A new energy station cluster broadband oscillation monitoring control method and device

Abstract

This invention relates to the field of new energy power plant control technology, specifically a method and device for monitoring and controlling broadband oscillations in a new energy power plant cluster. The method includes: synchronizing the clock signal in the new energy power plant cluster based on a satellite synchronous clock signal; constructing a master-slave distributed architecture for each station, where each slave processes the acquired sampling data to obtain the oscillation power and frequency of the corresponding branch, and sends the processed data to the corresponding master; each master processes the acquired data to determine whether the broadband oscillation power component of each slave reaches the corresponding preset threshold; if so, it measures the impedance of the corresponding branch of the target slave at the broadband oscillation frequency in real time and calculates the corresponding total impedance of the target system; based on the total impedance of the target system, it performs stable control of the corresponding branch through a hierarchical switching control strategy. This invention relies on the BeiDou satellite synchronous clock to achieve wide-area synchronous sampling and uses a hierarchical approach with minimal cut-off to suppress broadband oscillations in the new energy power plant cluster.

Description

Technical Field

[0001] This invention relates to the field of new energy power station control technology, specifically to a broadband oscillation monitoring and control method and device for new energy power station clusters. Background Technology

[0002] With the continuous increase in the penetration rate of new energy sources and the level of power electronics, the characteristics of power systems are undergoing profound changes. The electromagnetic oscillation process dominated by power electronic equipment control is becoming increasingly prominent, and it is intertwined with electromechanical transient processes, leading to many new stability problems in the dynamic process of the power grid. Among them, the broadband oscillation problem caused by the interaction between power electronic equipment and the power grid is particularly prominent, and has become a key factor restricting the safe and stable operation of large-scale new energy grid-connected power systems.

[0003] The broadband oscillations of the new power system characterized by large-scale grid integration of renewable energy ("dual-high") exhibited by high proportions of renewable energy and high proportions of power electronic equipment are distinctly different from the single-mode local oscillations of traditional power systems. Spatially, they involve multiple regions and numerous electrical devices, with interconnected influences between these devices, complicating the problem. Frequency-wise, they cover a range from a few hertz to several kilohertz, encompassing an extremely wide frequency band. In nature, they are divergent oscillations induced by negative damping, exhibiting multi-mode, amplitude-frequency-time-varying, and globally complex characteristics. Traditional control methods, relying on local monitoring and single strategies, are ill-suited to the complex scenarios of broadband, multi-mode oscillations. Faced with such complex and variable broadband oscillations, they cannot effectively guarantee the stable operation of the power system.

[0004] Existing technologies face several challenges in addressing broadband oscillations: (1) Time synchronization and sampling accuracy issues. Traditional monitoring devices lack sufficient accuracy when synchronously sampling over a wide area, leading to distorted oscillation feature extraction. Accurate extraction of oscillation features is fundamental to analyzing and controlling broadband oscillations; insufficient sampling accuracy severely impacts the accuracy of subsequent analysis and control strategies, significantly hindering effective solutions to broadband oscillations. (2) Limitations of control strategies. Existing control strategies often rely on damping compensation or parameter adjustments for specific frequency bands, failing to dynamically adapt to time-varying oscillation frequencies. In actual power systems, oscillation frequencies change with time, operating conditions, and other factors; traditional control strategies cannot adjust in a timely manner, resulting in suboptimal control performance. (3) Response speed and architecture issues. Centralized control architectures are used in some existing technologies, but they have significant drawbacks. This architecture requires substantial data transmission and processing, leading to decision delays. Rapid suppression is crucial in the early stages of broadband oscillation divergence; decision delays cause control measures to miss their optimal timing, making it difficult to effectively control oscillation development. While some scholars have attempted to solve power system oscillation problems, limitations exist. For example, although existing technologies propose suppression or identification methods based on local detection, they lack consideration of the overall characteristics of broadband oscillations and do not address effective countermeasures such as how to cut off corresponding lines based on detected broadband oscillations, making it difficult to solve the complex broadband oscillation problem brought about by large-scale grid connection of new energy sources. Summary of the Invention

[0005] The purpose of this invention is to provide a broadband oscillation monitoring and control method and device for new energy power station clusters in order to solve at least one of the above-mentioned technical problems. This invention has strong adaptability, high accuracy, and high reliability, and can perform effective prevention and control measures such as corresponding line disconnection based on the detected broadband oscillations to ensure the safe and stable operation of the power system.

[0006] The present invention achieves the above objectives through the following technical solutions:

[0007] A method for monitoring and controlling broadband oscillations in a cluster of new energy power plants includes:

[0008] Synchronize the clock signals in the new energy power station cluster based on satellite synchronous clock signals;

[0009] A master-slave distributed architecture is constructed for each site, consisting of one master and multiple slaves; wherein, the slaves are used to collect voltage and current signals of each electrical interval, and the master is used to make decisions and judgments based on the data obtained from each slave to control the operation of each branch within the site;

[0010] Each slave device processes the acquired sampling data to obtain the oscillation power and oscillation frequency of the corresponding branch, and sends the processed data to the corresponding master device.

[0011] Each host processes the acquired data and determines whether the broadband oscillation power component of each slave device reaches the corresponding preset threshold.

