Cross-system monitoring and tracing method and system for power oscillation of thermal power generating unit

Abstract

The invention discloses a cross-system monitoring and tracing method and system for power oscillation of a thermal power generating unit, and the method comprises the steps: obtaining corresponding physical quantities of a plurality of independent control systems in a monitoring substation by each monitoring substation, gathering the corresponding physical quantities into a channel, and carrying out the preprocessing; constructing a joint matrix based on the plurality of pre-processed channels, performing singular value decomposition to extract a signal subspace, and further solving to obtain an optimal modal parameter and angular frequency; the domain participation degree and the inter-domain average phase difference are calculated based on the optimal modal parameters, the average direction intensity is calculated based on the angular frequency, and then an oscillation source score is constructed; and judging the oscillation source based on the oscillation source score, and uploading the judgment to a monitoring master station to generate a diagnosis result. The problems that a traditional monitoring method is single in data source and cannot accurately trace are solved, cooperative control from early monitoring to accurate traceability is achieved, and the safety and stability of a power grid are improved.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis technology, specifically to a cross-system monitoring and tracing method and system for power oscillation in thermal power units. Background Technology

[0002] With the expansion of the power grid, low-frequency oscillations are becoming increasingly common, especially in thermal power plants. During grid-connected operation, multiple independent systems, such as the thermal system and excitation system, can be disturbed, leading to low-frequency oscillations. These oscillations are typically continuous decaying or oscillating electromechanical oscillation modes within the range of 0.2–2.5 Hz. Failure to suppress these oscillations in a timely manner can result in increased generator power angle fluctuations, abnormal current, and unstable main steam pressure and power output. In severe cases, it can cause unit tripping, endangering grid security. Therefore, real-time monitoring and rapid source tracing of low-frequency oscillations are crucial requirements for the safe and stable operation of power plant units.

[0003] However, traditional power grid dispatching can only observe oscillations in electrical quantities such as line power and frequency, but cannot identify the source of the oscillations. Existing low-frequency oscillation monitoring technologies have limitations in monitoring range. They typically use PMUs (Phasor Measurement Units) or DFRs (Fault Recorders) to identify electrical oscillation modes, such as frequency, power, and power angle oscillation mode analysis. However, the data source for this monitoring technology comes from the electrical measurements of the power grid and cannot reach multiple independent control systems within the power plant unit, such as the turbine regulating system (DEH), boiler coordinated control system (CCS), and excitation regulating system (AVR). This results in insufficient acquisition of dynamic behavior and makes it difficult to determine whether the oscillations are caused by the control systems within the power plant unit. On the other hand, there is the problem of isolated monitoring data from multiple systems. Most power plants use the historical curves and state variables of multiple control systems, such as DCS (Distributed Control System), DEH (Digital Electro-hydraulic Control System), or excitation system, for independent monitoring. However, these systems are isolated from each other and cannot be jointly analyzed, making it impossible to determine whether the low-frequency oscillation source comes from the excitation system, the turbine regulating system, or the coupling between the two. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a cross-system monitoring and tracing method and system for power oscillations in thermal power units, which solves the problem of "information silos" in traditional monitoring systems, which cannot analyze the source of oscillations from a global perspective.

[0005] The first aspect of this invention discloses a cross-system monitoring and source tracing method for power oscillations in thermal power units, comprising: Each monitoring substation acquires the corresponding physical quantities from multiple independent control systems within its station, aggregates the corresponding physical quantities into channels, and performs preprocessing. A joint matrix is ​​constructed based on the preprocessed multiple channels, and singular value decomposition is performed to extract the signal subspace, thereby solving for the optimal modal parameters and angular frequency; The domain participation degree and the average phase difference between domains are calculated based on the optimal modal parameters, and the average directional intensity is calculated based on the angular frequency, thereby constructing an oscillation source score; The oscillation source is determined based on the oscillation source score, and the determination result is uploaded to the monitoring master station to generate a diagnostic result.

