Generator fault three-dimensional visual early warning and construction method

Abstract

The invention discloses a three-dimensional visual early warning and construction method for generator faults, and relates to the technical field of digital twinning, and the method comprises the steps: obtaining generator equipment parameters, operation data and environment data through a sensor, and constructing a generator digital twinning model through the combination of the digital twinning technology and a three-dimensional visual technology; continuously collecting operation data of the generator to correct the constructed digital twinborn model of the generator until the constructed digital twinborn model of the generator can be matched with an actual generator; analyzing operation data simulated by the corrected generator digital twinborn model, identifying generator fault conditions, mapping the identified generator fault conditions and the generator digital twinborn model in real time, displaying the identified generator fault conditions to equipment parts with fault conditions of working personnel in a three-dimensional visual form, and sending early warning reminding to the working personnel. And a worker is informed of timely maintenance.

Description

Technical Field

[0001] The present invention belongs to the technical field of digital twins, and specifically relates to a three-dimensional visualization early warning and construction method for generator failure. Background Art

[0002] With the transformation of energy structure and the enhancement of environmental awareness around the world, the widespread use of electric energy has gradually replaced the traditional energy structure, whether it is wind power generation, thermal power generation or nuclear power plant, the most important thing is the generator, the operating status of the generator directly affects the power generation efficiency and reliability. Therefore, the research on the status monitoring and fault diagnosis technology of the generator is of great significance.

[0003] Generator status detection and fault diagnosis technology, as the key research object in the field of new energy in the current society, has received much attention both at home and abroad. Existing generator fault diagnosis and operation status detection methods have certain limitations. On the one hand, traditional methods mainly focus on the real-time operation status of the generator; on the other hand, some detection methods rely on daily inspections of generator components, or indirectly judge the operating status of the generator by monitoring the generator's power generation, operating voltage, and operating current. These methods cannot directly monitor and warn the generator itself and the equipment components of the generator, and it is difficult to meet the needs of accurate and efficient fault diagnosis and warning. In order to solve the above problems, the present invention provides a three-dimensional visualization warning and construction method for generator faults. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a three-dimensional visual early warning and construction method for generator failure, which solves the problem that the prior art is difficult to monitor in real time and warn of generator failure in advance.

[0005] The purpose of the present invention can be achieved by the following technical solutions: A three-dimensional visual early warning method for generator faults, characterized in that the method comprises the following steps: S1. Obtain generator equipment parameters, operation data, and environmental data and use digital twin technology to build a generator digital twin model; S2. Continuously collect actual generator operation data to calibrate the constructed generator digital twin model until the constructed generator digital twin model matches the actual generator; S3. Obtain the operating data of the generator digital twin model simulation operation for analysis, identify the generator fault condition, map the identified generator fault condition with the generator digital twin model in real time, and issue early warning reminders to the staff.

[0006] As a further solution of the present invention, the specific method of continuously collecting generator operation data to calibrate the constructed generator digital twin model is: S21. Based on the constructed digital twin model of the generator, As the starting moment, the simulation runs a simulation cycle. The duration of the simulation cycle is set by the staff. A simulation cycle contains j moments; S22. In one simulation operation cycle, the operation data obtained by simulating the operation of the digital twin model of the generator is obtained, and a scatter plot of the change of each operation data over time is drawn on a two-dimensional coordinate system, with time as the horizontal axis and the value of each type of operation data as the vertical axis, as a set of change scatter plot set A; S23, obtain the actual generator at the current time As the starting moment, run the operation data of a simulation operation cycle, and repeat the method described in S32 to obtain a set of change scatter plots B; S24, taking the change scatter plot set A and the change scatter plot set B as the same group, and recording them as a double-sample scatter plot feature vector; S25, based on the obtained double sample scatter plot feature vector, take a moment in two of the voltage change scatter plots The two voltage values ​​corresponding to the vertical axis and ; Pick The voltage value corresponding to the previous n consecutive moments The straight-line distance between ,Pick The voltage value corresponding to the next n consecutive moments The straight-line distance between ,by As the radius length Draw a circle, obtain the sum of all voltage values ​​within the circle and average them to get the average voltage value ; Similarly, get the voltage average value ; like and The difference rate exceeds the difference rate preset by the staff , think that the moment The two voltage change scatter plots do not match, and the voltage parameters of the actual generator within this circle are overwritten with the voltage parameters of the generator digital twin model simulation operation, where n is the value preset by the staff, and n is less than i, and i is less than j; S26. Convert the two voltage change scatter plots after overlay processing into two voltage change line graphs, and fit them into the same voltage change line graph for further analysis.