[0012] When the target host determines that the broadband oscillation power component of its corresponding target slave reaches a preset threshold, the impedance of the corresponding branch of the target slave at the broadband oscillation frequency is measured in real time and the corresponding total impedance of the target system is calculated.

[0013] Based on the total impedance of the target system, the corresponding branches are stabilized using a graded switching control strategy.

[0014] Furthermore, synchronizing the clock signals in the new energy power station cluster based on satellite synchronous clock signals includes:

[0015] The local satellite clock crystal oscillator is tamed by satellite pulses to ensure clock signal synchronization at each station;

[0016] When the external reference signal is interrupted, the system switches to the local satellite clock to provide the time reference signal.

[0017] Furthermore, the slave device generates a synchronization sampling pulse with a time deviation of ≤200ns using the 1PPS signal of the satellite clock as a reference point;

[0018] The voltage and current of each electrical bay within a wide area are sampled synchronously, and the sampling data is transmitted through a collaborative mechanism between FPGA and CPU.

[0019] Furthermore, each slave device processes the acquired sampling data to obtain the oscillation power and oscillation frequency of the corresponding branch, including:

[0020] Each slave device uses the instantaneous values ​​of three-phase voltage and three-phase current as inputs to calculate the instantaneous power, and performs real-time FFT spectrum analysis on the instantaneous power to obtain the oscillation amplitude and oscillation frequency of the broadband oscillation dominant component;

[0021] The real-time FFT uses 1s continuous sampling points as the data window, with a fixed number of points of 1024 and a frequency resolution of 1Hz. Linear interpolation is used to correct the peak frequency of the main lobe.

[0022] Furthermore, each host processes the acquired data, including:

[0023] The host normalizes the amplitude and frequency of the broadband oscillation dominant component sent by each slave device, and calculates the broadband oscillation power of each slave device based on the normalized oscillation frequency and amplitude.

[0024] Further, it is determined whether the broadband oscillation power component of each slave device reaches the corresponding preset threshold, including:

[0025] Each station sets a corresponding broadband oscillation power threshold for each branch based on the actual application scenario, which serves as its corresponding preset threshold.

[0026] Furthermore, the control method further includes:

[0027] When the target host determines that the broadband oscillation power component of its corresponding target slave device reaches a preset threshold, the target host sends a transient recording command to the target slave device, and the target slave device records the original sampling information of the current segment.

[0028] Furthermore, the total impedance of the target system is the sum of the impedances of each branch in the field station corresponding to the slave unit where the broadband oscillation power component reaches a preset threshold.

[0029] Furthermore, based on the total impedance of the target system, the corresponding branches are stabilized using a tiered switching control strategy, including:

[0030] The target slave device's corresponding branches are sorted from largest to smallest according to the oscillation amplitude;

[0031] Determine whether the planned number of branches to be cut exceeds the maximum number of branches threshold. If it does, cut the branch corresponding to the target slave according to the maximum number of branches threshold.

[0032] If the number of planned branch cuts is not exceeded, the number of branch cuts is gradually increased according to the order. When a branch is added, the total impedance of the target system under the corresponding cut-off scheme is calculated, and it is determined whether the total impedance of the target system satisfies that the real part is greater than zero.

[0033] If the conditions are met, stop further attempts to remove more branches.

[0034] A broadband oscillation monitoring and control device for a cluster of new energy power plants includes:

[0035] The clock synchronization module is used to synchronize the clock signals in the new energy power station cluster based on the satellite synchronization clock signal;

[0036] The architecture building module is used to build a master-slave distributed architecture for each site, which includes one master and multiple slaves; wherein, the slaves are used to collect voltage and current signals of each electrical interval, and the master is used to make decisions and judgments based on the data obtained from each slave to control the operation of each branch within the site;

[0037] The signal acquisition module is used by each slave device to process the acquired sampling data to obtain the oscillation power and oscillation frequency of the corresponding branch, and then send the processed data to the corresponding master device.

[0038] The oscillation judgment module is used by each host to process the acquired data and determine whether the broadband oscillation power component of each slave reaches the corresponding preset threshold.

[0039] The oscillation control module is used to measure the impedance of the corresponding branch of the target slave at the wideband oscillation frequency in real time and calculate the corresponding total impedance of the target system when the target host determines that the wideband oscillation power component of its corresponding target slave reaches a preset threshold.

[0040] The oscillation control module is also used to perform stable control on the corresponding branches based on the total impedance of the target system using a graded switching control strategy.

[0041] The beneficial effects of this invention are as follows:

[0042] This invention achieves wide-area, precise synchronous sampling through a master-slave distributed architecture, accurately extracts full-frequency band oscillation characteristics using FFT analysis, and employs a tiered switching strategy to quickly and effectively suppress wide-frequency oscillations. This invention overcomes the bottlenecks of slow response and limited frequency band coverage inherent in traditional control strategies, making it particularly suitable for large-scale renewable energy grid integration scenarios. It demonstrates significant advantages in high-precision measurement, rapid response, and full-frequency band monitoring and control, effectively ensuring the safe and stable operation of the power system. Attached Figure Description

[0043] Figure 1 This is a flowchart of a broadband oscillation monitoring and control method for a new energy power station cluster according to one embodiment of the present invention;

[0044] Figure 2 This is a flowchart of a broadband oscillation monitoring and control method for a new energy power station cluster according to another embodiment of the present invention;

[0045] Figure 3 This is a flowchart illustrating the data transmission process between an FPGA and a CPU according to one embodiment of the present invention.