[0006] Preferably, the step of constructing a joint matrix based on multiple preprocessed channels and extracting the signal subspace through singular value decomposition, and then solving for the optimal modal parameters and angular frequencies, includes: For each channel, construct a Hankel matrix. Stack all Hankel matrices according to the channel direction to construct a joint matrix. Perform singular value decomposition on the joint matrix and construct a signal subspace. Construct a shift subspace based on the signal subspace. Solve the shift subspace to obtain the modal poles. A Vandermonde matrix is ​​constructed based on the modal poles, and an objective function is constructed using the modal parameters as optimization variables. The optimal modal parameters are obtained by solving the objective function, and the modal poles are mapped to complex modal poles. The angular frequency is obtained by extracting the complex modal poles.

[0007] Preferably, the Vandermonde matrix satisfies the following formula: in, For the first Line number The matrix parameters of the column.

[0008] Preferably, the objective function satisfies the following formula: in, For modal parameters, The observed data matrix obtained through preprocessing, The Vandermonde matrix, For all the modal parameters of the first One modality, , The regularization coefficient is . This is a physical prior.

[0009] Preferably, the physical prior terms satisfy the following formula: in, The modal parameters are the first The mode in the th ... Phase on each channel; The modal parameters are the first The mode in the th ... Phase on each channel For the preset first The first channel and the first The expected phase difference between the channels This is the tolerance threshold. For all channel pairs that require phase constraints A set of.

[0010] Preferably, the step of calculating the domain participation degree and the average phase difference between domains based on the optimal modal parameters, calculating the average directional intensity based on the angular frequency, and then constructing the oscillation source score includes: The oscillation amplitude and oscillation phase are extracted based on the optimal modal parameters. The domain participation of multiple independent systems is calculated based on the oscillation amplitude. The average phase difference between domains is calculated using the oscillation phase. The directional transfer function of all channels is obtained through a vector autoregression model, and the average directional intensity is calculated by combining the angular frequency. Independent oscillation source scores and coupled oscillation source scores are calculated based on domain participation, average phase difference between domains, and average directional intensity.

[0011] Preferably, the domain participation degree satisfies the following formula: in, In functional domain The Middle Domain participation of each modality, the functional domain For the corresponding independent control system, The first optimal modal parameter The mode in the th ... The oscillation amplitude on each channel In functional domain All channels.

[0012] Preferably, the average phase difference between the domains satisfies the following formula: in, For the first functional domain Second functional domain In the Inter-domain average phase difference across modes For the first functional domain In the Domain-averaged phase over each mode For the second functional domain In the Domain-averaged phase over each mode.

[0013] Preferably, the domain-averaged phase satisfies the following formula: in, The optimal modal parameters are the first... The mode in the th ... The oscillation phase on each channel.

[0014] Preferably, the direction transfer function satisfies the following formula: in, For the corresponding angular frequency Next The first channel to the first The direction transfer function value for each channel; To derive the first... The first channel to the first The corresponding transmitted value for each channel; The normalization factor derived from the vector autoregressive model represents all channels up to the 1st... Total transmission level of each channel.

[0015] Preferably, the average directional intensity satisfies the following formula: Among them, the To the corresponding angular frequency First functional domain To the second functional domain The average directional intensity, This represents the number of channels for the corresponding functional domain.

[0016] Preferably, the score of the coupled oscillation source satisfies the following formula: in, , , The coupling scoring weight coefficients are: X is the required scoring functional domain one, T is the required scoring functional domain two, and E is the preset reference functional domain.

[0017] Preferably, the score of the independent oscillation source satisfies the following formula: in, , where is the independent scoring weight coefficient for functional domain D.

