each distribution network through the transmission and distribution network collaborative communication link; based on the load characteristics of each transformer substation in the jurisdiction, the distribution network dispatch center decomposes the load reduction amount of each region into specific load reduction instructions for each transformer substation and sends them to the IoT control terminal of the corresponding transformer substation. S4. Based on the load identification algorithm and load importance level, the IoT control terminal filters and cuts off non-critical loads with small frequency regulation effects; the actual load cut-off amount is collected and fed back to the distribution network dispatch center and the transmission network dispatch center through the IoT communication link. S5. The power grid dispatch center summarizes the actual load shedding amount of each distribution area to obtain the cumulative load reduction of the system; based on the real-time frequency data of the system in the synchronous operation data, it determines whether the system frequency has recovered to the safe threshold; if it has recovered, the load shedding process is terminated; if it is lower than the safe threshold and the maximum load shedding round has not been reached, the power deficit estimation and correction process is iteratively performed based on the slope of the current system low-frequency response curve and the remaining power deficit, until the system frequency recovers to the safe range or the full load shedding round is completed.

[0030] In this embodiment, in step S1, a wide-area measurement network covering the entire power grid is constructed by deploying synchronous phasor measurement units and wide-area measurement terminals at key nodes of the power grid; when a low-frequency over-limit is detected in the power system, synchronous operation data of key nodes of the entire power grid is collected through the wide-area measurement network; based on the synchronous operation data, a multi-machine equivalent model is constructed; based on the multi-machine equivalent model, the system power balance equation, the rotor motion equation, and the system frequency response model considering the load regulation effect, the initial total power deficit of the power system is calculated.

[0031] Specifically, firstly, key nodes in the transmission network are selected, such as power plant buses, regional tie lines, and load center buses, which are sensitive to system frequency changes and can reflect the overall operating status of the power grid. Synchronous phasor measurement units (PMUs) and wide-area measurement terminals are deployed at these nodes. All terminals are uniformly connected to the wide-area control center, forming a wide-area measurement network covering the entire transmission network to ensure real-time acquisition of synchronized operation data across the entire network. When the wide-area control center detects that the system frequency is below the rated frequency lower limit (e.g., 49.5Hz for a 50Hz system), it triggers the data acquisition process, synchronously collecting data such as voltage, current, frequency, phase angle, generator output, and total load at each key node through the wide-area measurement network, ensuring data synchronization and integrity. Subsequently, based on the collected synchronized operation data, a co-regulating generator group division method is adopted, classifying generators with similar dynamic characteristics and small relative angle changes into co-regulating generator groups. Each group is merged into an equivalent generator. Combining the real-time topology of the power grid, the primary frequency regulation characteristics of the generators, and the static characteristics of the load, a multi-generator equivalent model is constructed to recreate the actual dynamic characteristics of the complex power grid. Finally, using the multi-machine equivalent model as a carrier, the system power balance equation is used to decompose the power deficit composition, and the rotor motion equation is used to calculate the observed power deficit value. With the help of the source and load core parameters provided by the system frequency response model that takes into account the load adjustment effect, the initial total power deficit is calculated in a comprehensive manner, thus providing the initial data support for the load reduction decision.

[0032] Because the power system is a dynamic system, it maintains energy balance at every moment, and frequency changes are the process of converting kinetic energy into electrical energy. The system's frequency variation is jointly determined by the characteristics of the generating units and the load, and their effects are mutually coupled. The simplest dynamic frequency model is a first-order inertial element. Based on this, considering the load characteristics and the frequency regulation characteristics of the generating units, the improved system power balance equation is as follows: (1) In the formula, The system frequency deviation at time t; K is the load voltage regulation effect coefficient; K is the unit primary frequency regulation effect coefficient. This refers to the system voltage deviation. The power deficit is the observed value; D is the system damping coefficient; This is the load frequency regulation effect coefficient; ξ is the amplitude coefficient of the transient oscillation component; ξ is the damping ratio; The undamped natural angular frequency of the system; This is the damped oscillation angular frequency of the system; The initial phase angle of the transient oscillation component.

[0033] The power deficit observation ΔP is proportional to the rate of frequency change and is determined based on the equation of motion of the system's center of inertia. The equation of the rotor in the system is expressed as: (2)

[0034] In the formula, The equivalent inertial time constant of the generator is the larger the inertia, the slower the frequency drops, which is beneficial to system stability. The system's rated frequency is used. The frequency response of all loads is factored into the system based on load size. The voltage responses of each load node are summed to calculate the total power compensation.

[0035] In this embodiment, the active power balance of the system is disrupted, causing a change in the system frequency and subsequently a frequency dynamic process. To calculate the frequency change after the disturbance, a System Frequency Response (SFR) model needs to be established. By solving this model, the relationship between power deficit and frequency change is obtained. In system dynamic frequency analysis, the power system is typically simplified to an equivalent single-machine model. The simplified SFR model governor is approximated by a first-order model. The SFR model block diagram is shown below. Figure 2 As shown.

[0036] This section describes the dynamic frequency variation characteristics of a power system when an active power imbalance occurs, considering system inertia and damping. The input is the system's active power deficit. The output is the system frequency deviation. This link is essentially a first-order inertial link.

[0037] In the formula, D represents the system damping coefficient. When the system speed decreases, the losses caused by damping decrease, which can alleviate the power deficit to some extent. Primary frequency regulation is differential regulation.

[0038] This is a first-order model of a speed controller, which adjusts the mechanical power output based on the frequency deviation.

[0039] In the formula, K represents the ratio of the active power increment to the frequency offset after the first frequency adjustment. The larger K is, the smoother the power frequency characteristics of the unit, and the closer the frequency is to the rated level after the balance is restored. T is the adjustment time constant caused by valve opening, etc.

[0040] The minus sign "-" indicates negative feedback—the frequency decreases. When <0), the speed controller increases the mechanical power output, thereby suppressing frequency drops.

[0041] The simplified SFR model includes only the first-order response of the system inertia and the feedback model of the unit speed governor, neglecting the regulation effect of load changes in frequency and voltage. The load frequency characteristic can be regarded as a closed-loop feedback, with the input being the frequency offset and the output being the power compensation. The input is multiplied by its coefficient. The power compensation is then fed back to the output, at which point the SFR model changes from a single closed loop to a double closed loop. Voltage changes are then treated as a disturbance input and multiplied by a coefficient. This directly impacts the power deficit.

[0042] The improved SFR model is as follows: Figure 3 As shown. Assuming that both the power deficit and voltage disturbance are step signals, the transfer function of this model is: (3) In the formula, This is the load voltage regulation effect coefficient; This refers to the system voltage deviation. is the system frequency deviation; s is the Laplace operator; D is the system damping coefficient; is the load frequency regulation effect coefficient; K represents the ratio of active power increment to frequency offset after primary frequency regulation; T is the regulation time constant caused by valve opening, etc.