[0046] Figure 4 This is an FPGA timing diagram according to one embodiment of the present invention;

[0047] Figure 5 This is a flowchart of a slave oscillation component calculation algorithm according to one embodiment of the present invention;

[0048] Figure 6 This is a schematic diagram illustrating the relationship between oscillation power and frequency according to one embodiment of the present invention;

[0049] Figure 7 A flowchart illustrating the application of this invention to a certain new energy power station cluster;

[0050] Figure 8 This is a schematic diagram of the structure of a broadband oscillation monitoring and control device for a new energy power station cluster according to one embodiment of the present invention;

[0051] Figure 9 This is a schematic diagram of the system structure for applying the present invention to a certain new energy power station cluster. Detailed Implementation

[0052] The invention will now be discussed with reference to exemplary embodiments. It should be understood that the described embodiments are merely intended to enable those skilled in the art to better understand and thus implement the invention, and are not intended to imply any limitation on the scope of the invention.

[0053] As used herein, the term "comprising" and its variations are to be interpreted as open-ended terms meaning "including but not limited to". The term "based on" is to be interpreted as "at least partially based on". The terms "one embodiment" and "an embodiment" are to be interpreted as "at least one embodiment".

[0054] Example 1, Figure 1 This is a flowchart of a broadband oscillation monitoring and control method for a new energy power station cluster according to one embodiment of the present invention. Figure 1 As shown, according to one embodiment of the present invention, a broadband oscillation monitoring and control method for a new energy power station cluster includes the following steps:

[0055] Step S102: Synchronize the clock signal in the new energy power station cluster based on the satellite synchronization clock signal;

[0056] Step S104: Construct a master-slave distributed architecture for each site, which includes one master and multiple slaves; wherein, the slaves are used to collect voltage and current signals of each electrical interval, and the master is used to make decisions and judgments based on the data obtained from each slave to control the operation of each branch within the site.

[0057] Step S106: Each slave device processes the acquired sampling data to obtain the oscillation power and oscillation frequency of the corresponding branch, and sends the processed data to the corresponding master device.

[0058] Step S108: Each host processes the acquired data and determines whether the broadband oscillation power component of each slave device reaches the corresponding preset threshold.

[0059] Step S110: When the target host determines that the broadband oscillation power component of its corresponding target slave reaches the preset threshold, the impedance of the corresponding branch of the target slave at the broadband oscillation frequency is measured in real time and the corresponding total impedance of the target system is calculated.

[0060] Step S112: Based on the total impedance of the target system, the corresponding branches are stabilized using a graded switching control strategy.

[0061] This embodiment proposes a broadband oscillation monitoring and control method for new energy power station clusters. It utilizes the 1PPS signal from the BeiDou satellite synchronous clock to discipline the crystal oscillators across the entire network, achieving 100ns-level wide-area synchronous sampling and providing a time reference for phase consistency measurement of the 5Hz–2.5kHz broadband oscillation. At the power station level, a distributed architecture of "one master and multiple slaves" is adopted. The slaves, through FPGA+CPU collaboration, collect voltage and current data from each electrical interval at a rate of 12.8kHz, completing 1024-point FFT every 1 second. Instantaneous power is analyzed by spectrum to extract oscillation amplitude and frequency, which are then uploaded to the master via a semaphore mechanism. Upon receiving the data, the master first processes it according to branch rates. The system is normalized to a fixed capacity and then compared with its preset broadband oscillation power threshold. If a branch exceeds the limit, transient recording is immediately initiated and the impedance of that branch at the oscillation frequency is calculated in real time. The total system impedance is obtained by summing the grid-side impedance. If the real part of the total impedance is negative, it indicates a risk of negative damping divergence. In this case, the wind farm feeders are cut off in rounds according to the oscillation amplitude from large to small. Only one branch is added in each round and the total impedance is recalculated. The tripping is stopped when the real part is positive for the first time. If the frequency shifts due to the tripping and the real part becomes negative again, the tripping is iteratively added at the new frequency until the oscillation is suppressed. The whole process achieves fast, accurate and stable control of the broadband oscillation of the regional power grid with the minimum amount of tripping.

[0062] This invention relies on the BeiDou satellite synchronous clock to achieve wide-area synchronous sampling, adopts a "one master and multiple slaves" distributed architecture to quickly extract broadband oscillation parameters, and uses a hierarchical switching control strategy to achieve rapid, accurate and stable suppression of broadband oscillations in new energy power plant clusters with minimal switching amount.

[0063] According to one embodiment of the present invention, step S102 includes:

[0064] The local satellite clock crystal oscillator is tamed by satellite pulses to ensure clock signal synchronization at each station;

[0065] When the external reference signal is interrupted, the system switches to the local satellite clock to provide the time reference signal.

[0066] In this embodiment, step S102 continuously outputs 1PPS pulses through the BeiDou satellite receiving module to perform sliding average training on the local isothermal crystal oscillator, lock the frequency deviation, and keep the time error between stations within 100ns~200ns. When the external reference is interrupted due to weather or obstruction, the system automatically and seamlessly switches to the trained local satellite clock to continue providing a time reference of the same accuracy, ensuring that the phase drift of the wide-area sampling pulse edge is much smaller than the phase error, and providing a continuous, reliable, microsecond-level synchronized time frame for subsequent broadband oscillation feature extraction and control decisions.

