Method and system for constructing digital twinborn model of phase modifier based on real-time data fitting

Abstract

The invention discloses a method and a system for constructing a digital twinborn model of a phase modifier based on real-time data fitting. The method comprises the following steps: acquiring real-time operation data of the phase modifier; based on vibration data in operation data of the phase modifier, a vibration spectral density characteristic quantity is extracted through wavelet transform processing, then reactive power in the operation data serves as an input variable, the vibration spectral density characteristic quantity serves as an output variable, function fitting is carried out through a least square method, and a phase modifier is obtained. Constructing a digital twinborn model between reactive power and vibration spectral density; inputting reactive power in operation data collected in real time into the digital twinborn model, and generating a predicted value of the vibration spectral density; calculating a standard deviation between the real-time vibration spectral density and the predicted value; when the standard deviation exceeds a preset threshold value, it is judged that the phase modifier is in an abnormal operation state, and an alarm mechanism is triggered. The method has the advantages of improving the monitoring efficiency of the running state of the phase modifier, reducing the operation and maintenance cost of the phase modifier and the like.

Description

Technical Field

[0001] This invention mainly relates to the field of camera condenser detection technology, specifically to a method and system for constructing a digital twin model of a camera condenser based on real-time data fitting. Background Technology

[0002] A synchronous condenser is essentially a type of synchronous motor operating under special conditions. It functions like a motor but without mechanical load, specifically designed to provide or absorb reactive power to the power system, thereby improving the power factor and maintaining grid voltage levels. In a power system, asynchronous motors and transformers, as primary loads, draw significant reactive power from the grid for their own excitation. This results in the grid carrying a large amount of inductive reactive current, leading to a decrease in the power factor. A reduced power factor not only prevents generators and transmission / distribution equipment from fully utilizing their capabilities but also increases line losses and voltage drops, affecting transmission quality and even threatening transmission stability.

[0003] With the high proportion of new energy sources being integrated into the power grid, synchronous condensers, as important reactive power support equipment, play a crucial role in improving grid stability, cross-regional power transmission, and the absorption capacity of new energy sources. They must cope with complex operating conditions such as high-frequency voltage fluctuations and long-term full-load operation, significantly increasing the difficulty of monitoring their operational status, providing early warning of faults, and optimizing operation and maintenance. Traditional mechanism-based physical modeling methods, due to their excessive complexity and the large number of parameters requiring adjustment, struggle to accurately reflect the dynamic characteristics of the equipment and the coupling effects of multiple physics fields. Meanwhile, purely data-driven models lack a deep understanding of the inherent attributes of the equipment, failing to meet the management needs of synchronous condensers in new power systems. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention provides a method and system for constructing a digital twin model of a synchronous condenser based on real-time data fitting, which improves the monitoring efficiency of the synchronous condenser's operating status and reduces the operation and maintenance costs of the synchronous condenser.

[0005] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows: A method for constructing a digital twin model of a camera adjustment camera based on real-time data fitting includes the following steps: Acquire real-time operating data of the synchronous condenser; the operating data includes vibration data and reactive power. Based on the vibration data in the operation data of the synchronous condenser, wavelet transform is used to extract the vibration spectral density feature. Then, the reactive power in the operation data is used as the input variable, and the vibration spectral density feature is used as the output variable. The least squares method is used to fit the function to construct a digital twin model between reactive power and vibration spectral density. The reactive power from the real-time collected operational data is input into the digital twin model to generate a predicted value of the vibration spectral density. Calculate the standard deviation between the real-time vibration spectral density and the predicted value; when the standard deviation exceeds the preset threshold, determine that the synchronous condenser is in an abnormal operating state and trigger the alarm mechanism.

[0006] Preferably, the specific process of extracting vibrational spectral density features using wavelet transform is as follows: The acquired raw vibration signals are preprocessed to remove DC and trend. The preprocessed signal is subjected to continuous wavelet transform using complex-valued wavelet basis functions to obtain the time-frequency distribution map. The frequency-energy spectrum is obtained by averaging the time dimension of the time-frequency distribution map. The energy peaks corresponding to the fundamental frequency, second fundamental frequency, and third fundamental frequency are identified from the frequency-energy spectrum to complete the extraction of vibrational spectral density features.

[0007] Preferably, the specific process of constructing a digital twin model between reactive power and vibration spectral density by using the least squares method for function fitting is as follows: The collected vibration spectral density data were plotted as a scatter plot; The least squares method is used to fit the scattered data to obtain a digital twin model between reactive power and vibration spectral density, so as to generate a prediction curve of vibration spectral density.

[0008] Preferably, standard deviation The calculation formula is:

[0009] in, For the first One data point; The average of all data; The total number of data points; This is the sum of squares of the differences between all data points and the mean.