[0018] Preferably, the step of determining the oscillation source based on the oscillation source score includes: If the score of the independent oscillation source in the required scoring functional domain is greater than the first comparison benchmark, then the required scoring functional domain is determined to be an oscillation source. When the score of the independent oscillation source in the required scoring function domain one is less than or equal to the first comparison benchmark, the required scoring function domain two will be analyzed: If the score of the independent oscillation source in the second required scoring function domain is greater than the second comparison benchmark, then the second required scoring function domain is determined to be an oscillation source. When the score of the independent oscillation source in the required scoring functional domain 2 is less than or equal to the second comparison benchmark, the coupled side of the required scoring functional domain 1 and the required scoring functional domain 2 is determined to be the oscillation source. The first comparison benchmark is the maximum value of the independent oscillation source score of the required scoring functional domain two and the coupled oscillation source score of the required scoring functional domain one and the required scoring functional domain two, plus a stability threshold. The second comparison benchmark is the maximum value of the independent oscillation source score of the required scoring functional domain one and the coupled oscillation source score of the required scoring functional domain one and the required scoring functional domain two, plus a stability threshold.

[0019] The second aspect of this invention discloses a cross-system monitoring and tracing system for power oscillations in thermal power units, and the aforementioned monitoring and tracing method is implemented. The monitoring system includes: a main station server connected to multiple substation servers, with the substation servers set up in the monitoring substations and the main station server set up in the monitoring main station; Each substation server acquires the corresponding physical quantities of multiple independent control systems within its station, sets the corresponding physical quantities into channels and preprocesses them, constructs a joint matrix based on the preprocessed channels and performs singular value decomposition to extract the signal subspace, and then solves for the optimal modal parameters and angular frequency. Based on the optimal modal parameters, the domain participation degree and the average phase difference between domains are calculated, and the average directional intensity is calculated based on the angular frequency. Then, an oscillation source score is constructed, the oscillation source is judged based on the oscillation source score, and the judgment result is uploaded to the main station server, which generates the corresponding diagnostic result.

[0020] The beneficial effects of this invention are that, compared with the prior art, This invention achieves deep integration of multiple information across systems and regions, breaking through the limitations of traditional monitoring systems that can only monitor a single independent control system. It can simultaneously collect and process physical quantities of multiple independent control systems, accurately identify oscillation sources through oscillation source scoring, and calculate oscillation factors based on multiple physical quantities, greatly improving monitoring accuracy and information integrity, and realizing full-link visualization of the oscillation behavior of power grid units.

[0021] This invention achieves a leap from "early monitoring" to "source localization" of oscillations. On the one hand, it uses the joint analysis of introducing physical priors and sparse regularization to accurately extract oscillation modes. On the other hand, it quantifies oscillation source indicators from three dimensions: spatial distribution, time series, and causal influence, through domain participation, average phase difference between domains, and average directional intensity, and integrates multiple indicators into an oscillation source score. Furthermore, it combines a stability threshold to achieve robust and accurate judgment of oscillation sources, enabling precise identification of whether the oscillation source is generated by an independent control system or by the coupling of two independent control systems.

[0022] This invention achieves centralized analysis and distributed deployment. By connecting multiple monitoring substations and the main monitoring station, a two-level linkage mechanism is realized. The monitoring substations can quickly detect oscillation sources and transmit oscillation source information to the main monitoring station for early warning. This enables rapid, accurate, and quantifiable oscillation source location and control across the entire network, significantly improving the stability of the power system and the reliability of power supply. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of a cross-system monitoring and tracing system for power oscillation in thermal power units according to the present invention; In the diagram: 1. Main station server; 2. Sub-station server; 3. DCS side; 4. DEH side; 5. Excitation side. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of this invention.

[0025] Embodiment 1 of this invention discloses a cross-system monitoring and source tracing method for power oscillations in thermal power units, comprising the following steps: Step 1: Each monitoring substation acquires the corresponding physical quantities of multiple independent control systems within its station, aggregates the corresponding physical quantities into channels, and performs preprocessing. Specifically, each monitoring substation acquires the corresponding physical quantities of multiple independent control systems within its station, and aligns these corresponding physical quantities according to time to form a channel; The channels are then subjected to detrending, bandpass filtering, mean removal, and normalization to obtain standardized channels. The standardized channel satisfies the following formula: in, For the first The average of the physical quantities of each channel For the first The standard deviation of each channel The standardized channel is obtained through mean removal and normalization. For the first One channel, This represents the total number of samples. This is a discrete-time index.