[0007] As a further solution of the present invention, the specific method of converting the two voltage change scatter plots after the overlay processing into two voltage change line graphs and fitting them into the same voltage change line graph for further analysis in step S26 is: S31, based on the two voltage change scatter plots, connect any two adjacent points with the shortest line to obtain two voltage change line graphs, and fit the two voltage change line graphs into the same voltage change line graph to obtain a first line graph. With the second fold line , corresponding to the actual generator and the generator digital twin model respectively; S32, obtaining the first moment after the start moment as the initial moment, constructing an initial moment line passing through the initial moment and perpendicular to the horizontal axis, calculating the initial moment line and the vertical axis, the first fold line And the second fold line The area characteristics formed between ; S33, copy the initial time line to the moment after the initial time, record it as the second time line, calculate the initial time line and the second time line, the first fold line And the second fold line The area characteristics formed between ; S34, repeat step S33 until the last moment of this simulation operation cycle, and obtain a total of j area features, recorded as , and compare the obtained j area features with the area feature thresholds preset by the staff Compare, if any area feature , then it is considered that the moment The voltage parameters of the generator digital twin model simulation operation at the previous moment do not match the voltage parameters of the actual generator, and the voltage parameters of the generator digital twin model simulation operation are overwritten with the voltage parameters of the actual generator; S35. Place the voltage parameters of the generator digital twin model simulation operation after the overwriting operation in the generator digital twin model to run a simulation operation cycle and perform data correction.

[0008] As a further solution of the present invention, the voltage parameters of the simulated operation of the digital twin model of the generator after the overwriting operation are placed in the digital twin model of the generator to run a simulated operation cycle, and data correction is performed. In addition, other operating data are processed according to the method of processing voltage parameters to perform data correction on the digital twin model of the generator.

[0009] As a further solution of the present invention, a specific method for determining whether the constructed digital twin model of the generator is compatible with the actual generator is as follows: If the operating data of the digital twin model of the generator simulated in C consecutive simulation operation cycles are consistent with the actual generator, the digital twin model of the generator is considered to be consistent with the actual generator. The value of C is set by the staff.

[0010] As a further solution of the present invention, the specific method of identifying the generator fault condition in step S3 is: The generator digital twin model is divided according to the total number of equipment components, and u equipment components are obtained after division, which are recorded as , where u is a count index, indicating the total number of equipment components, For u equipment components: Any one of , and v is also a counting index, v starts from 1, and the maximum value of v is u; The digital twin model of the generator is simulated and runs a simulation operation cycle; Extraction of equipment parts At any time during the simulation cycle The time series running data: ,in is a data vector, Representative parts At the moment The values ​​of all running data at the time, there are q different running data in total; Based on j moments, we get the equipment components The j-time series running data is For example, Associated datasets , calculate the mean of the data set and standard deviation , through the normalization formula: The dataset All data in are normalized to the interval of 0 and 1 to obtain a normalized data set , and similarly, we get the normalized The respective associated normalized datasets; Put the normalized running data into the time series running data corresponding to each time, and get the time Normalized time series running data: Similarly, we obtain j normalized time series operation data corresponding to j moments, and fit each normalized time series operation data into the equipment component The equipment status characteristics at that moment, and draw the equipment component operation status change curve on the two-dimensional coordinate system and obtain the fluctuation degree of the curve; The simulation is run for several simulation cycles to obtain the fluctuation degree of several curves; The fluctuation degree of several curves is compared with the fluctuation degree threshold set by the staff. If the fluctuation degree of any curve exceeds the fluctuation degree threshold set by the staff, it is determined that the equipment component in the simulation operation cycle A fault condition occurs; According to the above method, fault identification processing is performed on all equipment components in the actual generator, the component equipment with fault conditions is identified, and the simulated operation cycle of the component equipment with fault conditions is obtained.