[0010] Preferably, the method also includes periodically verifying the digital twin model, including revising the fitting parameters and recording parameter changes, and triggering an alarm when the parameter change rate exceeds a threshold.

[0011] Preferably, the conditions for triggering the alarm include at least one of the following: The standard deviation between the digital twin model's predicted values ​​and the real-time collected data exceeds the limit; Real-time collected values ​​exceeded limits; The digital twin model's predicted values ​​exceeded the limit; The coefficients of the digital twin model have exceeded the limit.

[0012] Preferably, the operating data further includes electrical parameters, mechanical parameters, and thermal parameters; the electrical parameters include stator current, voltage, power factor, active power, excitation current, and voltage; the mechanical parameters include rotor speed and bearing displacement; and the thermal parameters include stator winding temperature, rotor winding temperature, stator cooling outlet temperature, rotor cooling outlet temperature, synchronous condenser housing temperature, and lubricating oil temperature.

[0013] The present invention also discloses a computer program product, comprising a computer program that, when executed by a processor, performs the steps of the method described above.

[0014] The present invention further discloses a computer-readable storage medium having a computer program stored thereon, the computer program executing the steps of the method described above when run by a processor.

[0015] The present invention also discloses a computer system including a memory and a processor interconnected thereon, wherein the memory stores a computer program that, when run by the processor, performs the steps of the method described above.

[0016] This invention further discloses a system for constructing a digital twin model of a camera adjustment camera based on real-time data fitting, comprising: The data acquisition module is used to acquire real-time operating data of the synchronous condenser; the operating data includes vibration data and reactive power. The model building module is used to extract vibration spectral density features from the vibration data in the operation data of the synchronous condenser using wavelet transform. Then, the reactive power in the operation data is used as the input variable, and the vibration spectral density features are used as the output variable. The least squares method is used to fit the function to build a digital twin model between reactive power and vibration spectral density. The prediction module is used to input the reactive power from the real-time collected operating data into the digital twin model to generate a predicted value of the vibration spectral density. The early warning module is used to calculate the standard deviation between the real-time vibration spectral density and the predicted value; when the standard deviation exceeds the preset threshold, it is determined that the synchronous condenser is in an abnormal operating state and an alarm mechanism is triggered.

[0017] Preferably, in the model building module, the specific process of extracting vibrational spectral density features using wavelet transform is as follows: The acquired raw vibration signals are preprocessed to remove DC and trend. The preprocessed signal is subjected to continuous wavelet transform using complex-valued wavelet basis functions to obtain the time-frequency distribution map. The frequency-energy spectrum is obtained by averaging the time dimension of the time-frequency distribution map. The energy peaks corresponding to the fundamental frequency, second fundamental frequency, and third fundamental frequency are identified from the frequency-energy spectrum to complete the extraction of vibrational spectral density features.

[0018] Preferably, in the model building module, the process of using the least squares method to fit the function and construct a digital twin model between reactive power and vibration spectral density is as follows: The collected vibration spectral density data were plotted as a scatter plot; The least squares method is used to fit the scattered data to obtain a digital twin model between reactive power and vibration spectral density, so as to generate a prediction curve of vibration spectral density.

[0019] Preferably, in the early warning module, the standard deviation The calculation formula is:

[0020] in, For the first One data point; The average of all data; The total number of data points; This is the sum of squares of the differences between all data points and the mean.

[0021] Preferably, it also includes a verification module for periodically verifying the digital twin model, including revising the fitting parameters and recording parameter changes, and triggering an alarm when the parameter change rate exceeds a threshold.

[0022] Preferably, in the early warning module, the conditions for triggering the alarm include at least one of the following: The standard deviation between the digital twin model's predicted values ​​and the real-time collected data exceeds the limit; Real-time collected values ​​exceeded limits; The digital twin model's predicted values ​​exceeded the limit; The coefficients of the digital twin model have exceeded the limit.

[0023] Preferably, in the data acquisition module, the operating data further includes electrical parameters, mechanical parameters, and thermal parameters; the electrical parameters include stator current, voltage, power factor, active power, excitation current, and voltage; the mechanical parameters include rotor speed and bearing displacement; and the thermal parameters include stator winding temperature, rotor winding temperature, stator cooling outlet temperature, rotor cooling outlet temperature, synchronous condenser housing temperature, and lubricating oil temperature.