[0026] like Figure 1 As shown, it can be understood that the corresponding physical quantities include not only the physical quantities on the PMU side such as power angle and power, but also the unit load on the DCS side. Actual power reactive power Main steam pressure Parameters; Deviation speed on DEH side 4 Deviation frequency Mechanical power Valve commands, valve feedback, and other parameters; excitation voltage on excitation side 5. Excitation current Generator terminal voltage Generator extreme current Parameters such as AVR (microcontroller) output and PSS (motor) speed.

[0027] Understandably, when the monitoring substation is set up inside a thermal power plant, the parameters on the DCS side reflect the boiler operating status, the parameters on the DEH side reflect the power source of the prime mover, and the parameters on the excitation side reflect the system damping.

[0028] Step 2: Construct a joint matrix based on the preprocessed multiple channels and perform singular value decomposition to extract the signal subspace, and then solve for the optimal modal parameters and angular frequency; Step 2.1: Construct a Hankel matrix for each channel, and stack all Hankel matrices according to the channel direction to construct a joint matrix; Specifically, the Hankel matrix conforms to That is, the number of rows is the preset window length. The number of columns is The matrix.

[0029] The joint matrix conforms to That is, the joint matrix has the number of rows equal to the number of channels. With preset window length The product of the products, the number of columns is The matrix.

[0030] Step 2.2: Decompose the joint matrix into a left singular vector matrix, a singular value matrix, and a right singular vector matrix using singular value decomposition, and determine the effective dimension of the signal subspace; Specifically, singular value decomposition is performed by substituting the joint matrix into the following formula: in, It is a left singular vector matrix. It is a singular value matrix. It is a right singular vector matrix.

[0031] Step 2.3: Select the first column vectors in the left singular vector matrix that correspond to the effective dimension, and combine them to form the signal subspace; Specifically, the signal subspace conforms to That is, the number of rows is equal to the number of channels. With preset window length The product of the columns is the effective dimension that is set or identified. The matrix.

[0032] Step 2.4: Construct a shift subspace based on the signal subspace, and obtain the optimal modal parameters and angular frequency by solving the shift subspace.

[0033] First, the signal subspace is multiplied by the first selection matrix and the second selection matrix respectively to obtain the first shift subspace and the second shift subspace offset by a preset sampling interval; The first shifted subspace satisfies the following formula: in, For the first shift subspace, This is the first selection matrix. This is the signal subspace.

[0034] The second shifted subspace satisfies the following formula: in, For the second shift subspace, This is the second selection matrix.

[0035] Then, the first shifted subspace is multiplied by the pseudo-inverse of the second shifted subspace to obtain the incidence matrix, and the modal poles are obtained by decomposing the incidence matrix. The correlation matrix satisfies the following formula: in, It is an incidence matrix. It is the pseudo-inverse of the second shifted subspace.

[0036] The modal poles satisfy the following formula: in, For the first The discrete-time modal poles of a mode are also called eigenvalues. For the corresponding eigenvectors.

[0037] Then, a Vandermonde matrix is ​​constructed based on the modal poles, and an objective function is constructed using the modal parameters as optimization variables. The optimal modal parameters are obtained by solving the objective function. The Vandermonde matrix satisfies the following formula: in, For the first Line number The matrix parameters of the column.

[0038] The objective function satisfies the following formula: in, These are the modal parameters, i.e., the modal parameter matrix. For data fitting terms, The observed data matrix obtained through preprocessing, The Vandermonde matrix, For sparse regularization, For sparse regularization coefficients, For the modal parameters All of the One modality, For all the The number of modes The physical prior regularization coefficient is... This is a physical prior.

[0039] It is understandable that the observation data matrix The rows correspond to the time series, and the columns correspond to the channels.

[0040] The physical prior terms satisfy the following formula: in, The modal parameters are the first The mode in the th ... Phase on each channel; The modal parameters are the first The mode in the th ... Phase on each channel For the preset first The first channel and the first The expected phase difference between the channels This is the tolerance threshold. For all channel pairs that require phase constraints A set of.