[0043] The load SER model considering load characteristics is a second-order system, and the standard form of the response is obtained by rearranging: (4) in, (5) (6) Performing an inverse Laplace transform on the output yields the time-domain expression of the frequency response: (7) in, (8) (9) (10) In the formula, The system frequency deviation at time t; The standard form of the response; K is the load voltage regulation effect coefficient; K is the unit primary frequency regulation effect coefficient. This refers to the system voltage deviation. The power deficit is the observed value; D is the system damping coefficient; This is the load frequency regulation effect coefficient; ξ is the amplitude coefficient of the transient oscillation component; ξ is the damping ratio; The undamped natural angular frequency of the system; This is the damped oscillation angular frequency of the system; The initial phase angle of the transient oscillation component.

[0044] As can be seen from the above formula (7), without considering low-frequency load shedding, under only one frequency regulation and load adjustment, the frequency change is divided into two parts: steady state and transient state. The numerator of the steady state component amplitude is affected by the voltage regulation effect, while the denominator is composed of the unit regulation effect coefficient K, the unit damping D, and the load regulation effect coefficient. They are composed of three components. After conversion to per-unit system, the three parameters in the denominator have equal weights, all of which suppress frequency drops and are superimposed on each other. The transient component exhibits sinusoidal fluctuations, with its amplitude decaying exponentially, and its variation pattern is closely related to the load characteristics.

[0045] In this embodiment, in step S2, the system low-frequency response curve is plotted in real time based on the synchronous operation data; the real-time frequency change rate of the low-frequency response curve is calculated using a numerical differentiation algorithm; a threshold range for the frequency change rate is set based on the proportion of new energy access; it is determined whether the real-time frequency change rate exceeds the threshold range; if it exceeds, a power deficit correction mechanism is activated to dynamically correct the initial total power deficit and obtain an accurate power deficit value; if it does not exceed, the initial total power deficit is used as the accurate power deficit value.

[0046] Specifically, firstly, based on the real-time frequency data from the synchronous operation data collected in step S1, a low-frequency response curve of the system is plotted in real time according to the time series, intuitively presenting the trend of system frequency change over time and providing visual support for subsequent frequency change rate calculation. Then, a numerical differential algorithm is used to process the low-frequency response curve to calculate the real-time rate of change of frequency (ROCOF), quantifying the speed of system frequency change and reflecting the severity of power deficit. Next, considering the current proportion of renewable energy access in the power grid, the proportion of renewable energy access is divided into 2-3 intervals, each interval corresponding to a different frequency change rate threshold. A reasonable threshold range is set to adapt to the impact of the low inertia characteristics of renewable energy on system frequency changes. Next, the calculated real-time frequency change rate is compared with the set threshold range. If the real-time frequency change rate exceeds the threshold range, it indicates that there is a deviation in the initial total power deficit estimation. The power deficit correction mechanism needs to be activated. By combining the node load fluctuation, unit output deviation, and tie-line power exchange data collected in real time by the wide-area measurement network, the initial total power deficit is dynamically adjusted to eliminate the estimation deviation and obtain an accurate power deficit value. If the real-time frequency change rate does not exceed the threshold range, it indicates that the initial total power deficit estimation is consistent with the current system state and can be directly used as the accurate power deficit value for subsequent load reduction task allocation.

[0047] When the system is disturbed, the initial value of the rate of change of frequency, i.e., at time t=0, is the system's ROCOF (Rate of Change of Frequency): (11) In the formula, ROCOF is the rate of change of the initial frequency of the system; This represents the system voltage deviation.

[0048] Inverse solution to power deficit for: (12) The actual power deficit can be calculated from the measured frequency change using the above formula. It can be seen that voltage fluctuations instantaneously affect frequency changes, and the power deficit calculation is affected by the coupling effect of unit inertia and load characteristics. If the proportion of constant power load in the system is small, ignoring voltage characteristics may lead to the mistaken assumption that frequency drops are not suppressed, thus underestimating the actual power deficit. Furthermore, the initial ROCOF is not affected by load frequency regulation effects or the unit's primary frequency regulation parameters. This is because at the instant the power deficit occurs, the system frequency remains constant due to inertia.

[0049] When the system experiences a power deficit, the system frequency gradually decreases, and primary frequency regulation begins. Here, we assume sufficient system reserve capacity and do not consider low-frequency load shedding; we calculate the time when the lowest frequency peak occurs.

[0050] According to the frequency change curve, at the lowest frequency point, the rate of frequency change is 0. (13) According to the above formula, we can obtain: (14) In the formula, The time it takes for the frequency response curve to reach the nth peak. It is the equivalent phase angle of the transient oscillation component.

[0051] As shown in the above formula, the time of the lowest point is independent of the magnitude of the power deficit caused by the fault. Therefore, if the node voltage can remain unchanged, it will not affect the time of the lowest point.

[0052] Substituting the time of the lowest point into the equation yields the maximum frequency offset: (15) In the formula, This represents the maximum frequency offset.

[0053] As shown in equation (15), the regulation effect coefficient, damping coefficient, and load regulation effect coefficient of primary frequency regulation all play a supporting role in frequency drop. Low-frequency load shedding requires judging the severity of the fault by the maximum frequency deviation. If the load LFR parameter is ignored, the severity of the accident will be overestimated. In some cases, this will cause the load shedding device to malfunction, resulting in unnecessary losses.

[0054] According to the final value theorem, the steady-state frequency offset is: (16) In the formula, This represents the steady-state frequency offset.

[0055] In this embodiment, the impact of new energy sources on the system frequency characteristics must also be considered. After decades of development, China's power system now features large-capacity generating units, ultra-high / extra-high voltage lines, and extensive interconnection. Due to the increased system capacity, small random disturbances within the system can be regulated by the rotational inertia of the generating units. However, from another perspective, the system's ability to cope with high-power surges is constantly weakening.

[0056] Compared to smaller generating units of the same capacity, larger generating units have a relatively smaller equivalent moment of inertia. Especially with the large-scale centralized development of new energy sources and the replacement of many conventional units, the proportion of new energy output in the power grid is constantly increasing, leading to a further reduction in the overall system inertia. In the event of a serious fault causing a large power deficit in the system, the fluctuations in system frequency will be more severe, further testing the system's control measures in emergency situations.

[0057] In complex systems, to simplify calculations, generators whose relative angle changes are not significant during dynamic processes are referred to as a coherent generator group, which can be combined into a single equivalent generator for calculation. The following formula describes the equivalent inertial time constant. The equation: (17) In the formula, H i The inertial time constant of unit i; S i S is the rated generating capacity of unit i; s N represents the total rated output power of the units in the system; N represents the total number of generators in the system participating in the coordinating group division.

[0058] Describe the unit's inertial time constant H i The equation is: (18) In the formula, J is the rotor's moment of inertia; Ω N This is the rated mechanical angular velocity of the rotor.