[0024] Compared with the prior art, the advantages of the present invention are as follows: The digital twin modeling of the vibration spectral density of synchronous condensers in this invention achieves multi-dimensional technological breakthroughs in the field of fault monitoring through the deep integration of virtual-real mapping, real-time simulation, and intelligent diagnosis. Its main benefits are reflected in four aspects: improved fault diagnosis accuracy, enhanced operation and maintenance efficiency, optimized full lifecycle management, and improved power grid safety. Virtual-real data comparison and diagnosis: the model collects the operating voltage, reactive power, and excitation current of the synchronous condenser in real time through physical layer sensors, while simultaneously reproducing the equipment's operating status in the digital twin. By monitoring and comparing the vibration amplitude of the synchronous condenser, it can be determined whether the current operating state is abnormal. If the standard deviation between the real-time vibration spectral density and the fitted predicted vibration spectral density exceeds 5%, it is determined to be an abnormal operating state, and an alarm response is initiated. This invention significantly improves the monitoring efficiency of the synchronous condenser's operating status and reduces the operation and maintenance costs of the synchronous condenser.

[0025] This invention presents a method for constructing a digital twin model of a synchronous condenser based on real-time data fitting. By building a closed-loop system of "physical entity-virtual model-data interaction," it can realize real-time mapping of the synchronous condenser's operating status, early warning of faults, and simulation optimization of operation and maintenance strategies. Constructing a high-precision digital twin model of a synchronous condenser not only breaks through the limitations of traditional monitoring methods and improves the reliability of equipment operation, but also reduces operation and maintenance costs (expected to reduce downtime maintenance time by 20% to 30%), which is of great engineering significance for ensuring the voltage stability and safe and economical operation of the power system.

[0026] This invention monitors the operating status of synchronous condensers and provides real-time predictions and early warnings through a digital twin model. By considering and studying the types, causes, and impacts of synchronous condenser failures, and existing modern intelligent sensors capable of detecting their operating status, a digital twin model is developed to monitor the synchronous condenser's operating status and provide early warnings for abnormal vibrations. This allows for timely detection of problems and rapid resolution of potential faults during synchronous condenser operation, improving equipment reliability, reducing downtime, and lowering maintenance costs. Attached Figure Description

[0027] Figure 1 This is a diagram illustrating the construction of the digital twin model of the present invention.

[0028] Figure 2 This is a diagram of the digital twin model architecture of the camera converter of the present invention.

[0029] Figure 3 This is a schematic diagram of data acquisition according to the present invention.

[0030] Figure 4 This diagram illustrates four alarm scenarios according to the present invention.

[0031] Figure 5 The flowchart below shows the construction method of the present invention in an embodiment.

[0032] Figure 6 The following are vibration waveform diagrams of the present invention: (a) is the synthesized vibration signal; (b) is the fundamental frequency signal; (c) is the second fundamental frequency signal; and (d) is the third fundamental frequency signal.

[0033] Figure 7 This is a vibration twin data diagram from the present invention.

[0034] Figure 8 (a) is a comparison diagram of the amplitude collected and the predicted amplitude by the synchronous condenser in this invention; (b) is a curve of the reactive power output of the synchronous condenser; and (c) is a monitoring diagram of the vibration spectral density of the synchronous condenser. Detailed Implementation

[0035] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0036] like Figure 1 As shown, the essence of digital twin modeling is "to replicate the physical world using digital means and to optimize physical operations using data and intelligence." Its core value lies in breaking the limitations of "physical entities being invisible and unpredictable," and achieving a shift "from passive response to proactive prediction." The working principle of digital twins is a complex and sophisticated process, specifically: First, operational data from the synchronous condenser is collected using sensors and IoT technology and transmitted to a data processing center. This data serves as the foundational information for the digital twin model, supporting subsequent model construction and analysis. Next, a digital model of the physical entity is constructed based on the collected data and simulated in a virtual environment, mimicking the entity's behavior in the real world. By monitoring the entity's status in real time, its performance and potential problems are analyzed to support decision-making. Based on simulation results and data analysis, the entity's operation can be optimized, and future performance and potential risks can be predicted. Finally, these analytical results are transformed into decision support to guide the maintenance, upgrades, and operation of the physical entity.

[0037] like Figure 5 As shown in the embodiment of the present invention, the method for constructing a digital twin model of a synchronous condenser based on real-time data fitting is based on digital twin modeling to monitor the operating status of the synchronous condenser, and includes the following steps: S1. Real-time acquisition of synchronous condenser operating data; in terms of electrical parameters, including stator current, voltage, power factor, active power, reactive power, and excitation current and voltage; in terms of mechanical parameters, including rotor speed, bearing amplitude, and bearing displacement; in terms of thermal parameters, including stator winding temperature, rotor winding temperature, stator cooling outlet temperature, rotor cooling outlet temperature, synchronous condenser housing temperature, and lubricating oil temperature; Specifically, synchronous condensers generate a rich variety of data during operation, which forms a crucial foundation for building digital twin models. Data acquisition covers several key aspects, including electrical parameters such as stator current, voltage, power factor, active power, reactive power, and excitation current and voltage. Monitoring stator current and voltage reflects the load and operating status of the synchronous condenser; the power factor directly reflects the synchronous condenser's reactive power compensation effect; and real-time active and reactive power data help analyze the energy exchange within the power system.