[0067] This invention uses the 1PPS pulse from the BeiDou satellite to discipline the local crystal oscillator and achieve seamless switching when the external reference is interrupted, providing a continuous and reliable synchronous time reference for broadband oscillation monitoring and control.

[0068] According to one embodiment of the present invention, the slave device generates a synchronization sampling pulse with a time deviation of ≤200ns using the 1PPS signal of the satellite clock as a reference point;

[0069] The voltage and current of each electrical bay within a wide area are sampled synchronously, and the sampling data is transmitted through a collaborative mechanism between FPGA and CPU.

[0070] In this embodiment, the slave device uses the rising edge of the BeiDou satellite clock (1PPS) as the absolute reference. After digital phase shifting and PLL frequency multiplication, it generates a 12.8kHz sampling pulse with edge jitter ≤200ns, ensuring that the voltage and current of each electrical interval in the wide area are latched at the same time. The FPGA completes oversampling and recursive DFT at 256 points / 20ms. Every 10ms, it pushes 128 original sampled values ​​and 32 spectral data to the CPU via interrupt. The CPU reads and packages the data through dual-port RAM with zero copy, realizing full-link microsecond-level synchronization of sampling-calculation-upload, ensuring that the phase consistency required for subsequent broadband oscillation analysis is not destroyed by transmission delay.

[0071] This invention generates synchronous sampling pulses with edge jitter ≤200ns based on the BeiDou satellite clock 1PPS. Through the collaboration of FPGA and CPU, it realizes microsecond-level synchronous acquisition and zero-copy transmission of wide-area electrical signals, ensuring phase consistency of broadband oscillation analysis.

[0072] According to one embodiment of the present invention, step S106 includes:

[0073] Each slave device uses the instantaneous values ​​of three-phase voltage and three-phase current as inputs to calculate the instantaneous power, and performs real-time FFT spectrum analysis on the instantaneous power to obtain the oscillation amplitude and oscillation frequency of the dominant broadband oscillation component;

[0074] The real-time FFT uses 1 second of continuous sampling points as the data window, with a fixed number of points of 1024 and a frequency resolution of 1Hz. Linear interpolation is used to correct the peak frequency of the main lobe.

[0075] In this embodiment, the slave device calculates the instantaneous power of the three-phase voltage and current instantaneous values ​​using a 1-second gapless data window (1024 points), and obtains a 1Hz resolution spectrum through fixed-point FFT. To address the main lobe energy leakage, linear interpolation is used to correct the peak frequency and its adjacent spectral lines, accurately extracting the frequency and amplitude of the dominant oscillation component in the range of 5-2500Hz, providing highly reliable original characteristic quantities for subsequent normalization threshold judgment and impedance calculation.

[0076] This invention uses a fixed 1024-point FFT and linear interpolation correction to accurately extract the amplitude and frequency of a 5-2500Hz wideband oscillation component, providing highly reliable characteristic parameters for subsequent control.

[0077] According to one embodiment of the present invention, in step S108, each host processes the acquired data, including:

[0078] The host normalizes the amplitude and frequency of the broadband oscillation dominant component sent by each slave device, and calculates the broadband oscillation power of each slave device based on the normalized oscillation frequency and amplitude.

[0079] Preferably, determining whether the broadband oscillation power component of each slave device reaches the corresponding preset threshold includes:

[0080] Each station sets a corresponding broadband oscillation power threshold for each branch based on the actual application scenario, which serves as its corresponding preset threshold.

[0081] In this embodiment, after receiving the oscillation amplitude and frequency from each slave unit, the master unit first normalizes the amplitude based on the rated capacity of the corresponding branch, and then calculates the broadband oscillation power of each slave unit based on the normalized oscillation frequency and amplitude. The station side independently sets a broadband oscillation power threshold for each branch based on the line damping capacity, convergence impedance, and dispatching procedures. Subsequent impedance measurement and tripping procedures are only triggered when the broadband oscillation power component of a branch continuously exceeds the limit, thereby avoiding insufficient sensitivity or false triggering and achieving accurate classification and identification of oscillation risk.

[0082] This invention achieves accurate calculation and hierarchical discrimination of broadband oscillation power by normalizing the dominant component of the branch broadband oscillation and setting an independent threshold, effectively avoiding false triggering and improving the reliability of oscillation risk identification.

[0083] According to one embodiment of the present invention, the control method further includes:

[0084] When the target host determines that the broadband oscillation power component of its corresponding target slave device reaches a preset threshold, the target host sends a transient recording command to the target slave device, and the target slave device records the original sampling information of the current segment.

[0085] In this embodiment, when the target host determines that the normalized broadband oscillation power of a certain slave branch continues to exceed the limit, it immediately sends a "transient recording command" to the slave through a GOOSE message at the station control layer. After receiving the command, the slave packs and freezes the current 1s original sampling points (256 points / 20ms, a total of 12800 points) in the FPGA buffer along with the time standard, and stores them in the local solid-state disk. This provides distortion-free data for subsequent impedance identification and retains high-precision original data for scheduling accident inversion, while not affecting the continuous execution of real-time FFT calculations.