[0059] The rotational inertia of conventional generating units and the effective rotational inertia of renewable energy generating units constitute the total system inertia. After renewable energy generating units are connected to the grid and replace conventional generating units, the installed capacity S of unit i... s The moment of inertia J remains unchanged, but the moment of inertia J decreases significantly, resulting in H i This reduction further reduces the system's equivalent time constant. Reduce, therefore, The system's equivalent inertial time constant will decrease as the proportion of new energy units in the power grid increases. The system's equivalent inertial time constant when new energy is connected can be calculated according to equations (17) and (18).

[0060] like Figure 4 The figure shows the trend of system frequency variation under different renewable energy ratios with the same power deficit (5%). Figure 4 Analysis of the results shows that as the proportion of new energy units in the system increases, the system frequency drops faster and to a greater extent. The same degree of power deficit has a more severe frequency response in a grid with a high proportion of new energy. At the same time, due to the lower technical standards for grid connection of new energy power generation, its withstand rate and voltage withstand capabilities are worse than those of conventional thermal / hydropower units. During an accident, abnormal frequency / voltage fluctuations may cause large-scale grid disconnection of units, further amplifying the power deficit.

[0061] Traditional tuning methods do not consider the impact of the system's overall inertia on frequency. If the system's overall inertia is small, the system frequency will drop more significantly at the initial moment of a power deficit, potentially leading to excessive load shedding by the low-frequency load shedding device and causing unnecessary economic losses. Furthermore, traditional low-frequency load shedding schemes require the system frequency to be exactly below the operating frequency of a specific load cycle to initiate shedding. When the power deficit is large, this may cause the optimal shedding timing to be missed, resulting in cascading failures and further amplifying the fault. Therefore, it is necessary to propose an optimized low-frequency load shedding scheme to adapt to the future grid integration of large-scale renewable energy sources.

[0062] In this embodiment, in step S3, based on the accurate power deficit value, combined with the load scale, power supply reliability level and grid security constraints of each distribution network, the total load reduction task is decomposed into the partition load reduction amount of each distribution network and load reduction instructions are generated through the load balancing allocation algorithm; the load reduction instructions are sent to the dispatch centers of each distribution network through the transmission and distribution network collaborative communication link; based on the load characteristics of each transformer substation in the jurisdiction, the distribution network dispatch center decomposes the partition load reduction amount into the specific load reduction instructions of each transformer substation and sends them to the IoT control terminal of the corresponding transformer substation.

[0063] Specifically, firstly, using the accurate power deficit value obtained in step S2 as the total base, and comprehensively considering the load scale, power supply reliability level, and grid security constraints of each distribution network, including distribution line current carrying capacity constraints and voltage constraints, to avoid grid security risks caused by load reduction, a load balancing allocation algorithm is adopted to fairly and reasonably decompose the total load reduction task into the regional load reduction amount of each distribution network. At the same time, corresponding load reduction instructions are generated, specifying the load reduction time, total load reduction amount, and other requirements for each distribution network. Subsequently, through the transmission and distribution network collaborative communication link, the load reduction instructions are quickly and accurately sent to the dispatch centers of each distribution network to ensure the real-time and reliable transmission of instructions. Finally, after receiving the regional load reduction instructions, each distribution network dispatch center, combined with the load characteristics of each transformer substation within its jurisdiction, further refines the regional load reduction amount into the specific load reduction amount for each transformer substation, generating specific load reduction instructions for each substation, specifying the load type and load amount to be cut off in the substation, and then sending them to the corresponding substation's IoT control terminal through the IoT communication link to complete the step-by-step penetration of the load reduction instructions, preparing for subsequent load shedding.

[0064] According to the "Technical Regulations for Automatic Low-Frequency Load Shedding in Power Systems," the initiation threshold for low-frequency load shedding in power systems should be set at a high value. This is because low-frequency load shedding is the last line of defense for maintaining the safety and stability of the power grid. It requires disconnecting appropriate loads when there is a severe power deficit and no available capacity to prevent the frequency from continuing to drop. A higher initiation frequency allows the load shedding device to disconnect a certain amount of load at the initial stage of the frequency drop, effectively raising the minimum point of the system frequency and reducing the risk of system instability.

[0065] However, in small-scale power deficit faults, the load regulation effect and the primary frequency regulation characteristics of the unit should be fully utilized, as well as frequency regulation methods that do not cause load outages. Examples include DC LFC and low-frequency pump switching and rapid regulation of renewable energy sources in pumped storage stations. If the starting frequency is set too high, system reserves and other regulation measures will not be fully utilized, leading to malfunctions of load shedding devices in situations where load shedding is not necessary, resulting in unnecessary load shedding losses.

[0066] Taking into account the low-frequency tolerance of the unit and load, the safety of the system, and the starting speed of spinning standby, the low-frequency load shedding start-up cycle in my country is generally set at 49.0Hz. Furthermore, without considering low-frequency load shedding, the power deficit when the lowest point of the system's transient frequency is higher than 49.0Hz and the steady-state frequency is higher than 49.5Hz is defined as the system reference power deficit.

[0067] Based on the above transfer function model, the frequency evolution of the power system's inertia center is calculated using a wide-area measurement system during the initial stage of frequency sag. The minimum frequency is calculated as follows: (19) From this equation, we can obtain the steady-state frequency of the system. for: (20) Based on the frequency of each node measured by the wide-area measurement system and the SFR model of the control center, if the predicted lowest point of the system frequency is lower than 49.0Hz, or if the steady-state value of the frequency does not reach 49.5Hz after frequency recovery, it is determined that the power deficit has reached the load reduction start-up threshold.

[0068] Furthermore, the coordination between different rounds of low-frequency load shedding under changes in frequency change rate and system operating state should also be considered. Therefore, the load shedding initiation criterion for the adaptive low-frequency load shedding strategy is: (twenty one) In the formula, The union symbol represents the set of sets. The cumulative time since the last load shedding operation; This is the delay for the next round of load shedding. It can be seen that as the frequency begins to decrease, the increment of frequency change begins to shrink. Low-frequency load shedding can be initiated when the estimated minimum frequency is less than 49Hz and the steady-state frequency is less than 49.5Hz, and the cumulative time of the previous low-frequency load shedding is greater than the action delay.

[0069] In this embodiment, considering the frequency characteristics of a power grid with a high proportion of renewable energy facing power deficits, and given that current low-frequency load shedding schemes do not account for the frequency deficit estimation errors caused by the low inertia of renewable energy, it is necessary to correct and optimize the estimated power deficit when renewable energy output is high. An action criterion for accelerating load shedding should be established at the load shedding point to improve the speed and accuracy of load shedding.