[0038] Monitoring excitation current and voltage is crucial for the excitation control and performance adjustment of synchronous condensers. Real-time acquisition of these parameters allows for timely adjustments to the excitation system, ensuring stable operation of the synchronous condenser. Mechanical parameters include speed, vibration, and shaft displacement. Speed ​​is a critical indicator of synchronous condenser operation; a stable speed is fundamental for its proper functioning. Monitoring vibration and shaft displacement can promptly detect potential faults in the synchronous condenser's mechanical components, such as bearing wear or rotor imbalance. When vibration values ​​or shaft displacement exceed normal ranges, it may indicate a potential mechanical failure requiring timely inspection and maintenance.

[0039] S2. Construct a digital twin model of the camera adapter based on operational data, such as Figure 2 As shown; specifically, fitting methods such as least squares method, polynomial fitting, power function fitting, and machine learning fitting are used to establish digital twin models of each operating state of the synchronous condenser. Taking the vibration of a synchronous condenser as an example, wavelet transform is used to process the frequency domain analysis of the condenser's vibration, resulting in a time-frequency distribution map. The fundamental frequency is then identified from the energy concentration region in the time-frequency distribution map. Double the fundamental frequency and three times the base frequency In order to obtain the vibration spectral density characteristic quantity that can characterize the physical state of the mechanical operation of the camera; Then, using the reactive power collected in S1 as the input variable and the vibration spectral density characteristic quantity as the output variable, the least squares method is used to fit the function to construct a quantitative prediction model between reactive power and vibration spectral density, and the vibration spectral density prediction curve is obtained. The digital twin model of the camera is calibrated at regular intervals, such as on the 1st of each month, to calibrate and revise the fitting parameters and record the changes in the fitting parameters; if the change exceeds a threshold, an alarm is triggered; the threshold can be manually adjusted, for example, to 5%; Taking the vibration spectral density of a synchronous condenser during operation as an example, a data-fitting digital twin model of the vibration spectral density of the synchronous condenser is constructed.

[0040] The vibration of the synchronous condenser is a combination of mechanical and electrical characteristics during operation, primarily manifested in vibration frequency and amplitude. The dominant frequency is the rotational speed frequency, accompanied by small amounts of second and third harmonic components; the rotational speed frequency component accounts for 70%–80% of the total vibration energy. The vibration amplitude distribution shows the vertical acceleration of the bearing housing is typically 1–2g, the horizontal acceleration is 0.8–1.5g, and the axial acceleration (rotor movement direction) is the smallest (0.5–1g). The stator housing vibration amplitude is only 1 / 3 to 1 / 2 that of the bearing housing (due to the rigid support of the housing). This type of vibration is caused by the inherent characteristics of the mechanical system (such as minor rotor imbalance) and conventional electromagnetic coupling effects (such as minor non-uniformity of the air gap magnetic field). If the amplitude is stable and does not increase significantly, it falls within the normal operating range.

[0041] Wavelet transform, through scaling and translation operations, decomposes a signal into components at different frequency scales. It reflects frequency characteristics while preserving time information, making it particularly suitable for extracting the fundamental frequency and harmonics of non-stationary signals (such as noisy vibration signals). Its core principle is to obtain a time-frequency distribution map using continuous wavelet transform (CWT) to identify the fundamental frequency from areas of concentrated energy. Double the fundamental frequency and three times the base frequency In order to obtain the vibration spectral density characteristic quantity that can characterize the physical state of the mechanical operation of the camera.

[0042] Specifically, let the signal be... The wavelet basis functions are Then the signal is at scale and time The CWT at this location is defined as: (1) in: scale parameter This determines the frequency resolution (small scales correspond to high frequencies, and large scales correspond to low frequencies). The translation parameter represents the reaction time and location. wavelet basis functions Complex conjugate; integral result modulus This indicates the energy intensity at that time and frequency point.

[0043] In wavelet transform, scaling With actual frequency Through the center frequency of the wavelet fundamental Related: (2) in, The signal sampling frequency is used to calculate the scale using this formula. Convert to actual frequency .

[0044] The specific process of extracting vibrational spectral density features using wavelet transform is as follows: The original signal undergoes preprocessing such as DC removal and trend removal to eliminate baseline drift. The preprocessed signal is denoted as [signal name missing]. ; The complex-valued wavelet function is selected to perform continuous wavelet transform on the preprocessed signal to obtain the time-frequency distribution map. The complex-valued wavelet function has a stronger ability to resolve frequencies, and its complex-valued wavelet formula is: (3) in, The imaginary unit; The center frequency.