[0086] This invention triggers the slave device to freeze high-precision raw sampling data via a GOOSE message when the broadband oscillation power exceeds the limit, providing a distortion-free basis for impedance identification and fault inversion without affecting real-time analysis.

[0087] According to one embodiment of the present invention, the total impedance of the target system is the sum of the impedances of each branch in the field station corresponding to the slave unit where the broadband oscillation power component reaches a preset threshold.

[0088] In this embodiment, the total impedance of the target system is composed of the "problem branches" of the field station where the oscillation over-limit slave is located: the master first screens out all branches whose broadband oscillation power components reach the preset threshold, and calculates the branch impedance of each branch at the current oscillation frequency using the synchronous voltage and current phasors, and then obtains the total impedance of the target system by algebraic summation. This is used as the sole input for the subsequent impedance real part criterion, avoiding the mixing of normally operating branches and causing the stability judgment to be "diluted", and ensuring that the switching strategy is accurately applied to the negative damping source.

[0089] This invention precisely locates the negative damping source by screening out-of-limit branches of oscillation and calculating the total impedance of the target system algebraically, thus avoiding interference from normal branches and ensuring that the switching strategy effectively suppresses broadband oscillations.

[0090] According to one embodiment of the present invention, step S112 includes:

[0091] Sort the target slave circuits by oscillation amplitude from largest to smallest;

[0092] Determine whether the planned number of branches to be cut exceeds the maximum number of branches threshold. If it does, cut the corresponding branch of the target slave according to the maximum number of branches threshold.

[0093] If the number of planned branch cuts is not exceeded, the number of branch cuts is gradually increased according to the order. When a branch is added, the total impedance of the target system under the corresponding cut-off scheme is calculated, and it is determined whether the total impedance of the target system satisfies that the real part is greater than zero.

[0094] If the conditions are met, stop further attempts to remove more branches.

[0095] In this embodiment, in step S112, the host first sorts the over-limit branches according to their oscillation power from high to low according to their oscillation amplitude to form a queue to be cut. When increasing the number of cuts step by step, it first checks whether the planned number of cuts in this round exceeds the "maximum number of branches threshold" given by the scheduling procedure. The maximum number of branches threshold is a preset proportion of the total number of branches in the entire station (such as 30%) or the total number of over-limit branches in the station. In this embodiment, the latter is used. If the planned number of cut branches exceeds the maximum number of branches threshold, it will be executed directly according to the upper limit. Otherwise, the total system impedance of the remaining branches will be recalculated for each additional cut. Once the real part of the total system impedance turns from negative to positive, the search will stop. This will quickly eliminate the negative damping with the minimum breaking amount, ensuring that the oscillation is calmed and avoiding the risk of over-cutting.

[0096] This invention achieves rapid positive conversion of the real part of the total system impedance with minimal cut-off by sorting oscillation amplitudes and verifying the maximum number of branches, thus accurately eliminating negative damping and avoiding the risk of over-cutting.

[0097] Example 2, Figure 2 This is a flowchart of a broadband oscillation monitoring and control method for a new energy power station cluster, according to another embodiment of the present invention. Figure 2 As shown, according to one embodiment of the present invention, a broadband oscillation monitoring and control method for a new energy power station cluster includes the following steps:

[0098] Step S201: Wide-area synchronous sampling of the new energy power station cluster is achieved using satellite synchronization clock signals;

[0099] The new energy power station cluster adopts a master-slave distributed architecture, with each station including one master and multiple slaves. The slaves are responsible for collecting voltage and current signals from each electrical bay, processing the collected data and sending it to the master. The master is used to make decisions based on the data obtained from each slave, and control the operation of each branch within the station based on the decision results.

[0100] Step S201 specifically includes:

[0101] Step S2011: Time synchronization of the new energy power station cluster;

[0102] The local BeiDou satellite clock crystal oscillator is tamped by high-precision satellite pulses to ensure that the BeiDou satellite clock can still provide a high-precision time reference signal when the external reference signal is interrupted.

[0103] Step S2012: Based on the actual status of each site, construct a master-slave distributed architecture for each site to collect voltage and current signals from each electrical bay.

[0104] Step S2013: Collect voltage and current signals from each electrical bay via the slave device.

[0105] The slave device generates synchronous sampling pulses using the 1PPS signal of the Beidou satellite clock as a reference point, and synchronously samples the voltage and current of each electrical interval in a wide area. Its accuracy can reach about 100ns~200ns, which meets the requirements of the broadband oscillation monitoring and control device of new energy power station cluster for continuous synchronous sampling. The sampling data transmission is realized through the FPGA and CPU collaborative mechanism.

[0106] Figure 3 This is a flowchart illustrating the FPGA and CPU data transmission process according to one embodiment of the present invention. Figure 3 As shown, the FPGA notifies the CPU to read the original sampling points and the recursive DFT calculation results via an interrupt. The number of sampling points per cycle (20ms) and the number of interrupt points of the FPGA are written to the FPGA register by the CPU during device initialization.

[0107] Since the slave device's sampling rate is 12.8K, which is 256 points per cycle (20ms), performing recursive DFT calculations on every 256 points would be computationally intensive and unnecessary. The accuracy achieved with 48 points of sampling is sufficient. Following an integer multiple relationship, the recursive DFT is performed every 4 sampling points, resulting in a sampling rate of 64 points for the recursive DFT. To maintain compatibility with previous FPGA register settings, a sampling rate of 64 is written to the FPGA registers during device initialization, and the number of sampling points generating interrupts is 32.