[0011] As a further solution of the present invention, the specific method of mapping the identified generator fault condition with the generator digital twin model in real time and issuing an early warning reminder to the staff is: For equipment components that have faults, the generator digital twin model uses flashing highlight blocks of different colors from other equipment components to inform staff when the equipment component will fail, and provides an early warning to remind staff to repair or replace the equipment component in time before it fails.

[0012] As a further solution of the present invention, a method for constructing a three-dimensional visualization of a generator fault includes the following steps: Obtain generator equipment parameters, which include equipment components and the three-dimensional shapes and positional relationships between equipment components, and build a generator digital twin model based on the three-dimensional modeling software combined with the acquired equipment parameters; Acquire generator operation data, including voltage, current, speed, vibration, stator temperature, and rotor temperature, and map them to the generator digital twin model in real time through digital twin technology to achieve simulated operation; Environmental parameters are obtained, including ambient temperature, ambient humidity, ambient air pressure, and ambient electric field strength, and are fitted in real time into the operating environment of the generator digital twin model simulation operation through digital twin technology.

[0013] Beneficial effects of the present invention: (1) The present invention provides a method for constructing a three-dimensional visual warning of generator faults by combining actual generators with digital twin technology, constructing a digital twin model of the generator through digital twin technology, and providing a method for correcting the digital twin model of the generator based on actual generator operation data until the digital twin model of the generator is highly consistent with the actual generator. This precise modeling and mapping method can ensure a comprehensive and accurate understanding of the generator, and provide a solid and reliable basic data and model reference for subsequent fault diagnosis and warning; (2) The present invention provides a method for identifying fault conditions based on analyzing the operating data simulated by the digital twin model of the generator after correction processing, and maps the fault information with the digital twin model of the generator in real time, and promptly issues early warning reminders to the staff. Different from the previous methods of indirect monitoring or only detecting faults that have occurred, this method realizes early warning of potential faults, enabling the staff to take timely measures, greatly improving the reliability and safety of the generator operation, and reducing the downtime, equipment damage risk and maintenance costs caused by faults; (3) The present invention constructs a three-dimensional visualized generator model through digital twin technology, and uses a three-dimensional visualized method to warn of faults, so that the staff can more intuitively understand the fault location of the generator and the faulty equipment components. Compared with traditional complex data reports or indicator light alarms, the readability and comprehensibility of fault information are improved, which facilitates the staff to quickly make accurate judgments and decisions based on the faulty equipment components, thereby improving work efficiency and response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention will be further described below in conjunction with the accompanying drawings.

[0015] Figure 1 is a schematic flow chart of the method described in Example 1 of the present invention; Figure 2 Schematic diagram of the process of the method described in Example 2 of the present invention; Figure 3 Schematic diagram of the process of the method described in Example 3 of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0017] Example 1 A three-dimensional visualization early warning method for generator faults, such as Figure 1 As shown, specifically including the following: Step 1: Obtain the generator equipment parameters, operating data, and environmental data; the generator equipment parameters include (including but not limited to) the generator housing, rotor, stator, bearings, and the three-dimensional shape and position relationship between them, which are obtained through the generator appearance description and 3D point cloud measurement technology. The purpose of obtaining the equipment parameters is to use , The three-dimensional composition technology is used to create an accurate geometric model according to the actual size and equipment components of the corresponding generator, and the digital twin technology is used to further build a digital twin model of the generator based on the constructed three-dimensional model; The operating data is collected by various sensors installed on the actual generator when the actual generator is working normally, including: voltage, current, speed, vibration, stator temperature, and rotor temperature. These data collected from the actual generator are mapped to the corresponding generator digital twin model through digital twin technology to achieve simulated operation; Environmental data refers to the actual working environment of the current generator, which is collected by sensors in the current working environment, including ambient temperature, ambient humidity, ambient air pressure, and ambient electric field strength. The measured environmental data is fitted into the working environment of the generator digital twin model through digital twin technology; What needs to be explained here is that environmental data can be summarized into certain rules through periodic data, and the summarized environmental data is applied to the digital twin model of the generator to simulate the environmental data in the future. Before the simulation, correction processing is required. The correction processing of the environmental data is regarded as prior art and will not be elaborated in this article.