[0045] Based on the complex-valued wavelet formula above, for Within the preset scale range The time-frequency matrix is ​​calculated internally to obtain... The time-frequency distribution diagram; In the time-frequency distribution diagram, the frequency component with the most concentrated energy is the fundamental frequency. The frequency-energy spectrum is obtained by averaging the time dimension of the time-frequency distribution plot. Read directly from the frequency-energy spectrum , , The corresponding energy peak value is used to extract the vibrational spectral density characteristic quantity.

[0046] The least squares method used to construct the prediction model is a core mathematical method for data fitting and parameter estimation. Its core idea is to find the mathematical model (such as a straight line or curve) that best reflects the inherent laws of the data by minimizing the sum of squares of the errors between the observed values ​​and the model's predicted values. The least squares method has the advantages of convenient solution, mathematical simplicity, and error minimization, and is the most classic tool for connecting "data" and "linear model".

[0047] In this embodiment, the prediction model for vibration spectral density and reactive power established using the least squares method specifically includes: The collected vibration spectral density data were plotted as a scatter plot; The least squares method was used to fit the scattered data to obtain a quadratic function relationship model between the vibration spectral density and reactive power. The predicted curve of vibrational spectral density is generated based on the fitted model.

[0048] Specifically, reactive power Q and vibration amplitude The expression is: (4) Fitting by least squares method formula: make (5) The original model then becomes a linear form: (6) Assume there are n sets of observation data After conversion, it becomes in .

[0049] The objective of the least squares method is to minimize the sum of squared residuals. : (7) In order to seek The minimum value of the parameter and Find the partial derivatives of each and set them equal to 0: (8) (9) (10) (11) like Figure 6 As shown, the first curve represents the combined vibration signal, exhibiting a complex vibration pattern; it is the effect of superimposing the fundamental frequency, its second harmonic, and its third harmonic. The second curve represents the fundamental frequency signal, with a frequency of... , is the fundamental frequency component of the combined signal. The third curve is the second harmonic of the fundamental frequency signal, with a frequency of . The waveform is denser than the fundamental frequency, representing the second harmonic of the fundamental frequency. The fourth curve is the third harmonic of the fundamental frequency, with a frequency of [missing value]. The most dense waveform is the third harmonic of the fundamental frequency. The combined vibration signal is composed of the fundamental frequency and its integer multiples of harmonics (second harmonic, third harmonic). The higher the harmonic order, the smaller the amplitude and the higher the frequency (the denser the waveform).

[0050] like Figure 7 As shown, the blue scatter plots represent the real-time vibration spectral density data of the synchronous condenser, and the red curve is the trend fit of the vibration spectral density scatter plot data. When the reactive power Q gradually changes from 0 to negative, the vibration spectral density gradually decreases from approximately 10 mm; when Q gradually increases from 0 to positive, the vibration spectral density decreases from its lowest value (5 mm). The reactive power gradually increases from 6mm. ) and vibrational spectral density (unit: The vibration spectral density exhibits a quadratic function relationship: when the reactive power is greater than 0, the vibration spectral density increases with the increase of reactive power; when the reactive power is less than 0, the vibration spectral density also increases with the increase of reactive power. The vibration spectral density is only related to the magnitude of the reactive power and is independent of the sign of the reactive power.

[0051] S3. Use the constructed digital twin model to perform simulation and generate predicted values ​​of the synchronous condenser's operating status; Specifically, the reactive power collected in real time in S1 is input into the quantitative prediction model constructed in S3, and the predicted value of the vibration spectral density is output. S4. Calculate the standard deviation between the real-time acquired operating data and the predicted value to determine the operating status of the synchronous condenser; if the standard deviation is less than 5%, the synchronous condenser is in normal operating condition; if the standard deviation is greater than or equal to 5%, the synchronous condenser may be in abnormal operating condition.

[0052] Specifically, if the standard deviation between the collected vibration spectral density (obtained directly from the collected vibration data) and the fitted predicted vibration spectral density is less than 5%, the synchronous condenser is determined to be in normal operating condition; if the standard deviation between the real-time vibration spectral density and the predicted curve data is greater than or equal to 5%, the synchronous condenser is determined to be in abnormal operating condition, and an alarm mechanism needs to be activated.

[0053] The formula for calculating the standard deviation is: (12) in, For the first One data point; The average of all data; The total number of data points; This is the sum of squares of the differences between all data points and the mean.

[0054] like Figure 8 As shown in (a), the curve represents the reactive power output of the synchronous condenser. Above the X-axis is "reactive power generation," during which the condenser is in a reactive power generation state most of the time, corresponding to the blue curve mostly above the X-axis. Below the X-axis is "reactive power absorption," during which the condenser is in a reactive power absorption state only for a short period, at which time the blue curve extends downwards into the orange area, indicating negative reactive power. The light blue area represents reactive power generation, occupying the vast majority of the time, indicating that the synchronous condenser mainly operates in a reactive power generation state, while reactive power absorption (light orange area) is extremely rare. The proportion of reactive power generation is 97.5%, while the proportion of reactive power absorption is only 2.5%, further confirming the condenser's primary reactive power generation characteristic.