[0041] Understandably, the optimal modal parameters can be obtained by solving the objective function using a solver.

[0042] This invention introduces physical prior terms and sparse regularization terms in the process of solving modal parameters to achieve joint approximation of cross-domain signal modal consistency and sparse distribution of oscillation energy, thereby improving the accuracy of modal recognition and the ability to detect weak modes.

[0043] Simultaneously, the modal poles are mapped to complex modal poles, and the angular frequency is obtained by extracting the real and imaginary parts of the complex modal poles. The complex modal poles satisfy the following formula: in, For the corresponding The continuous-time complex modal poles; The sampling time interval; The attenuation factor is the real part; The imaginary part Angular frequencies of each mode This is a virtual part unit.

[0044] It is understandable that through modal poles The complex modal poles were calculated. Furthermore, by extracting the poles of the complex modes The angular frequency is obtained directly from the imaginary and real parts. and attenuation factor .

[0045] It is understandable that this invention can utilize angular frequency. The frequency and damping ratio are calculated to better understand the operating status of each independent control system; The frequency satisfies the following formula: in, For the corresponding angular frequency The frequency.

[0046] The damping ratio satisfies the following formula: in, For the corresponding angular frequency Damping ratio.

[0047] Step 3: Calculate the domain participation degree and the average phase difference between domains based on the optimal modal parameters, and calculate the average directional intensity based on the angular frequency, and then construct the oscillation source score; Step 3.1: Extract the oscillation amplitude and oscillation phase based on the optimal modal parameters, calculate the domain participation of multiple independent systems based on the oscillation amplitude, and calculate the average phase difference between domains using the oscillation phase; Specifically, the domain participation degree satisfies the following formula: in, In functional domain The Middle Domain participation of each modality, the functional domain For the corresponding independent control system, The first optimal modal parameter The mode in the th ... The oscillation amplitude on each channel In functional domain All channels.

[0048] The average phase difference between the domains satisfies the following formula: in, For the first functional domain Second functional domain In the Inter-domain average phase difference across modes For functional domains In the Domain-averaged phase over each mode For functional domains In the Domain-averaged phase over each mode.

[0049] The average phase of the domain satisfies the following formula: in, The optimal modal parameters are the first... The mode in the th ... The oscillation phase on each channel.

[0050] Step 3.2: Obtain the direction transfer function for all channels through the vector autoregression model, and calculate the average directional intensity by combining it with the angular frequency; Specifically, the direction transfer function satisfies the following formula: in, For the corresponding angular frequency Next The first channel to the first The direction transfer function value for each channel; To derive the first... The first channel to the first The corresponding transmitted value for each channel; The normalization factor derived from the vector autoregressive model represents all channels up to the 1st... Total transmission level of each channel For all possible directions to the first A channel for transmitting data.

[0051] The average directional intensity satisfies the following formula: Among them, the To the corresponding angular frequency First functional domain To the second functional domain The average directional intensity, For the first functional domain The number of channels, For the second functional domain The number of channels.

[0052] Step 3.3: Calculate the scores for independent oscillation sources and coupled oscillation sources based on domain participation, average phase difference between domains, and average directional intensity; Specifically, the score of the coupled oscillation source satisfies the following formula: in, , , The coupling scoring weight coefficients are: X is the required scoring functional domain one, T is the required scoring functional domain two, and E is the preset reference functional domain.

[0053] The score of the independent oscillation source satisfies the following formula: in, These are the independent scoring weight coefficients for the corresponding functional domain D.

[0054] It is understood that the oscillation source scoring of this invention is used to construct a comprehensive scoring index for multiple independent control systems by utilizing inter-domain participation, phase leadership relationship and direction transfer function based on multiple channels. It performs joint analysis between the independent control systems, solves the problem of not being able to determine whether the oscillation source comes from an independent control system or is caused by the coupling of multiple independent control systems, realizes accurate source tracing and factor analysis of oscillation sources, and facilitates timely processing afterward.