[0070] according to Figure 4 It can be seen that the higher the proportion of new energy access, the better. It will become smaller, but It remains basically unchanged. According to equations (12), (16), and (21), it can be concluded that the increase in ROCOF due to the access of new energy sources leads to an increase in the estimated ROCOF. Smaller, power deficit and system steady-state frequency As the load increases, the corresponding cumulative load reduction of the system can be obtained from formula (25). An increase in ROCOF may lead to over-cutting. It is necessary to adjust the corresponding startup or acceleration criteria based on the size of ROCOF.

[0071] As the proportion of renewable energy gradually increases, the ROCOF value also increases. Therefore, considering the proportion of renewable energy in the system and comparing df / dt, if the frequency variation is large and the renewable energy proportion is high, the power deficit should be reduced; if the frequency variation is small and the renewable energy proportion is relatively small, the power deficit should be slightly reduced; if the renewable energy proportion is very small, no adjustment to the estimated power deficit is needed. The specific implementation method is as follows... Figure 5 As shown.

[0072] The system experienced a power deficit. Afterwards, when the penetration rate of new energy... When the inertia increases, the equivalent inertia decreases, and under the same deficit... It is magnified. Therefore, utilizing During the offline phase, respectively , The frequency change rate threshold corresponding to the reference deficit measured at the penetration rate , ( During online testing, actual measurements will be taken. The power deficit is finally corrected by comparing it with the threshold. Where: This refers to the system's real-time frequency. The proportion of real-time power output from new energy sources to the system capacity (penetration rate). , These are the boundary values ​​for the offline permeability intervals (e.g., 10%, 30%); ΔROCOF1 and ΔROCOF2 are the permeability... and In this scenario, the difference between the actual ROCOF of the system and the ROCOF without renewable energy access; , Is , Apply reference deficit The values ​​for adjusting the power change rate and power deficit estimate are measured in real time. Since the proportion of real-time renewable energy output in the grid cannot be accurately measured, they can be adjusted based on the low-frequency response curves when renewable energy is connected and not connected. This is typically divided into 2-3 intervals, with each interval having a renewable energy connection rate of 20%-30%. For example, Figure 5 The system is divided into three intervals: when the renewable energy source accounts for 50% of the system capacity at maximum output, the intervals can be set as less than 10%, greater than 10% and less than 30%, and greater than 30%. The frequency change rate df / dt is adjusted according to the frequency change of the system's low-frequency response when the renewable energy source accounts for 10% and 30%, respectively, denoted as df1 and df2. When a low frequency is detected, the system compares df / dt with the preset thresholds df1 and df2 to determine the corresponding system operating interval. Simultaneously, the ROCOF value is modified according to the preset ΔROCOF to calculate the power deficit.

[0073] In this embodiment, in step S4, based on the load identification algorithm and the load importance level, the IoT control terminal filters and cuts off non-critical loads with small frequency regulation effects; the actual load cut-off amount is collected and fed back to the distribution network dispatch center and the transmission network dispatch center in sequence through the IoT communication link.

[0074] Specifically, firstly, after receiving the specific load reduction command issued in step S3, the IoT control terminal activates the load identification algorithm to identify all loads within the distribution area, calculates the frequency regulation effect coefficient of each load, and filters out loads with small frequency regulation effects. Simultaneously, based on the load importance level classification standard, it eliminates first- and second-level important loads, determining the final list of non-important loads to ensure that power supply to important users is not affected. Subsequently, the IoT control terminal triggers the control switch action of the corresponding non-important load to execute the load shedding operation, achieving precise load shedding. After shedding is completed, the IoT control terminal collects the actual load shedding amount through its built-in measurement module to confirm whether the shedding effect meets the command requirements. Then, through the IoT communication link, it feeds back the actual load shedding amount to the distribution network dispatch center. The distribution network dispatch center then aggregates the actual shedding data of all distribution areas within its jurisdiction and feeds it back to the transmission network dispatch center, forming a closed loop of "shearing-feedback" to ensure that the dispatch center has real-time control over the load reduction execution status of each distribution area.

[0075] The load power-frequency static characteristic refers to the relationship between the power consumption of a load and the node frequency in a power system when the load maintains its rated voltage and total output remains constant. During large-scale power deficit faults, the frequency drop is significant, and the impact of frequency changes on the load cannot be ignored. These loads are called frequency-sensitive loads, typically composed of induction motors. The frequency drop causes a decrease in rotor speed, resulting in reduced power consumption. The change in load power caused by frequency changes near the rated frequency is called the load frequency response (LFR). The load frequency response coefficient is the ratio of a 1% frequency change to the percentage change in active power at the node. The reciprocal of the response coefficient is called the droop coefficient, which reflects the characteristics of the speed governor in primary frequency regulation. The primary frequency regulation characteristics of the unit and the load power-frequency characteristics are as follows: Figure 6 As shown, the two trends are opposite, and their intersection is the steady-state value of the system.

[0076] in, The first priority load set to be removed consists of loads with low frequency regulation effect coefficients (frequency insensitive) and low importance. The second priority load set to be removed consists of loads with high frequency regulation effect coefficients or high importance. Only remove when insufficient; The frequency trigger setting for the first round of low-frequency load shedding (e.g., 49.0 Hz) is set; the system will activate when the frequency drops below this value. resection; The frequency trigger setting for the second round of low-frequency load shedding (e.g., 48.5 Hz) is set; the frequency will be activated if it continues to drop below this value. resection; for The total amount of load power that can actually be cut off; for The total amount of load power that can actually be cut off.

[0077] The load regulation effect coefficient depends on the load composition. Different load types also have different power-frequency static characteristics, specifically categorized as follows: 1. Loads whose active power is independent of frequency are usually purely resistive loads, such as heaters and incandescent lamps.

[0078] Second, loads whose active power is proportional to frequency, such as ball mills.

[0079] 3. Loads with active power proportional to the square of the frequency, such as transformer losses.

[0080] Fourth, loads whose active power is proportional to a power higher than the load, such as fans and water pumps.

[0081] The expression describing the combined frequency characteristics of various loads is as follows: (twenty two) In the formula, Let be the actual active power of the i-th load node; The index of the set of load nodes; This represents the actual power consumed by the load. Let be the rated power of node i at the rated frequency; The system's rated frequency; Let be the real-time power of node i; , , … The load percentage is proportional to the frequency raised to the power of 0, 1, 2...n, and satisfies a0 + a1 + a2 + ... + a n-1 =1.

[0082] For frequency Taking the derivative, we obtain the frequency regulation effect coefficient of the load node: (twenty three) In the formula, This is the per-unit value for frequency.

[0083] As can be seen from the above formula, loads with insensitive frequency characteristics are prioritized for shedding based on their frequency characteristics. Simultaneously, loads are prioritized based on their importance, with more important loads being shedding last. The total load power to be shedding is determined by the power deficit estimated by the adaptive load shedding strategy. Distribution networks in areas with concentrated less important loads are selected based on load importance ranking. Then, the proportion of load ZIP in each distribution area is used via the IoT system to selectively shed loads with low frequency regulation coefficients through smart control switches within the distribution areas. After load shedding is completed, the total amount of shedding load is fed back to the main grid level by level, completing a full round of low-frequency load shedding. The specific load shedding process is as follows: Figure 7 As shown.