[0018] Step 2: Continuously collect the operating data of the actual generator to calibrate the constructed digital twin model of the generator. The calibration method is to use the digital twin model of the generator to simulate the time of a simulated operation cycle with the current time as the start time, and obtain the operating data obtained by the simulated operation of the digital twin model of the generator during this simulated operation cycle, and then obtain the operating data of the actual generator during this simulated operation cycle. The two data are compared and the difference between the two operating data is analyzed. If the difference is large, the digital twin model of the generator needs to be calibrated. If the operating data of several consecutive simulated operations are consistent with the operating data of the actual generator operation, then the calibration is considered successful. At this time, it is considered that the operating data of the simulated operation of the digital twin model of the generator is equivalent to the operating data of the actual generator.

[0019] Step 3: Analyze the operating data of the simulated operation of the calibrated generator digital twin model, identify the generator fault, and map the identified generator fault with the generator digital twin model in real time, showing the location of the fault to the staff in a three-dimensional visual form, and issuing an early warning reminder to the staff.

[0020] Example 2 This embodiment further explains step 2 in embodiment 1 in detail. On the basis of step 2, a method for calibrating a generator digital twin model based on actual generator operation data is disclosed, such as Figure 2 As shown, specifically including the following: Based on step one in Example 1, a preliminary constructed digital twin model of the generator can be obtained. In this embodiment, the digital twin model of the generator will be calibrated to make the constructed digital twin model of the generator match the actual generator.

[0021] Get the current time point, that is, the current moment ,Will As the starting moment, a period of time is extended into the future. The duration of this period is set by the staff and is named the simulation operation cycle. A simulation operation cycle contains j moments (it needs to be explained that the time interval between two adjacent moments is not one second, and should be determined by the staff based on actual conditions. If a higher-precision digital twin model of the generator is required, the time between adjacent moments can be appropriately shortened, otherwise it can be extended; in addition, j moments refer to moments other than the starting moment, that is, the moment after the starting moment is taken as the first moment (the first moment is also called the initial moment, which is different from the starting moment), until a simulation operation cycle, that is, the jth moment, but the various data measured still include the operation data and environmental data from the start moment to the first moment).

[0022] Using the digital twin model of the generator that has been initially constructed, As the starting moment, a simulation operation cycle is simulated, and the operation data of the digital twin model of the generator in this simulation operation cycle is recorded in real time. For the voltage, current, speed, vibration, stator temperature, and rotor temperature in the operation data, a two-dimensional coordinate system is used to map the change scatter diagram. The construction method of the two-dimensional coordinate system is: with time (this simulation operation cycle, the scale is a moment) as the horizontal axis and the values ​​of various operation data (such as voltage value, current value, etc.) as the vertical axis, a scatter diagram of the change of each operation data over time is drawn; The voltage change scatter plot, current change scatter plot, speed change scatter plot, vibration change scatter plot, stator temperature change scatter plot, and rotor temperature change scatter plot are obtained respectively, and the scatter plots of the above generator digital twin model are taken as a group, which is called the change scatter plot set A.

[0023] Then make the actual generator have the same initial moment as the generator digital twin model The time for starting a simulation operation cycle is the same as the time for the simulation operation cycle of the actual generator and the simulation operation cycle of the generator digital twin model.

[0024] The operating data of the actual generator during operation is recorded in real time, and the corresponding voltage change scatter plot, current change scatter plot, speed change scatter plot, vibration change scatter plot, stator temperature change scatter plot, and rotor temperature change scatter plot are obtained according to the processing method of the generator digital twin model. The scatter plots of the above actual generators are taken as a group and called the change scatter plot set B.

[0025] A set of change scatter plots A drawn from the operating data obtained by simulating the operation of the digital twin model of the generator within the same simulation operation cycle and a set of change scatter plots B drawn from the operating data obtained by the actual operation of the generator are treated as the same set of data and recorded as the double-sample scatter plot feature vector.