[0108] Figure 4 This is a timing diagram of an FPGA according to one embodiment of the present invention. Figure 4 As shown, the FPGA internally implements a sample rate of 256 * 4, with the recursive DFT calculated at 64 points. The FPGA assigns a sequence number from 0 to 12799 to the sample points every second. The FPGA generates an interrupt every 10ms to notify the read function. The CPU's read function executes once every 10ms, reading 128 sample points and 32 recursive DFT calculation results.

[0109] Step S202: The slave device performs instantaneous power calculation, filtering, and FFT analysis on the sampled data to extract the oscillation frequency and amplitude; specifically including:

[0110] Step S2021: The slave device calculates the oscillation frequency and oscillation amplitude based on the collected data;

[0111] The slave device uses the instantaneous values ​​of three-phase voltage and three-phase current as inputs to calculate the instantaneous power, and performs real-time FFT spectrum analysis on the instantaneous power to obtain the amplitude and frequency of the broadband oscillation dominant component;

[0112] The slave device performs FFT calculations on the data in whole seconds, with a data window of 1 second and an FFT algorithm number of 1024 (2 to the power of 10). Figure 5This is a flowchart of a slave oscillation component calculation algorithm according to one embodiment of the present invention. Figure 5 As shown, the FFT has a resolution of 1Hz and can identify oscillations from 5 to 2500Hz. Wideband oscillations are calculated once per second, triggered by a semaphore mechanism. The calculated oscillation power and frequency are stored in global variables and sent to the host computer. The oscillation power... Y Calculate using the following formula:

[0113] ;

[0114] In the formula, Y max For maximum power, Y max-1 The power of the adjacent preceding frequency point;

[0115] Figure 6 This is a schematic diagram illustrating the relationship between oscillation power and frequency according to one embodiment of the present invention. Figure 6 As shown, the oscillation frequency Fr Calculated using the following formula:

[0116] Compare Y max+1 and Y max-1 The size, if Y max+1 > Y max-1 ,but:

[0117] ;

[0118] Frequency can be obtained Fr The value; if Y max+1 < Y max-1 Its linear interpolation logic is the same as above.

[0119] The step is S2022, where the slave device sends the oscillation frequency and amplitude to the master device. When the master device sends a transient recording command to the slave device, the slave device records the original sampling information of the current segment.

[0120] Step S203: The host normalizes the oscillation amplitude and determines whether the broadband oscillation power component of each slave device reaches the preset threshold.

[0121] The host normalizes the amplitude and frequency of the dominant broadband oscillation component sent by each slave device, and calculates the broadband oscillation power of each slave device based on the normalized oscillation frequency and amplitude. In combination with the actual application scenario, different sites set corresponding broadband oscillation power thresholds according to their load capacity. The broadband oscillation power of each slave device's corresponding branch is compared with its corresponding broadband oscillation power threshold to determine whether the branch has reached the preset threshold.

[0122] Step S204: When the broadband oscillation power component of the slave unit reaches the preset threshold, measure the impedance of each wind farm feeder of the slave unit at the broadband oscillation frequency in real time and calculate the total system impedance.

[0123] Calculating the system impedance requires obtaining the impedances of the grid side and each wind farm branch. The accuracy of impedance identification directly determines the accuracy of the stability assessment. The total system impedance is the sum of the impedances of each branch in the corresponding wind farm where the broadband oscillation power component reaches a preset threshold.

[0124] Step S205: Stabilize the corresponding branch according to the total impedance value of the system and implement a graded switching control strategy.

[0125] The stability of wideband oscillation at a given frequency can be determined by the sign of the product of the slope of the polymer resistance and the polymer reactance. For new energy grid-connected systems, the reactance slope at the wideband oscillation point is usually positive, so its stability is determined by the sign of the polymer resistance.

[0126] If the system impedance is changed by cutting off part of the wind farm feeder, so that the real part of the system impedance is greater than zero at the wideband oscillation frequency, the wideband oscillation can be eliminated. This is the action criterion for wideband oscillation monitoring and control based on the impedance method.

[0127] Figure 7 This is a flowchart illustrating the application of the present invention to a cluster of new energy power plants. Figure 7 As shown, when a large-scale wind power grid-connected system with N wind farm branches experiences broadband oscillations, and its oscillation component reaches a preset threshold (i.e., assuming that the broadband oscillation power components of N wind farm branches in the system exceed their corresponding preset thresholds), the impedance of each wind farm feeder at the broadband oscillation frequency is measured in real time, and the total system impedance is calculated. The N wind farm branches are sorted from largest to smallest oscillation amplitude. The total system impedance is calculated for each cut-off scheme, from 1 to N wind farm feeders. If the total system impedance when cutting off i feeders in the wind farm combination satisfies a real part greater than zero, it indicates that there is an effective cut-off method to suppress broadband oscillations, and further attempts to cut off more feeders will cease. Theoretically, as long as both positive and negative sequence impedances satisfy a real part greater than zero, the system is stable at that frequency. However, considering the errors in actual measurements and the possibility that cutting off part of the wind farm may cause a shift in the oscillation frequency, if broadband oscillations cannot be eliminated after cutting off wind farms according to the sequence impedance values ​​at the current broadband oscillation frequency, iterative "turbine cutting" cycles can be added to ensure that broadband oscillations are suppressed.