[0084] First, load identification is performed, classifying loads according to frequency characteristics and importance. Then, regional ZIP aggregation and network-wide ZIP aggregation are performed separately. This is based on the frequency change rate. Differential with time Estimated power deficit The total power to be cut off is then obtained through adjustments. Next, precise load shelving is performed: loads are prioritized based on importance, with priority given to shelving areas with concentrated, less important loads. Then, through the IoT and smart switches in the distribution network area, the shelving is adjusted at the regional distribution network level according to the regional ZIP shelving ratio and... Selectively disconnect loads with low frequency regulation coefficients; if the power deficit is still not met, further full-network ZIP disconnection is implemented, also proportionally and... The corresponding load is then removed. Finally, the total amount of load actually removed is fed back to the main grid level by level to complete the entire load reduction operation. Figure 7 In this context, load identification refers to the identification and classification of loads based on their frequency characteristics and importance; regional ZIP aggregation refers to the aggregation and statistics of loads at the regional distribution network level according to the ZIP model (constant impedance Z, constant current I, constant power P); and whole-network ZIP aggregation refers to the aggregation of all network loads at the entire main network level using the ZIP model. The estimate is based on The system estimates the current power deficit using parameters such as transformer substations; it uses substation areas to represent substation-level load units (i.e., the power supply area of ​​distribution transformers); regional distribution networks represent regional distribution networks, located between the main grid and substations, responsible for regional load aggregation and shelving; regional main grids represent the upper-level transmission network (main grid), responsible for global power balance and load shelving command issuance; precise load shelving refers to precise shelving based on load importance ranking, prioritizing the shelving of less important loads; regional and regional ZIP shelving refers to shelving at the regional and regional distribution network levels, according to the ZIP model ratio and power deficit. Remove the load.

[0085] In this embodiment, in step S5, the power grid dispatch center summarizes the actual load shedding amount of each distribution area to obtain the cumulative load reduction of the system; based on the real-time frequency data of the system in the synchronous operation data, it determines whether the system frequency has recovered to the safe threshold; if it has recovered, the load shedding process is terminated; if it is lower than the safe threshold and the maximum load shedding round has not been reached, the power deficit estimation and correction process is iteratively performed based on the slope of the current system low-frequency response curve and the remaining power deficit until the system frequency recovers to the safe range or the full load shedding round is completed.

[0086] Specifically, firstly, the transmission network dispatch center receives the actual load shedding data from each distribution network dispatch center, summarizes and calculates the actual shedding data for all transformer areas, obtains the cumulative load reduction of the system, and clarifies the scale of load reduction currently completed. Then, combining the system frequency data collected in real time by the wide-area measurement network, it determines whether the system frequency has recovered to the safe threshold: if the system frequency has recovered to the safe threshold, it indicates that the load reduction is sufficient and the system frequency has stabilized. At this point, the entire load reduction process is terminated, and key data from this load reduction is recorded to support subsequent load reduction strategy optimization. If the system frequency is still below the safe threshold, and the current load reduction round has not reached the preset maximum load reduction round, it indicates that the current load reduction is insufficient and iterative optimization is required. At this point, based on the slope of the current system low-frequency response curve and the remaining power deficit, the process returns to step S2 to re-execute the power deficit estimation and correction process, recalculate the accurate power deficit value, allocate load reduction tasks, and execute load shedding until the system frequency recovers to a safe range or all rounds of load shedding are completed, ensuring system frequency stability and avoiding safety hazards such as frequency collapse.

[0087] Among them, by using the improved system frequency response model and combining the initial frequency ROCOF of each node monitored by the wide-area measurement system, the actual power deficit caused by system faults is calculated.

[0088] Low-frequency load shedding strategies do not aim to remove the full amount of load lost to restore the system to its rated frequency. This is because it would result in excessively high load shedding costs, neglecting the economics of the load shedding scheme. Therefore, it is necessary to set a desired system recovery frequency. Assuming a frequency of 49.8Hz, while ensuring normal system operation, the frequency regulation effect of the load should be fully utilized to reduce load shedding. Furthermore, the load shedding should also consider the system's reserve capacity and the use of other regulation measures.

[0089] The load shedding amount, taking into account both the load regulation effect and the primary frequency regulation of the unit, is calculated as follows: (twenty four) By solving for the shear load, we can obtain: (25) In the formula, This represents the cumulative load reduction of the system. For accurate power deficit values; The target recovery frequency for the system; This represents the number of load nodes. Let be the rated active power of the i-th load node; The amount of active power released to compensate for other regulatory measures.

[0090] According to the dynamic principles of power system frequency, any change in generator output will cause a sudden change in the system frequency rate of change. When generator output increases or low-frequency load shedding devices activate, the system power deficit decreases accordingly, resulting in a step increase in the system frequency rate of change. Conversely, a decrease in generator output and a sudden increase in load will lead to a decrease in the system frequency rate of change, thereby accelerating the frequency drop.

[0091] Adaptive load shedding calculates the actual power deficit based on the initial frequency change rate and cuts it off in one go. If the load shedding is appropriate, the frequency will recover immediately until it eventually stabilizes at the desired recovery frequency. However, in reality, various operating conditions and prediction uncertainties often exist, leading to errors between the calculated and actual power deficit. This can result in two unexpected situations: First, after the load is shedding, the frequency continues to decrease. This indicates a deficit, meaning the calculated power deficit is too small. In this case, the electromagnetic power of the system is still less than the mechanical power, and the rotor is in a deceleration state. In this situation, it is necessary to re-estimate the system power deficit based on the frequency change rate after load shedding and proceed with the next load shedding. Second, after the load is shedding, the frequency begins to recover but ultimately fails to reach the desired recovery frequency, hovering for an extended period between 49.5Hz and 49.0Hz. In this case, special wheel operation needs to be considered. The special wheel, through offline tuning, determines the load shedding amount and, by shedding a small amount of load, allows the frequency hovering below 49.5Hz to recover to the desired recovery frequency.

[0092] The application scenarios of this invention are as follows: In provincial power grid dispatch scenarios with a high proportion of renewable energy connected to the grid, this invention can rely on a wide-area measurement network to perceive the power deficit caused by fluctuations in wind power and photovoltaic output in real time. By adapting the low inertia characteristics to a multi-machine equivalent model, it can dynamically adjust the load reduction strategy to ensure system frequency stability.

[0093] In the scenario of coordinated operation between large industrial parks and urban power distribution networks, this invention can extend load reduction decisions to the park and transformer area level loads, prioritize the removal of non-critical loads with weak frequency regulation effects, reduce the impact on power supply for key production and residential use, and improve the resilience of power distribution network operation.