[0026] Based on the obtained double sample scatter plot feature vector, take two change scatter plots of any of the same parameters. Here, take the voltage change scatter plot as an example. Since the horizontal axes are all time units and the scales are the same, they both correspond to the time of the same simulation operation cycle. Therefore, take any moment in this simulation operation cycle. , the two voltage values ​​corresponding to the vertical axis on the two voltage change scatter plots and ; For voltage value , get The voltage value corresponding to the previous n consecutive moments The straight-line distance between ,as well as The voltage value corresponding to the next n consecutive moments The straight-line distance between , and then As the radius length , draw a circle and get the voltage value Neighborhood radius The sum of all voltage values ​​within the range (i.e. all voltage values ​​within this circular area) is averaged to obtain the average voltage value. ; Similarly, get the voltage value Neighborhood radius The voltage average of all voltage values ​​within the range ; What needs to be explained here is that by calculating the average voltage within the neighborhood radius, it is possible to quickly and effectively identify whether abnormal data occurs. If the digital twin model of the generator is consistent with the actual generator, then the average voltage values ​​within the neighborhood radius of their voltage values ​​at a certain moment should be smaller. If the difference is larger, it means that there is abnormal data within the neighborhood radius of the voltage value at that moment simulated by the digital twin model of the generator. If the abnormal data is missing data, then the average voltage value within the neighborhood radius of the voltage value at that moment simulated by the digital twin model of the generator will be smaller. If the abnormal data is redundant data, then the average voltage value within the neighborhood radius of the voltage value at that moment simulated by the digital twin model of the generator will be larger, so as to detect whether abnormal data occurs.

[0027] The voltage average value is calculated by the staff and Set a preset difference rate , if the average voltage and The difference rate between the two exceeds the preset difference rate , then it is regarded as the moment The two voltage change scatter plots do not match (that is, the voltage parameters in the operating data simulated by the generator digital twin model at this moment do not match the voltage parameters in the actual generator operating data). As the radius length The voltage parameters within the range of the circle are used to preliminarily cover the voltage parameters of the simulated operation of the digital twin model of the generator, where n is the value preset by the staff, and n is less than i, and i is less than j.

[0028] Perform preliminary coverage processing on the voltage parameters at each moment of the digital twin model simulation operation, and connect any two adjacent points in the voltage change scatter diagram obtained by the digital twin model simulation operation after preliminary coverage processing and the voltage change scatter diagram obtained by the actual generator operation using the shortest line, and finally obtain two voltage change line graphs; Because the horizontal and vertical axes of the two voltage change line graphs are the same, the two voltage change line graphs can be fitted into the same voltage change line graph through image processing technology, that is, there are two broken lines in one voltage change line graph, which are recorded as the first broken line With the second fold line , which correspond to the voltage parameters of the actual generator and the generator digital twin model respectively; Get the moment after the start moment in this simulation operation cycle as the initial moment (first moment), and construct a straight line passing through the first moment and perpendicular to the horizontal axis on the voltage change line graph with two broken lines. The straight line passes through the first broken line. With the second fold line , recorded as the initial time line (first time line); Initial time line, vertical axis, first fold line And the second fold line The closed area is formed between them, and the area of ​​the closed area is calculated and recorded as the area feature ; It should be explained here that the larger the area of ​​the closed area, the greater the difference rate between the voltage parameters of the actual generator and the generator digital twin model corresponding to this area, and vice versa. It should be noted here that the first broken line And the second fold line There will be intersections between them, resulting in multiple closed areas in adjacent moments. For example, between the initial time line and the vertical axis, the first broken line With the second fold line If multiple intersections occur, multiple closed areas will appear. The sum of the areas of the multiple closed areas is calculated and the sum is recorded as the area feature. .