[0055] like Figure 8As shown in (b), the curve represents the vibration spectral density monitoring of the camera. The pink and red scatter points represent the collected and predicted vibration spectral density values ​​exceeding the threshold, respectively. The green curve of collected vibration spectral density values ​​fluctuates within a certain range, generally revolving around 5~18. The blue predicted vibration spectral density curve and the green acquired value curve show similar trends, with similar fluctuation rhythms and approximate ranges. For most of the time, the acquired vibration spectral density values ​​(green curve) and the predicted vibration spectral density values ​​(blue curve) are close to each other, indicating a good match between the predicted and actual acquired values. However, there are also periods where there are certain deviations, reflecting that the vibration spectral density of the condenser camera changes dynamically during the monitoring period.

[0056] The vibrational spectral density threshold set in the model is 15. ,from Figure 8 From the data, five points of both the collected and predicted values ​​exceeded the threshold, indicating that the synchronous condenser may be operating abnormally and needs to be stopped immediately to investigate the problem.

[0057] like Figure 3 As shown, the synchronous condenser collects a certain amount of data in real time. The corresponding sensors collect the data and send it to the PC host computer for processing and fitting. All collected data, fitted calculation prediction data, and coefficients of the synchronous condenser's digital twin model are checked at regular intervals, such as on the 1st of each month, to revise the fitting parameters and record changes in the fitting parameters. If the change exceeds a threshold, an alarm is triggered. The threshold can be manually adjusted, for example, to 10%.

[0058] The digital twin model of the synchronous condenser will issue an alarm in the following four situations, alerting technicians that there may be an abnormality in the synchronous condenser's operating status. (See Table 1 and...) Figure 4 As shown.

[0059] Table 1 Abnormal situations of four types of alarms

[0060] The digital twin modeling of the vibration spectral density of synchronous condensers in this invention achieves multi-dimensional technological breakthroughs in the field of fault monitoring through the deep integration of virtual-real mapping, real-time simulation, and intelligent diagnosis. Its main benefits are reflected in four aspects: improved fault diagnosis accuracy, enhanced operation and maintenance efficiency, optimized full lifecycle management, and improved power grid safety. Virtual-real data comparison and diagnosis: the model collects the operating voltage, reactive power, and excitation current of the synchronous condenser in real time through physical layer sensors, while simultaneously reproducing the equipment's operating status in the digital twin. By monitoring and comparing the vibration amplitude of the synchronous condenser, it can be determined whether the current operating state is abnormal. If the standard deviation between the real-time vibration spectral density and the fitted predicted vibration spectral density exceeds 5%, it is determined to be an abnormal operating state, and an alarm response is initiated. This invention significantly improves the monitoring efficiency of the synchronous condenser's operating status and reduces the operation and maintenance costs of the synchronous condenser.

[0061] This invention presents a method for constructing a digital twin model of a synchronous condenser based on real-time data fitting. By building a closed-loop system of "physical entity-virtual model-data interaction," it can realize real-time mapping of the synchronous condenser's operating status, early warning of faults, and simulation optimization of operation and maintenance strategies. Constructing a high-precision digital twin model of a synchronous condenser not only breaks through the limitations of traditional monitoring methods and improves the reliability of equipment operation, but also reduces operation and maintenance costs (expected to reduce downtime maintenance time by 20% to 30%), which is of great engineering significance for ensuring the voltage stability and safe and economical operation of the power system.

[0062] This invention also provides a system for constructing a digital twin model of a camera adjustment camera based on real-time data fitting, comprising: The data acquisition module is used to acquire real-time operating data of the synchronous condenser; the operating data includes vibration data and reactive power. The model building module is used to extract vibration spectral density features from the vibration data in the operation data of the synchronous condenser using wavelet transform. Then, the reactive power in the operation data is used as the input variable, and the vibration spectral density features are used as the output variable. The least squares method is used to fit the function to build a digital twin model between reactive power and vibration spectral density. The prediction module is used to input the reactive power from the real-time collected operating data into the digital twin model to generate a predicted value of the vibration spectral density. The early warning module is used to calculate the standard deviation between the real-time vibration spectral density and the predicted value; when the standard deviation exceeds the preset threshold, it is determined that the synchronous condenser is in an abnormal operating state and an alarm mechanism is triggered.

[0063] Specifically, in the model building module, the process of extracting vibrational spectral density features using wavelet transform is as follows: The acquired raw vibration signals are preprocessed to remove DC and trend. The preprocessed signal is subjected to continuous wavelet transform using complex-valued wavelet basis functions to obtain the time-frequency distribution map. The frequency-energy spectrum is obtained by averaging the time dimension of the time-frequency distribution map. The energy peaks corresponding to the fundamental frequency, second fundamental frequency, and third fundamental frequency are identified from the frequency-energy spectrum to complete the extraction of vibrational spectral density features.