[0094] In emergency support scenarios for inter-provincial interconnected power grids, this invention can combine tie-line power exchange data to quickly assess inter-regional support capabilities and local load reduction needs, achieving three-level coordination of transmission, distribution, and application, and avoiding network-wide frequency instability caused by disturbances in a single region.

[0095] In scenarios of large-scale power deficit caused by extreme weather or major faults, this invention can predict the system's recovery capability by forecasting the steady-state frequency offset value, iteratively optimize the load reduction amount, and prevent insufficient load reduction from causing frequency collapse or excessive disconnection from causing power outage.

[0096] It should be noted that the method of this disclosure embodiment can be executed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method of this disclosure embodiment, and the multiple devices will interact with each other to complete the method described.

[0097] It should be noted that the above description describes some embodiments of this disclosure. In some cases, the described actions or steps can be performed in a different order than that shown in the above embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0098] Example 2 See Figure 8 Embodiment 2 of the present invention also provides an adaptive power grid low-frequency load shedding device based on three-level coordination of transmission and distribution, comprising: The initial total power deficit calculation module 001 is used to construct a wide-area measurement network covering the entire power transmission network by deploying synchronous phasor measurement units and wide-area measurement terminals at key nodes of the power transmission network; when a low-frequency over-limit is detected in the power system, synchronous operation data of key nodes of the entire power transmission network is collected through the wide-area measurement network; based on the synchronous operation data, a multi-machine equivalent model is constructed; based on the multi-machine equivalent model, the system power balance equation, the rotor motion equation, and the system frequency response model considering load regulation effects, the initial total power deficit of the power system is calculated. The accurate power deficit value acquisition module 002 is used to draw the system low-frequency response curve in real time based on the synchronous operation data; calculate the real-time frequency change rate of the low-frequency response curve through a numerical differentiation algorithm; set a threshold range for the frequency change rate based on the proportion of new energy access; determine whether the real-time frequency change rate exceeds the threshold range; if it exceeds, activate the power deficit correction mechanism to dynamically correct the initial total power deficit and obtain an accurate power deficit value; if it does not exceed, use the initial total power deficit as the accurate power deficit value. The load reduction instruction generation and distribution module 003 is used to decompose the total load reduction task into the regional load reduction amount of each distribution network based on the accurate power deficit value, combined with the load scale, power supply reliability level and grid security constraints of each distribution network, and generate load reduction instructions through a load balancing allocation algorithm; and to send the load reduction instructions to the dispatch centers of each distribution network through the transmission and distribution network collaborative communication link; based on the load characteristics of each transformer substation in the jurisdiction, the distribution network dispatch center decomposes the regional load reduction amount into the specific load reduction instructions of each transformer substation and sends them to the IoT control terminal of the corresponding transformer substation. The IoT control terminal load shedding and feedback module 004 is used to screen and shed non-critical loads with small frequency regulation effects based on load identification algorithms and load importance levels; the actual load shedding amount is collected and fed back to the distribution network dispatch center and the transmission network dispatch center through the IoT communication link. The load iteration shedding module 005 is used by the power grid dispatch center to summarize the actual load shedding amount of each distribution area to obtain the cumulative load reduction of the system; based on the real-time frequency data of the system in the synchronous operation data, it determines whether the system frequency has recovered to the safe threshold; if it has recovered, the load shedding process is terminated; if it is lower than the safe threshold and the maximum load shedding round has not been reached, the power deficit estimation and correction process is iteratively performed based on the slope of the current system low-frequency response curve and the remaining power deficit until the system frequency recovers to the safe range or the full load shedding round is completed.

[0099] In this embodiment, the key nodes of the transmission network in the initial total power deficit calculation module 001 include: power plant busbars, regional tie lines and load center busbars; the synchronous operation data include: voltage, current, frequency, phase angle, unit output and total load.

[0100] In this embodiment, the initial total power deficit calculation module 001 calculates the initial total power deficit of the power system by using the system frequency response model that takes load regulation effects into account to obtain the source-load core characteristic parameters; integrating the source-load core characteristic parameters into the multi-machine equivalent model to obtain the system comprehensive equivalent parameters; based on the system comprehensive equivalent parameters, calculating the observed power deficit value through the rotor motion equation of the system inertial center; and decomposing and deducting the compensation amount from the observed power deficit value through the system power balance equation to obtain the initial total power deficit of the power system.

[0101] The time-domain expression of the system frequency response model that takes into account load regulation effects is as follows: ; In the formula, The system frequency deviation at time t; K is the load voltage regulation effect coefficient; K is the unit primary frequency regulation effect coefficient. This refers to the system voltage deviation. The power deficit is the observed value; D is the system damping coefficient; This is the load frequency regulation effect coefficient; ξ is the amplitude coefficient of the transient oscillation component; ξ is the damping ratio; The undamped natural angular frequency of the system; This is the damped oscillation angular frequency of the system; The initial phase angle of the transient oscillation component; The expression for the system power balance equation is: ; In the formula, These are power deficit observations; The power deficit is caused by a fault. This is for primary frequency modulation power compensation; Power compensation due to load frequency regulation effect; This is power compensation caused by the load voltage regulation effect; This represents the coupling amount between frequency and voltage;

[0102] The expression for the rotor motion equation is as follows: ; In the formula, The equivalent inertial time constant of the generator; This is the system's rated frequency.

[0103] In this embodiment, in the accurate power deficit value acquisition module 002, during the process of calculating the real-time frequency change rate of the low-frequency response curve using the numerical differentiation algorithm, the initial value calculation formula for the real-time frequency change rate is as follows: ; In the formula, ROCOF is the rate of change of the initial frequency of the system; This represents the system voltage deviation.

[0104] In this embodiment, in the IoT control terminal load cut-off and feedback module 004, during the process of screening and cutting off non-important loads with small frequency regulation effects based on the load identification algorithm and load importance level, the magnitude of the load's frequency regulation effect is determined by calculating the load's frequency regulation effect coefficient. The formula for calculating the frequency modulation effect coefficient is as follows: ; ; In the formula, This is the load frequency regulation effect coefficient; This represents the actual power consumed by the load. Let be the actual active power of the i-th load node; The index of the set of load nodes; Let be the rated power of node i at the rated frequency; Let be the real-time power of node i; , , … The load percentage is proportional to the frequency raised to the power of 0, 1, 2...n. This is the per-unit value for frequency.

[0105] In this embodiment, the formula for calculating the cumulative load reduction of the system in the load iteration and removal module 005 is as follows: ; In the formula, This represents the cumulative load reduction of the system. For accurate power deficit values; The target recovery frequency for the system; This represents the number of load nodes. Let be the rated active power of the i-th load node; The amount of active power released to compensate for other regulatory measures.