[0128] This invention employs a regional hierarchical interconnection and graded information flow oscillation control strategy. Broadband monitoring and control devices are deployed at key nodes of the regional power grid to optimize data transmission and control links, achieving rapid response to wide-area broadband oscillations. For broadband monitoring and control devices deployed in the substation domain, the invention coordinates the processing of wide-area control commands from remote broadband monitoring and control devices with graded and round-robin precise control commands from local access elements, achieving precise control at the element level and suppressing broadband oscillations with minimal cutoff.

[0129] Example 3, Figure 8 This is a schematic diagram of the structure of a broadband oscillation monitoring and control device for a new energy power station cluster according to one embodiment of the present invention; Figure 9 This is a schematic diagram of the system structure for applying the present invention to a certain new energy power station cluster.

[0130] like Figures 8-9 As shown, according to one embodiment of the present invention, a broadband oscillation monitoring and control device for a new energy power station cluster includes:

[0131] Clock synchronization module 10 is used to synchronize the clock signal in the new energy power station cluster based on the satellite synchronization clock signal;

[0132] Architecture building module 20 is used to build a master-slave distributed architecture for each site, which includes one master and multiple slaves; wherein, the slaves are used to collect voltage and current signals of each electrical interval, and the master is used to make decisions and judgments based on the data obtained from each slave to control the operation of each branch within the site;

[0133] The signal acquisition module 30 is used by each slave device to process the acquired sampling data to obtain the oscillation power and oscillation frequency of the corresponding branch, and send the processed data to the corresponding host.

[0134] The oscillation judgment module 40 is used by each host to process the acquired data and determine whether the broadband oscillation power component of each slave reaches the corresponding preset threshold.

[0135] The oscillation control module 50 is used to measure the impedance of the corresponding branch of the target slave in real time at the wideband oscillation frequency and calculate the corresponding total impedance of the target system when the target host determines that the wideband oscillation power component of its corresponding target slave reaches the preset threshold.

[0136] The oscillation control module 50 is also used to perform stable control on the corresponding branches based on the total impedance of the target system using a graded switching control strategy.

[0137] This embodiment discloses a broadband oscillation monitoring and control device for a new energy power station cluster. The clock synchronization module 10 relies on the 1PPS pulse from the BeiDou satellite to discipline and maintain timekeeping for the global crystal oscillator, outputting a synchronous sampling clock with an accuracy of 100ns~200ns. The architecture construction module 20 deploys a "one master, multiple slaves" distributed architecture at each power station. The slaves are responsible for collecting voltage and current data from each electrical interval, while the master centrally makes decisions and controls the branch circuits at its own station. The signal acquisition module 30 calculates the instantaneous power on the slave side, performs a 1024-point FFT after filtering, and extracts the oscillation power and frequency from 5Hz to 2.5kHz. The data is then sent to the host computer. The oscillation judgment module 40 normalizes the oscillation amplitude received by the host computer according to "instantaneous power fluctuation / component rated power" to obtain the relative amplitude of the wideband oscillation, and compares it with the branch's custom threshold to determine whether it exceeds the limit. The oscillation control module 50 measures the impedance of the branch that exceeds the limit in real time at the oscillation frequency, accumulates it to form the total impedance of the target system, and cuts off the generators in a graded manner according to "maximum amplitude priority": the number of generators to be cut off is increased in round by round. If the planned number exceeds the maximum number of branch numbers, the threshold is used. The real part of the total impedance is checked in each round. Once it turns from negative to positive, the process stops. Otherwise, the generator with the largest oscillation amplitude among the remaining components continues to be cut off until the oscillation subsides. This device has been mapped to the scenario where the sending-end network (conventional hydropower, thermal power, pumped storage, new energy base, "Shago" base) is connected to the receiving-end network (substation, distributed new energy and energy storage power station) through the hub substation, local power grid and source-grid-load-storage integrated area. This device is deployed in each master station and substation / execution station to realize rapid monitoring of wideband oscillation, suppression of minimum generator cut-off amount and stable control of the whole area.

[0138] This invention achieves wide-area synchronous sampling through a master-slave distributed architecture, extracts full-band oscillation features using FFT analysis, and employs a hierarchical switching strategy to achieve rapid suppression of wideband oscillations. It solves the problems of slow response and insufficient frequency band coverage of traditional control strategies, and is suitable for large-scale new energy grid connection scenarios. It has the advantages of high precision, fast response, and full-band coverage.