[0029] Then copy the initial time line to the moment after the first moment (i.e. the second moment), mark the straight line passing through the second moment as the second time line, and calculate the initial time line, the second time line, and the first broken line. And the second fold line The area of ​​the closed area formed between ; Continue to copy the initial timeline backward to the moment after the second moment (the third moment), and calculate the third timeline, the second timeline, and the first fold line And the second fold line The area of ​​the closed area formed between ; Until the initial time line is copied to the last moment (jth moment) of this simulation operation cycle, and the j area features calculated are recorded as , take any one of the area features Area feature thresholds preset by staff Compare, if the area feature Greater than the preset area feature threshold , then it is determined that at this moment The voltage parameters of the generator digital twin model simulation operation between the previous moment and the actual generator voltage parameters do not match. The voltage parameters of the actual generator between the previous moment overwrite the voltage parameters of the generator digital twin model simulation operation; The voltage data of the generator digital twin model simulation operation after two overlay operations is fitted into the generator digital twin model and a simulation operation cycle is repeated to perform data correction.

[0030] What needs to be explained here is that the voltage data of the simulated operation of the digital twin model of the generator after two overwriting operations is fitted into the digital twin model of the generator and a simulation operation cycle is repeatedly run for data correction, instead of directly using the actual generator operation data to fit into the digital twin model simulation for data correction. The reason is: the actual generator operation data is directly measured by the sensor, and the data contains noise, errors or outliers. If these data are directly used for the simulation operation of the digital twin model, it will cause the digital twin model to learn the wrong pattern or deviate from the actual operation rules, thereby affecting the accuracy and reliability of the digital twin model. By first performing anomaly detection on the simulated operation data of the digital twin model and covering part of the abnormal data with actual data, it can ensure that the data used to correct the simulated operation is more accurate and reliable. After several corrections, a digital twin model of the generator that is relatively close to the actual operating state of the actual generator can be obtained.

[0031] All voltage parameters after two overwriting operations are fitted into the digital twin model of the generator and a simulation operation cycle is repeatedly run to perform data correction. Other operating data are processed according to the method of processing voltage parameters, and other operating data are fitted into the digital twin model of the generator and a simulation operation cycle is repeatedly run to perform data correction.

[0032] If all parameters in the operating data of the generator digital twin model simulated within C consecutive simulation operation cycles can match the actual generator, it is deemed that the generator digital twin model can match the actual generator at this time, and the next step can be carried out. The value of C mentioned here is determined by the staff.

[0033] Example 3 This embodiment further explains step 3 in embodiment 1 in detail. Based on step 3, a method is disclosed for analyzing the operation data obtained by simulating the operation of the digital twin model of the generator, identifying the fault situation and notifying the staff, such as Figure 3 As shown, specifically including the following: Based on the generator digital twin model after correction processing in Example 2, the staff divides the generator digital twin model according to the equipment components based on the type of actual generator and the equipment parameters of the actual generator, and records the divided equipment components in sequence as ,in is the counting index, starting from 1, and the maximum value is , The total number of equipment components divided into the digital twin model of the generator, Any equipment component in the digital twin model of the generator.

[0034] Based on the corrected digital twin model of the generator, the current time is used as the start time to simulate a simulation operation cycle.

[0035] For any equipment component in the digital twin model of the generator within the simulation operation cycle , extract the device components At any moment The time series running data: ,in is a data vector, represented as , Representative equipment parts At the moment The values ​​of all operating data at that time.

[0036] For equipment parts For example, there are q different operating data. The q operating data associated with each moment form a data vector and together constitute a moment series operating data. The moment series operating data of a moment can represent the operating status of the equipment component at that moment. Different moment series operating data reflect the operating status of the equipment component. Operational status at different times; It should be explained here that q is not a fixed value, but a variable counting index, which is determined according to the number of operating data associated with different equipment components.