[0055] Step 4: Determine the oscillation source based on the oscillation source score, and upload the determination to the monitoring master station to generate diagnostic results; Specifically, the method of determining the oscillation source based on the oscillation source score includes: If the score of the independent oscillation source in the required scoring functional domain is greater than the first comparison benchmark, then the required scoring functional domain is determined to be an oscillation source. When the independent oscillator score of the required scoring function domain one is less than or equal to the first comparison benchmark, the independent oscillator score of the required scoring function domain two will be analyzed: If the score of the independent oscillation source in the second required scoring function domain is greater than the second comparison benchmark, then the second required scoring function domain is determined to be an oscillation source. When the score of the independent oscillation source in the required scoring functional domain 2 is less than or equal to the second comparison benchmark, the coupled side of the required scoring functional domain 1 and the required scoring functional domain 2 is determined to be the oscillation source. The first comparison benchmark is the maximum value of the independent oscillation source score of the required scoring functional domain two and the coupled oscillation source score of the required scoring functional domain one and the required scoring functional domain two, plus a stability threshold. The second comparison benchmark is the maximum value of the independent oscillation source score of the required scoring functional domain one and the coupled oscillation source score of the required scoring functional domain one and the required scoring functional domain two, plus a stability threshold.

[0056] In a specific embodiment, when Then it is determined that the required scoring function domain is the oscillation source; when The analysis will focus on the independent oscillation source scoring of the required scoring functional domain two, when... If so, then the required scoring function domain two is determined to be the oscillation source, when If so, then the coupling side of the required scoring functional domain one and the required scoring functional domain two is determined to be the oscillation source.

[0057] It is understood that the coupling side is an oscillation source generated by the interaction between the desired scoring functional domain one and the desired scoring functional domain two.

[0058] like Figure 1 As shown, Embodiment 2 of the present invention discloses a cross-system monitoring and tracing system for power oscillation of thermal power units, and runs the cross-system monitoring and tracing method for power oscillation of thermal power units. The monitoring system includes: a main station server 1 connected to multiple substation servers 2, the substation servers 2 being set in monitoring substations, and the main station server 1 being set in the monitoring main station. The substation server 2 is used to acquire the corresponding physical quantities of multiple independent control systems within the station, set the corresponding physical quantities into channels and preprocess them, construct a joint matrix based on the preprocessed channels and perform singular value decomposition to extract the signal subspace, and then solve for the optimal modal parameters and angular frequency. Based on the optimal modal parameters, the domain participation degree and the average phase difference between domains are calculated, and the average directional intensity is calculated based on the angular frequency. Then, an oscillation source score is constructed, the oscillation source is judged based on the oscillation source score, and the judgment is uploaded to the main station server 1. The main station server 1 generates the corresponding diagnostic results.

[0059] It is understood that the substation server 2 can be connected to multiple independent systems such as the DCS side 3, the DEH side, and the excitation side 5.

[0060] Understandably, the substation server 2 can package the corresponding physical quantities and processed key parameters into a standard format diagnostic result, and upload it to the main station server 1 along with the oscillation source judgment result, forming a panoramic view of low-frequency oscillation monitoring across multiple thermal power plants. Furthermore, since the oscillation source analysis is completed on the substation server 2, the substation server 2 can automatically execute oscillation control and suppression after judging the oscillation source, achieving rapid feedback and timely processing.

[0061] Understandably, the main station server 1 can display an early warning on the dispatch screen based on the vibration source judgment results of the entire network, informing the dispatcher of the global impact of the oscillation, the location of the root cause, and suggested measures. It also provides the dispatcher with a manual decision-making interface to confirm, stop, or amplify the control actions of the substations and issue globally unified dispatch instructions.

[0062] It is understandable that when the substation server 2 is set up in a thermal power plant, monitoring of all power plants can be achieved by connecting multiple substation servers 2 with the main station server 1.

[0063] The beneficial effects of this invention are that, compared with the prior art, This invention achieves deep integration of multiple information across systems and regions, breaking through the limitations of traditional monitoring systems that can only monitor a single independent control system. It can simultaneously collect and process physical quantities of multiple independent control systems, accurately identify oscillation sources through oscillation source scoring, and calculate oscillation factors based on multiple physical quantities, greatly improving monitoring accuracy and information integrity, and realizing full-link visualization of the oscillation behavior of power grid units.