[0064] Specifically, in the model building module, the process of constructing a digital twin model between reactive power and vibration spectral density by using the least squares method for function fitting is as follows: The collected vibration spectral density data were plotted as a scatter plot; The least squares method is used to fit the scattered data to obtain a digital twin model between reactive power and vibration spectral density, so as to generate a prediction curve of vibration spectral density.

[0065] Specifically, in the early warning module, the standard deviation The calculation formula is: (12) in, For the first One data point; The average of all data; The total number of data points; This is the sum of squares of the differences between all data points and the mean.

[0066] Specifically, it also includes a verification module for periodically verifying the digital twin model, including revising the fitting parameters and recording parameter changes, and triggering an alarm when the parameter change rate exceeds a threshold.

[0067] Specifically, in the early warning module, the conditions for triggering an alarm include at least one of the following: The standard deviation between the digital twin model's predicted values ​​and the real-time collected data exceeds the limit; Real-time collected values ​​exceeded limits; The digital twin model's predicted values ​​exceeded the limit; The coefficients of the digital twin model have exceeded the limit.

[0068] This invention monitors the operating status of synchronous condensers and provides real-time predictions and early warnings through a digital twin model. By considering and studying the types, causes, and impacts of synchronous condenser failures, and existing modern intelligent sensors capable of detecting their operating status, a digital twin model is developed to monitor the synchronous condenser's operating status and provide early warnings for abnormal vibrations. This allows for timely detection of problems and rapid resolution of potential faults during synchronous condenser operation, improving equipment reliability, reducing downtime, and lowering maintenance costs.

[0069] This invention also discloses a computer program product, including a computer program that, when run by a processor, executes the steps of the method described above.

[0070] The present invention further discloses a computer-readable storage medium having a computer program stored thereon, the computer program executing the steps of the method described above when run by a processor.

[0071] This invention also discloses a computer system including a memory and a processor interconnected, wherein the memory stores a computer program, and the computer program executes the steps of the method described above when run by the processor.

[0072] The products, media, and systems of the present invention, corresponding to the methods described above, also possess the advantages described above.

[0073] The present invention can implement all or part of the processes in the methods of the above embodiments, or it can be implemented by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium includes: any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. The memory is used to store computer programs and / or modules. The processor implements various functions by running or executing the computer programs and / or modules stored in the memory, and by calling data stored in the memory. The memory may include high-speed random access memory, as well as non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0074] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for constructing a digital twin model of a camera adjuster based on real-time data fitting, characterized in that, Including the following steps: Acquire real-time operating data of the synchronous condenser; the operating data includes vibration data and reactive power. Based on the vibration data in the operation data of the synchronous condenser, wavelet transform is used to extract the vibration spectral density feature. Then, the reactive power in the operation data is used as the input variable, and the vibration spectral density feature is used as the output variable. The least squares method is used to fit the function to construct a digital twin model between reactive power and vibration spectral density. The reactive power from the real-time collected operational data is input into the digital twin model to generate a predicted value of the vibration spectral density. Calculate the standard deviation between the real-time vibration spectral density and the predicted value; when the standard deviation exceeds the preset threshold, determine that the synchronous condenser is in an abnormal operating state and trigger the alarm mechanism.

2. The method for constructing a digital twin model of a camera based on real-time data fitting according to claim 1, characterized in that, The specific process of extracting vibrational spectral density features using wavelet transform is as follows: The acquired raw vibration signals are preprocessed to remove DC and trend. The preprocessed signal is subjected to continuous wavelet transform using complex-valued wavelet basis functions to obtain the time-frequency distribution map. The frequency-energy spectrum is obtained by averaging the time dimension of the time-frequency distribution map. The energy peaks corresponding to the fundamental frequency, second fundamental frequency, and third fundamental frequency are identified from the frequency-energy spectrum to complete the extraction of vibrational spectral density features.

3. The method for constructing a digital twin model of a camera based on real-time data fitting according to claim 2, characterized in that, The specific process of constructing a digital twin model between reactive power and vibration spectral density by using the least squares method for function fitting is as follows: The collected vibration spectral density data were plotted as a scatter plot; The least squares method is used to fit the scattered data to obtain a digital twin model between reactive power and vibration spectral density, so as to generate a prediction curve of vibration spectral density.

4. The method for constructing a digital twin model of a camera based on real-time data fitting according to claim 1, 2, or 3, characterized in that, Standard deviation The calculation formula is: in, For the first One data point; The average of all data; The total number of data points; This is the sum of squares of the differences between all data points and the mean.