[0037] Based on the j moments in a simulation operation cycle, a total of components can be obtained In j time series running data composed of j different time periods, the values ​​of j same running data are obtained from the j time series running data. For example, we can get Associated Datasets , calculate the average value of this set of data and standard deviation , and through the normalization formula: Will Associated Datasets All data in are normalized to the interval of 0 and 1, and the normalized Associated normalized dataset According to the above method, we can get The normalized data sets after normalization associated with each other; The normalized running data is placed in the time series running data corresponding to each time, and the time is obtained. Normalized time series running data: Similarly, we obtain j normalized time series operation data corresponding to j moments, and divide the equipment components into The normalized time series operation data corresponding to each moment is fitted into the equipment components Characteristics of the device state at that moment; What needs to be explained here is that all the operating parameter values ​​in the normalized moment series operating data jointly determine the operating status of the equipment component. These operating parameter values ​​are regarded as a data vector from a mathematical perspective at each moment. The serialization of the data vector over time fully reflects the operating status of the equipment. Therefore, by analyzing these multi-dimensional data vectors, the characteristic information of the operating status of the equipment components can be extracted. After normalization, data of different dimensions and magnitudes are unified to the same scale, eliminating the dimensional differences and magnitude differences between the data, which allows direct fitting between different operating data.

[0038] All equipment status features within this simulation operation cycle are counted, and a total of j equipment status features after fitting the j normalized time series operation data are obtained. Then, time is used as the horizontal axis of the two-dimensional coordinate system, and the fitted equipment status features are used as the vertical axis of the two-dimensional coordinate system. The obtained j equipment status features are plotted on the two-dimensional coordinate system and fitted into an equipment component operation status change curve. The horizontal axis of the two-dimensional coordinate system is time, representing a simulation operation cycle of this simulation operation, and the scale is each moment. The vertical axis is the equipment status feature, representing the equipment component at the corresponding moment. The running status of

[0039] Based on the equipment components determined during the current simulation run cycle The operating status change curve of the equipment components is obtained to obtain the fluctuation degree of the curve (the fluctuation degree of the curve can be analyzed by Fourier transform to analyze the frequency component of the curve, and the existing technology will not be described in detail). The smaller the fluctuation degree, the more stable the current equipment component is, and vice versa. Based on this principle, the generator digital twin model is used to simulate and run several simulation operation cycles to obtain several components. The equipment component operation status change curves are obtained to obtain the fluctuation degree of several curves; The staff sets a fluctuation degree threshold for the fluctuation degree of the curve based on the actual generator in the actual working environment, which is used to distinguish the equipment components in the normal operating state from the equipment components in the abnormal operating state, and compares the fluctuation degrees of several obtained curves with the fluctuation degree threshold set by the staff; If the fluctuation degree of any curve exceeds the fluctuation degree threshold set by the staff, it is determined that the equipment component is A fault condition has occurred.

[0040] According to the above method, all equipment components divided by the actual generator are subjected to fault identification processing, and the equipment components with fault conditions are recorded; Continue to use the digital twin model of the generator to simulate several simulation operation cycles, analyze the simulated operation data, and identify equipment components that will fail in a future simulation operation cycle.

[0041] Based on the equipment components that are determined to be about to fail, the generator digital twin model uses highlight blocks of different colors from other equipment components for rendering and annotation, flashing to remind staff when the equipment component will fail; For example, there is a generator whose digital twin model simulates the operation to the sixth simulation operation cycle. The rotor inside the generator fails. The staff presets a simulation operation cycle length of one week. The digital twin model renders the rotor in a color that is clearly distinguished from other equipment components and flashes to remind the staff. When the staff clicks to view the rotor, the staff will be informed that the rotor will fail in the sixth week after the current time and the equipment components need to be replaced or repaired before the sixth week.

[0042] Some of the data in the formulas described above are numerically calculated by removing their dimensions. At the same time, the contents not described in detail in this specification belong to the existing technologies well known to those skilled in the art.

[0043] The above contents are merely examples and explanations of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the invention or exceed the scope defined by the claims, they shall all fall within the protection scope of the present invention.

[0044] It should be stated that all user data collected in this application is collected with the user's consent and authorization. The use of user data is legal and compliant, and the use and processing of user data complies with the relevant laws, regulations and standards of the relevant regions.

Claims

1. A three-dimensional visualization early warning method for generator failure, characterized in that: The method comprises the following steps: S1. Obtain generator equipment parameters, operation data, and environmental data and use digital twin technology to build a generator digital twin model; S2. Continuously collect actual generator operation data to calibrate the constructed generator digital twin model until the constructed generator digital twin model matches the actual generator; S3. Obtain the operating data of the generator digital twin model simulation operation for analysis, identify the generator fault condition, map the identified generator fault condition with the generator digital twin model in real time, and issue early warning reminders to the staff.