[0106] It should be noted that the information interaction and execution process between the modules of the above system are based on the same concept as the method embodiment in Embodiment 1 of this application, and the resulting technical effects are the same as those in the method embodiment of this application. For details, please refer to the description in the method embodiment shown above in this application, and it will not be repeated here.

[0107] Example 3

[0108] Embodiment 3 of the present invention provides a non-transitory computer-readable storage medium storing program code for an adaptive grid low-frequency load shedding method based on three-level coordination of transmission and distribution. The program code includes instructions for executing the adaptive grid low-frequency load shedding method based on three-level coordination of transmission and distribution as described in Embodiment 1 or any possible implementation thereof.

[0109] Computer-readable storage media can be any available medium that a computer can access, or a data storage device such as a server or data center that integrates one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

[0110] Example 4

[0111] Embodiment 4 of the present invention provides an electronic device, including: a memory and a processor; The processor and the memory communicate with each other via a bus; the memory stores program instructions that can be executed by the processor, and the processor can execute the adaptive grid low-frequency load shedding method based on three-level coordination of transmission, distribution and utilization, as described in Embodiment 1 or any possible implementation thereof.

[0112] Specifically, a processor can be implemented in hardware or software. When implemented in hardware, the processor can be a logic circuit, an integrated circuit, etc. When implemented in software, the processor can be a general-purpose processor that reads software code stored in memory. This memory can be integrated into the processor or located outside the processor and exist independently.

[0113] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable system. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0114] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing systems. They can be centralized on a single computing system or distributed across a network of multiple computing systems. Optionally, they can be implemented using program code executable by a computing system, thereby storing them in a storage system for execution by the computing system. In some cases, the steps shown or described can be performed in a different order than those presented herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0115] Although the present invention has been described in detail above with general descriptions and specific embodiments, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention fall within the scope of protection claimed by the present invention.

Claims

1. An adaptive low-frequency load shedding method for power grids based on three-level coordination of transmission, distribution, and utilization, characterized in that, include: By deploying synchronous phasor measurement units and wide-area measurement terminals at key nodes of the power transmission network, a wide-area measurement network covering the entire power transmission network is constructed. When a low-frequency over-limit is detected in the power system, synchronous operation data of key nodes of the entire power grid is collected through the wide-area measurement network; based on the synchronous operation data, a multi-machine equivalent model is constructed; based on the multi-machine equivalent model, the system power balance equation, the rotor motion equation, and the system frequency response model considering the load regulation effect, the initial total power deficit of the power system is calculated. Based on the synchronous operation data, the system low-frequency response curve is plotted in real time; the real-time frequency change rate of the low-frequency response curve is calculated using a numerical differentiation algorithm; a threshold range for the frequency change rate is set based on the proportion of new energy access; and it is determined whether the real-time frequency change rate exceeds the threshold range. If the value is exceeded, a power deficit correction mechanism is activated to dynamically correct the initial total power deficit and obtain an accurate power deficit value. If the initial total power deficit is not exceeded, then the accurate power deficit value is taken as the initial total power deficit value. Based on the accurate power deficit value, combined with the load scale, power supply reliability level and grid security constraints of each distribution network, the total load reduction task is decomposed into the load reduction amount of each distribution network and load reduction instructions are generated through the load balancing allocation algorithm; the load reduction instructions are sent to the dispatch centers of each distribution network through the transmission and distribution network collaborative communication link; based on the load characteristics of each transformer substation in the jurisdiction, the distribution network dispatch center decomposes the load reduction amount of each region into specific load reduction instructions for each transformer substation and sends them to the IoT control terminal of the corresponding substation. Based on load identification algorithms and load importance levels, IoT control terminals screen and disconnect non-critical loads with small frequency regulation effects; the actual load disconnection amount is collected and fed back to the distribution network dispatch center and transmission network dispatch center through IoT communication links; The power grid dispatch center summarizes the actual load shedding amount of each distribution area to obtain the cumulative load reduction of the system; based on the real-time frequency data of the system in the synchronous operation data, it determines whether the system frequency has recovered to the safe threshold. If the load reduction process has been completed, then terminate the load reduction process. If the load is below the safety threshold and the maximum load shedding cycle has not been reached, the power deficit estimation and correction process is iteratively performed based on the slope of the current system low-frequency response curve and the remaining power deficit until the system frequency recovers to the safe range or the full load shedding cycle is completed.

2. The adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization as described in claim 1, is characterized in that, The key nodes of the power transmission network include: power plant busbars, regional interconnection lines, and load center busbars; the synchronous operation data include: voltage, current, frequency, phase angle, unit output, and total load.

3. The adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization as described in claim 2, is characterized in that, In the process of calculating the initial total power deficit of the power system, the core characteristic parameters of the source and load are obtained by calculating the system frequency response model that takes into account the load regulation effect; the core characteristic parameters of the source and load are integrated into the multi-machine equivalent model to obtain the comprehensive equivalent parameters of the system; based on the comprehensive equivalent parameters of the system, the observed value of the power deficit is calculated by the rotor motion equation of the system inertial center; the observed value of the power deficit is decomposed and the compensation amount is deducted by the system power balance equation to obtain the initial total power deficit of the power system. The time-domain expression of the system frequency response model that takes into account load regulation effects is as follows: ； In the formula, The system frequency deviation at time t; K is the load voltage regulation effect coefficient; K is the unit primary frequency regulation effect coefficient. This refers to the system voltage deviation. The power deficit is the observed value; D is the system damping coefficient; This is the load frequency regulation effect coefficient; ξ is the amplitude coefficient of the transient oscillation component; ξ is the damping ratio; The undamped natural angular frequency of the system; This is the damped oscillation angular frequency of the system; The initial phase angle of the transient oscillation component; The expression for the system power balance equation is: ； In the formula, These are power deficit observations; The power deficit is caused by a fault. This is for primary frequency modulation power compensation; Power compensation due to load frequency regulation effect; This is power compensation caused by the load voltage regulation effect; This represents the coupling amount between frequency and voltage; The expression for the rotor motion equation is as follows: ； In the formula, The equivalent inertial time constant of the generator; This is the system's rated frequency.

4. The adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization as described in claim 3, is characterized in that, In the process of calculating the real-time frequency change rate of the low-frequency response curve using the numerical differentiation algorithm, the initial value calculation formula for the real-time frequency change rate is as follows: ； In the formula, ROCOF is the rate of change of the initial frequency of the system; This represents the system voltage deviation.

5. The adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization as described in claim 4, is characterized in that, In the process of screening and removing non-critical loads with small frequency regulation effects based on the load identification algorithm and load importance level, the magnitude of the load's frequency regulation effect is determined by calculating the load's frequency regulation effect coefficient. The formula for calculating the frequency modulation effect coefficient is as follows: ； ； In the formula, This is the load frequency regulation effect coefficient; This represents the actual power consumed by the load. Let be the actual active power of the i-th load node; The index of the set of load nodes; Let be the rated power of node i at the rated frequency; Let be the real-time power of node i; , , … The load percentage is proportional to the frequency raised to the power of 0, 1, 2...n. This is the per-unit value for frequency.