[0064] This invention achieves a leap from "early monitoring" to "source localization" of oscillations. On the one hand, it uses the joint analysis of introducing physical priors and sparse regularization to accurately extract oscillation modes. On the other hand, it quantifies oscillation source indicators from three dimensions: spatial distribution, time series, and causal influence, through domain participation, average phase difference between domains, and average directional intensity, and integrates multiple indicators into an oscillation source score. Furthermore, it combines a stability threshold to achieve robust and accurate judgment of oscillation sources, enabling precise identification of whether the oscillation source is generated by an independent control system or by the coupling of two independent control systems.

[0065] This invention achieves centralized analysis and distributed deployment. By connecting multiple monitoring substations and the main monitoring station, a two-level linkage mechanism is realized. The monitoring substations can quickly detect oscillation sources and transmit oscillation source information to the main monitoring station for early warning. This enables rapid, accurate, and quantifiable oscillation source location and control across the entire network, significantly improving the stability of the power system and the reliability of power supply.

[0066] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0067] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0068] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0069] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A cross-system monitoring and source tracing method for power oscillation in thermal power units, characterized in that, include: Each monitoring substation acquires the corresponding physical quantities from multiple independent control systems within its station, aggregates the corresponding physical quantities into channels, and performs preprocessing. A joint matrix is ​​constructed based on the preprocessed multiple channels, and singular value decomposition is performed to extract the signal subspace, thereby solving for the optimal modal parameters and angular frequency; The domain participation degree and the average phase difference between domains are calculated based on the optimal modal parameters, and the average directional intensity is calculated based on the angular frequency, thereby constructing an oscillation source score; The oscillation source is determined based on the oscillation source score, and the determination result is uploaded to the monitoring master station to generate a diagnostic result.

2. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 1, characterized in that: The process of constructing a joint matrix based on preprocessed multiple channels and extracting the signal subspace through singular value decomposition, thereby solving for the optimal modal parameters and angular frequencies, includes: For each channel, construct a Hankel matrix. Stack all Hankel matrices according to the channel direction to construct a joint matrix. Perform singular value decomposition on the joint matrix and construct a signal subspace. Construct a shift subspace based on the signal subspace. Solve the shift subspace to obtain the modal poles. A Vandermonde matrix is ​​constructed based on the modal poles, and an objective function is constructed using the modal parameters as optimization variables. The optimal modal parameters are obtained by solving the objective function, and the modal poles are mapped to complex modal poles. The angular frequency is obtained by extracting the complex modal poles.

3. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 2, characterized in that: The Vandermonde matrix satisfies the following formula: in, For the first Line number The matrix parameters of the column.

4. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 2, characterized in that: The objective function satisfies the following formula: in, For modal parameters, The observed data matrix obtained through preprocessing, The Vandermonde matrix, For all the modal parameters of the modal parameters One modality, , The regularization coefficient is . This is a physical prior.

5. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 4, characterized in that: The physical prior terms satisfy the following formula: in, The modal parameters are the first The mode in the th ... Phase on each channel; The modal parameters are the first The mode in the th ... Phase on each channel For the preset first The first channel and the first The expected phase difference between the channels This is the tolerance threshold. For all channel pairs that require phase constraints A set of.

6. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 1, characterized in that: The domain participation degree and average inter-domain phase difference are calculated based on optimal modal parameters, and the average directional intensity is calculated based on angular frequency, thereby constructing an oscillation source score, including: The oscillation amplitude and oscillation phase are extracted based on the optimal modal parameters. The domain participation of multiple independent systems is calculated based on the oscillation amplitude. The average phase difference between domains is calculated using the oscillation phase. The directional transfer function of all channels is obtained through a vector autoregression model, and the average directional intensity is calculated by combining the angular frequency. Independent oscillation source scores and coupled oscillation source scores are calculated based on domain participation, average phase difference between domains, and average directional intensity.

7. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 6, characterized in that: The domain participation degree satisfies the following formula: in, In functional domain The Middle Domain participation of each modality, the functional domain For the corresponding independent control system, The first optimal modal parameter The mode in the th ... The oscillation amplitude on each channel In functional domain All channels.

8. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 6, characterized in that: The average phase difference between the domains satisfies the following formula: in, For the first functional domain Second functional domain In the Inter-domain average phase difference across modes For the first functional domain In the Domain-averaged phase over each mode For the second functional domain In the Domain-averaged phase over each mode.

9. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 8, characterized in that: The average phase of the domain satisfies the following formula: in, The optimal modal parameters are the first... The mode in the th ... The oscillation phase on each channel.

10. A cross-system monitoring and source tracing method for power oscillation in thermal power units according to claim 6, characterized in that: The direction transfer function satisfies the following formula: in, For the corresponding angular frequency Next The first channel to the first The direction transfer function value for each channel; To derive the first... The first channel to the first The corresponding transmitted value for each channel; The normalization factor derived from the vector autoregressive model represents all channels up to the 1st... Total transmission level of each channel.

11. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 6, characterized in that: The average directional intensity satisfies the following formula: Among them, the To the corresponding angular frequency First functional domain To the second functional domain The average directional intensity, This represents the number of channels for the corresponding functional domain.

12. The cross-system monitoring and source tracing method for power oscillation in thermal power units according to claim 6, characterized in that: The score of the coupled oscillation source satisfies the following formula: in, , , The coupling scoring weight coefficients are: X is the required scoring functional domain one, T is the required scoring functional domain two, and E is the preset reference functional domain.

13. The method for cross-system monitoring and source tracing of power oscillation in thermal power units according to claim 6, characterized in that: The score of the independent oscillation source satisfies the following formula: in, , where is the independent scoring weight coefficient for functional domain D.

14. The cross-system monitoring and source tracing method for power oscillation in thermal power units according to claim 1, characterized in that: The method of determining the oscillation source based on oscillation source scoring includes: If the score of the independent oscillation source in the required scoring functional domain is greater than the first comparison benchmark, then the required scoring functional domain is determined to be an oscillation source. When the score of the independent oscillation source in the required scoring function domain one is less than or equal to the first comparison benchmark, the required scoring function domain two will be analyzed: If the score of the independent oscillation source in the second required scoring function domain is greater than the second comparison benchmark, then the second required scoring function domain is determined to be an oscillation source. When the score of the independent oscillation source in the required scoring functional domain 2 is less than or equal to the second comparison benchmark, the coupled side of the required scoring functional domain 1 and the required scoring functional domain 2 is determined to be the oscillation source. The first comparison benchmark is the maximum value of the independent oscillation source score of the required scoring functional domain two and the coupled oscillation source score of the required scoring functional domain one and the required scoring functional domain two, plus a stability threshold. The second comparison benchmark is the maximum value of the independent oscillation source score of the required scoring functional domain one and the coupled oscillation source score of the required scoring functional domain one and the required scoring functional domain two, plus a stability threshold.

15. A cross-system monitoring and tracing system for power oscillations in thermal power units, operating the monitoring and tracing method as described in any one of claims 1 to 14, characterized in that: The monitoring system includes: a main station server (1) connected to multiple substation servers (2), the substation servers (2) being set up in monitoring substations, and the main station server (1) being set up in the monitoring main station; Each substation server (2) acquires the corresponding physical quantities of multiple independent control systems within the station, sets the corresponding physical quantities into channels and preprocesses them, constructs a joint matrix based on the preprocessed channels and extracts the signal subspace through singular value decomposition, and then solves for the optimal modal parameters and angular frequency. Based on the optimal modal parameters, it calculates the domain participation degree and the average phase difference between domains, calculates the average directional intensity based on the angular frequency, and then constructs an oscillation source score. Based on the oscillation source score, it judges the oscillation source and uploads the judgment result to the main station server (1). The main station server (1) generates the corresponding diagnostic result.

Images