5. The method for constructing a digital twin model of a camera based on real-time data fitting according to claim 1, 2, or 3, characterized in that, It also includes regularly verifying the digital twin model, including revising the fitting parameters and recording parameter changes, and triggering an alarm when the parameter change rate exceeds a threshold.

6. The method for constructing a digital twin model of a camera based on real-time data fitting according to claim 1, 2, or 3, characterized in that, The conditions for triggering an alarm include at least one of the following: The standard deviation between the digital twin model's predicted values ​​and the real-time collected data exceeds the limit; Real-time collected values ​​exceeded limits; The digital twin model's predicted values ​​exceeded the limit; The coefficients of the digital twin model have exceeded the limit.

7. The method for constructing a digital twin model of a camera based on real-time data fitting according to claim 1, 2, or 3, characterized in that, The operating data also includes electrical parameters, mechanical parameters, and thermal parameters; the electrical parameters include stator current, voltage, power factor, active power, excitation current, and voltage; the mechanical parameters include rotor speed and bearing displacement; and the thermal parameters include stator winding temperature, rotor winding temperature, stator cooling outlet temperature, rotor cooling outlet temperature, synchronous condenser housing temperature, and lubricating oil temperature.

8. A computer program product, comprising a computer program, characterized in that, The computer program is executed by the processor to perform the steps of the method as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1-7.

10. A computer system comprising a memory and a processor interconnected thereon, the memory storing a computer program, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1-7.

11. A system for constructing a digital twin model of a camera based on real-time data fitting, characterized in that, include: The data acquisition module is used to acquire real-time operating data of the synchronous condenser; the operating data includes vibration data and reactive power. The model building module is used to extract vibration spectral density features from the vibration data in the operation data of the synchronous condenser using wavelet transform. Then, the reactive power in the operation data is used as the input variable, and the vibration spectral density features are used as the output variable. The least squares method is used to fit the function to build a digital twin model between reactive power and vibration spectral density. The prediction module is used to input the reactive power from the real-time collected operating data into the digital twin model to generate a predicted value of the vibration spectral density. The early warning module is used to calculate the standard deviation between the real-time vibration spectral density and the predicted value. When the standard deviation exceeds the preset threshold, it is determined that the synchronous condenser is in an abnormal operating state and an alarm mechanism is triggered.

12. The system for constructing a digital twin model of a camera based on real-time data fitting according to claim 11, characterized in that, In the model building module, the specific process of extracting vibrational spectral density features using wavelet transform is as follows: The acquired raw vibration signals are preprocessed to remove DC and trend. The preprocessed signal is subjected to continuous wavelet transform using complex-valued wavelet basis functions to obtain the time-frequency distribution map. The frequency-energy spectrum is obtained by averaging the time dimension of the time-frequency distribution map. The energy peaks corresponding to the fundamental frequency, second fundamental frequency, and third fundamental frequency are identified from the frequency-energy spectrum to complete the extraction of vibrational spectral density features.

13. The system for constructing a digital twin model of a camera based on real-time data fitting according to claim 12, characterized in that, In the model building module, the least squares method is used for function fitting, and the specific process of constructing a digital twin model between reactive power and vibration spectral density is as follows: The collected vibration spectral density data were plotted as a scatter plot; The least squares method is used to fit the scattered data to obtain a digital twin model between reactive power and vibration spectral density, so as to generate a prediction curve of vibration spectral density.

14. The system for constructing a digital twin model of a camera based on real-time data fitting according to claim 11, 12, or 13, characterized in that, In the early warning module, standard deviation The calculation formula is: in, For the first One data point; The average of all data; The total number of data points; This is the sum of squares of the differences between all data points and the mean.

15. The system for constructing a digital twin model of a camera based on real-time data fitting according to claim 11, 12, or 13, characterized in that, It also includes a verification module for periodically verifying the digital twin model, including revising the fitting parameters and recording parameter changes, and triggering an alarm when the parameter change rate exceeds a threshold.

16. The system for constructing a digital twin model of a camera based on real-time data fitting according to claim 11, 12, or 13, characterized in that, In the early warning module, the conditions for triggering an alarm include at least one of the following: The standard deviation between the digital twin model's predicted values ​​and the real-time collected data exceeds the limit; Real-time collected values ​​exceeded limits; The digital twin model's predicted values ​​exceeded the limit; The coefficients of the digital twin model have exceeded the limit.

17. The system for constructing a digital twin model of a camera based on real-time data fitting according to claim 11, 12, or 13, characterized in that, In the data acquisition module, the operating data also includes electrical parameters, mechanical parameters, and thermal parameters; the electrical parameters include stator current, voltage, power factor, active power, excitation current, and voltage; the mechanical parameters include rotor speed and bearing displacement; the thermal parameters include stator winding temperature, rotor winding temperature, stator cooling outlet temperature, rotor cooling outlet temperature, synchronous condenser housing temperature, and lubricating oil temperature.

Images