[0139] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0140] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

[0141] It should be understood that the sequence number of each step in the invention and embodiments of the present invention does not absolutely imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

Claims

1. A method for monitoring and controlling broadband oscillations in a cluster of new energy power plants, characterized in that, include: Synchronize the clock signals in the new energy power station cluster based on satellite synchronous clock signals; A master-slave distributed architecture is constructed for each site, consisting of one master and multiple slaves; wherein the slaves are used to collect voltage and current signals of each electrical interval, and the master is used to make decisions and judgments based on the data obtained from each slave to control the operation of each branch within the site; Each slave device processes the acquired sampling data to obtain the oscillation power and oscillation frequency of the corresponding branch, and sends the processed data to the corresponding master device. Each host processes the acquired data and determines whether the broadband oscillation power component of each slave device reaches the corresponding preset threshold. When the target host determines that the broadband oscillation power component of its corresponding target slave reaches a preset threshold, the impedance of the corresponding branch of the target slave at the broadband oscillation frequency is measured in real time and the corresponding total impedance of the target system is calculated. Based on the total impedance of the target system, the corresponding branches are stabilized using a tiered switching control strategy, including: The target slave device's corresponding branches are sorted from largest to smallest according to the oscillation amplitude; Determine whether the planned number of branches to be cut exceeds the maximum number of branches threshold. If it does, cut the branch corresponding to the target slave according to the maximum number of branches threshold. If the number of planned branch cuts is not exceeded, the number of branch cuts is gradually increased according to the order. When a branch is added, the total impedance of the target system under the corresponding cut-off scheme is calculated, and it is determined whether the total impedance of the target system satisfies that the real part is greater than zero. If the conditions are met, stop further attempts to remove more branches.

2. The broadband oscillation monitoring and control method for new energy power station clusters according to claim 1, characterized in that, Synchronizing clock signals in a cluster of new energy power stations based on satellite synchronous clock signals includes: The local satellite clock crystal oscillator is tamed by satellite pulses to ensure that the clock signals of each station are synchronized; When the external reference signal is interrupted, the system switches to the local satellite clock to provide the time reference signal.

3. The broadband oscillation monitoring and control method for new energy power station clusters according to claim 1, characterized in that: The slave device generates a synchronous sampling pulse with a time deviation of ≤200ns using the 1PPS signal of the satellite clock as a reference point; The voltage and current of each electrical bay within a wide area are sampled synchronously, and the sampling data is transmitted through a collaborative mechanism between FPGA and CPU.

4. The broadband oscillation monitoring and control method for new energy power station clusters according to claim 3, characterized in that, Each slave device processes the acquired sampled data to obtain the oscillation power and oscillation frequency of the corresponding branch, including: Each slave device uses the instantaneous values ​​of three-phase voltage and three-phase current as inputs to calculate the instantaneous power, and performs real-time FFT spectrum analysis on the instantaneous power to obtain the oscillation amplitude and oscillation frequency of the broadband oscillation dominant component; The real-time FFT uses 1s continuous sampling points as the data window, with a fixed number of points of 1024 and a frequency resolution of 1Hz. Linear interpolation is used to correct the peak frequency of the main lobe.

5. The broadband oscillation monitoring and control method for new energy power station clusters according to claim 1, characterized in that, Each host processes the acquired data, including: The host normalizes the amplitude and frequency of the broadband oscillation dominant component sent by each slave device, and calculates the broadband oscillation power of each slave device based on the normalized oscillation frequency and amplitude.

6. The broadband oscillation monitoring and control method for new energy power station clusters according to claim 5, characterized in that, Determine whether the broadband oscillation power component of each slave device reaches the corresponding preset threshold, including: Each station sets a corresponding broadband oscillation power threshold for each branch based on the actual application scenario, which serves as its corresponding preset threshold.

7. The broadband oscillation monitoring and control method for new energy power station clusters according to claim 1, characterized in that, The control method further includes: When the target host determines that the broadband oscillation power component of its corresponding target slave device reaches a preset threshold, the target host sends a transient recording command to the target slave device, and the target slave device records the original sampling information of the current segment.

8. The broadband oscillation monitoring and control method for new energy power station clusters according to claim 1, characterized in that: The total impedance of the target system is the sum of the impedances of each branch in the field station corresponding to the slave unit where the broadband oscillation power component reaches a preset threshold.

9. A broadband oscillation monitoring and control device for a new energy power station cluster, characterized in that, include: The clock synchronization module is used to synchronize the clock signals in the new energy power station cluster based on the satellite synchronization clock signal; The architecture building module is used to build a master-slave distributed architecture for each site, which includes one master and multiple slaves; wherein, the slaves are used to collect voltage and current signals of each electrical interval, and the master is used to make decisions and judgments based on the data obtained from each slave to control the operation of each branch within the site; The signal acquisition module is used by each slave device to process the acquired sampling data to obtain the oscillation power and oscillation frequency of the corresponding branch, and then send the processed data to the corresponding master device. The oscillation judgment module is used by each host to process the acquired data and determine whether the broadband oscillation power component of each slave reaches the corresponding preset threshold. The oscillation control module is used to measure the impedance of the corresponding branch of the target slave at the wideband oscillation frequency in real time and calculate the corresponding total impedance of the target system when the target host determines that the wideband oscillation power component of its corresponding target slave reaches a preset threshold. The oscillation control module is also used to perform stabilization control on the corresponding branches based on the total impedance of the target system using a graded switching control strategy, including: The target slave device's corresponding branches are sorted from largest to smallest according to the oscillation amplitude; Determine whether the planned number of branches to be cut exceeds the maximum number of branches threshold. If it does, cut the branch corresponding to the target slave according to the maximum number of branches threshold. If the number of planned branch cuts is not exceeded, the number of branch cuts is gradually increased according to the order. When a branch is added, the total impedance of the target system under the corresponding cut-off scheme is calculated, and it is determined whether the total impedance of the target system satisfies that the real part is greater than zero. If the conditions are met, stop further attempts to remove more branches.

Images