6. The adaptive power grid low-frequency load shedding method based on three-level coordination of transmission, distribution, and utilization as described in claim 5, is characterized in that, The formula for calculating the cumulative load reduction of the system is as follows: ； In the formula, This represents the cumulative load reduction of the system. For accurate power deficit values; The target recovery frequency for the system; This represents the number of load nodes. Let be the rated active power of the i-th load node; The amount of active power released to compensate for other regulatory measures.

7. An adaptive power grid low-frequency load shedding device based on three-level coordination of transmission and distribution, employing the adaptive power grid low-frequency load shedding method based on three-level coordination of transmission and distribution as described in any one of claims 1-6, characterized in that, include: The initial total power deficit calculation module is used to construct a wide-area measurement network covering the entire power transmission network by deploying synchronous phasor measurement units and wide-area measurement terminals at key nodes of the power transmission network. When a low-frequency over-limit is detected in the power system, synchronous operation data of key nodes of the entire power grid is collected through the wide-area measurement network; based on the synchronous operation data, a multi-machine equivalent model is constructed; based on the multi-machine equivalent model, the system power balance equation, the rotor motion equation, and the system frequency response model considering the load regulation effect, the initial total power deficit of the power system is calculated. The accurate power deficit value acquisition module is used to plot the system low-frequency response curve in real time based on the synchronous operation data; calculate the real-time frequency change rate of the low-frequency response curve through a numerical differentiation algorithm; set a threshold range for the frequency change rate based on the proportion of new energy access; and determine whether the real-time frequency change rate exceeds the threshold range. If the value is exceeded, a power deficit correction mechanism is activated to dynamically correct the initial total power deficit and obtain an accurate power deficit value. If the initial total power deficit is not exceeded, then the accurate power deficit value is taken as the initial total power deficit value. The load reduction instruction generation and distribution module is used to decompose the total load reduction task into the regional load reduction amount of each distribution network based on the accurate power deficit value, combined with the load scale, power supply reliability level and grid security constraints of each distribution network, and generate load reduction instructions through a load balancing allocation algorithm; the load reduction instructions are sent to the dispatch centers of each distribution network through the transmission and distribution network collaborative communication link; based on the load characteristics of each transformer substation in the jurisdiction, the distribution network dispatch center decomposes the regional load reduction amount into the specific load reduction instructions of each transformer substation and sends them to the IoT control terminal of the corresponding transformer substation. The IoT control terminal load shedding and feedback module is used to screen and shed non-critical loads with small frequency regulation effects based on load identification algorithms and load importance levels; the actual load shedding amount is collected and fed back to the distribution network dispatch center and the transmission network dispatch center through the IoT communication link. The load iteration shedding module is used by the power grid dispatch center to summarize the actual load shedding amount of each distribution area to obtain the cumulative load reduction of the system; and to determine whether the system frequency has recovered to the safe threshold based on the real-time frequency data of the system in the synchronous operation data. If the load reduction process has been completed, then terminate the load reduction process. If the load is below the safety threshold and the maximum load shedding cycle has not been reached, the power deficit estimation and correction process is iteratively performed based on the slope of the current system low-frequency response curve and the remaining power deficit until the system frequency recovers to the safe range or the full load shedding cycle is completed.

8. The adaptive power grid low-frequency load shedding device based on three-level coordination of transmission, distribution, and application as described in claim 7, characterized in that, In the initial total power deficit calculation module, the key nodes of the transmission network include: power plant busbars, regional tie lines, and load center busbars; the synchronous operation data include: voltage, current, frequency, phase angle, unit output, and total load.

9. The adaptive power grid low-frequency load shedding device based on three-level coordination of transmission and distribution as described in claim 8, characterized in that, In the initial total power deficit calculation module, during the process of calculating the initial total power deficit of the power system, the source-load core characteristic parameters are calculated using the system frequency response model that takes load regulation effects into account; the source-load core characteristic parameters are integrated into the multi-machine equivalent model to obtain the system comprehensive equivalent parameters; based on the system comprehensive equivalent parameters, the observed power deficit value is calculated using the rotor motion equation of the system inertial center; the observed power deficit value is decomposed and compensated using the system power balance equation to obtain the initial total power deficit of the power system. The time-domain expression of the system frequency response model that takes into account load regulation effects is as follows: ； In the formula, The system frequency deviation at time t; K is the load voltage regulation effect coefficient; K is the unit primary frequency regulation effect coefficient. This refers to the system voltage deviation. The power deficit is the observed value; D is the system damping coefficient; This is the load frequency regulation effect coefficient; ξ is the amplitude coefficient of the transient oscillation component; ξ is the damping ratio; The undamped natural angular frequency of the system; This is the damped oscillation angular frequency of the system; The initial phase angle of the transient oscillation component; The expression for the system power balance equation is: ； In the formula, These are power deficit observations; The power deficit is caused by a fault. This is for primary frequency modulation power compensation; Power compensation due to load frequency regulation effect; This is power compensation caused by the load voltage regulation effect; This represents the coupling amount between frequency and voltage; The expression for the rotor motion equation is as follows: ； In the formula, The equivalent inertial time constant of the generator; This is the system's rated frequency.

10. The adaptive power grid low-frequency load shedding device based on three-level coordination of transmission and distribution as described in claim 9, characterized in that, In the accurate power deficit value acquisition module, during the process of calculating the real-time frequency change rate of the low-frequency response curve using the numerical differentiation algorithm, the initial value calculation formula for the real-time frequency change rate is as follows: ； In the formula, ROCOF is the rate of change of the initial frequency of the system; This refers to the system voltage deviation. In the load shedding and feedback module of the IoT control terminal, during the process of screening and shedding non-important loads with small frequency regulation effects based on the load identification algorithm and load importance level, the magnitude of the load's frequency regulation effect is determined by calculating the load's frequency regulation effect coefficient. The formula for calculating the frequency modulation effect coefficient is as follows: ； ； In the formula, This is the load frequency regulation effect coefficient; This represents the actual power consumed by the load. Let L be the actual active power of the i-th load node; L is the index of the load node set. Let be the rated power of node i at the rated frequency; Let be the real-time power of node i; , , … The load percentage is proportional to the frequency raised to the power of 0, 1, 2...n. Per-unit frequency; In the load iteration and shelving module, the formula for calculating the cumulative load reduction of the system is as follows: ； In the formula, This represents the cumulative load reduction of the system. For accurate power deficit values; M represents the system target recovery frequency; M represents the number of load nodes. Let be the rated active power of the i-th load node; The amount of active power released to compensate for other regulatory measures.

Images