2. A three-dimensional visualization early warning method for generator failure according to claim 1, characterized in that: The specific method of continuously collecting generator operation data and correcting the constructed generator digital twin model is as follows: S21. Based on the constructed digital twin model of the generator, As the starting moment, the simulation runs a simulation cycle. The duration of the simulation cycle is set by the staff. A simulation cycle contains j moments; S22. In one simulation operation cycle, the operation data obtained by simulating the operation of the digital twin model of the generator is obtained, and a scatter plot of the change of each operation data over time is drawn on a two-dimensional coordinate system, with time as the horizontal axis and the value of each type of operation data as the vertical axis, as a set of change scatter plot set A; S23, obtain the actual generator at the current time As the starting moment, run the operation data of a simulation operation cycle, and repeat the method described in S32 to obtain a set of change scatter plots B; S24, taking the change scatter plot set A and the change scatter plot set B as the same group, and recording them as a double-sample scatter plot feature vector; S25, based on the obtained double sample scatter plot feature vector, take a moment in two of the voltage change scatter plots The two voltage values ​​corresponding to the vertical axis and ; Pick The voltage value corresponding to the previous n consecutive moments The straight-line distance between ,Pick The voltage value corresponding to the next n consecutive moments The straight-line distance between ,by As the radius length Draw a circle, obtain the sum of all voltage values ​​within the circle and average them to get the average voltage value ; Similarly, get the voltage average value ; like and The difference rate exceeds the difference rate preset by the staff , think that the moment The two voltage change scatter plots do not match, and the voltage parameters of the actual generator within this circle are overwritten with the voltage parameters of the generator digital twin model simulation operation, where n is the value preset by the staff, and n is less than i, and i is less than j; S26. Convert the two voltage change scatter plots after overlay processing into two voltage change line graphs, and fit them into the same voltage change line graph for further analysis.

3. A three-dimensional visualization early warning method for generator failure according to claim 2, characterized in that: The specific method of converting the two voltage change scatter plots after the overlay processing into two voltage change line graphs and fitting them into the same voltage change line graph for further analysis in step S26 is: S31, based on the two voltage change scatter plots, connect any two adjacent points with the shortest line to obtain two voltage change line graphs, and fit the two voltage change line graphs into the same voltage change line graph to obtain a first line graph. With the second fold line , corresponding to the actual generator and the generator digital twin model respectively; S32, obtaining the first moment after the start moment as the initial moment, constructing an initial moment line passing through the initial moment and perpendicular to the horizontal axis, calculating the initial moment line and the vertical axis, the first fold line And the second fold line The area characteristics formed between ; S33, copy the initial time line to the moment after the initial time, record it as the second time line, calculate the initial time line and the second time line, the first fold line And the second fold line The area characteristics formed between ; S34, repeat step S33 until the last moment of this simulation operation cycle, and obtain a total of j area features, recorded as , and compare the obtained j area features with the area feature thresholds preset by the staff Compare, if any area feature , then it is considered that the moment The voltage parameters of the generator digital twin model simulation operation at the previous moment do not match the voltage parameters of the actual generator, and the voltage parameters of the generator digital twin model simulation operation are overwritten with the voltage parameters of the actual generator; S35. Place the voltage parameters of the generator digital twin model simulation operation after the overwriting operation in the generator digital twin model to run a simulation operation cycle and perform data correction.

4. A three-dimensional visualization early warning method for generator failure according to claim 3, characterized in that: The voltage parameters of the simulated operation of the digital twin model of the generator after the overwriting operation are placed in the digital twin model of the generator to run a simulated operation cycle for data correction. In addition, other operating data are processed according to the method of processing voltage parameters to perform data correction on the digital twin model of the generator.

5. A three-dimensional visualization early warning method for generator failure according to claim 4, characterized in that: The specific judgment method for whether the constructed generator digital twin model can match the actual generator is as follows: If the operating data of the digital twin model of the generator simulated in C consecutive simulation operation cycles are consistent with the actual generator, the digital twin model of the generator is considered to be consistent with the actual generator. The value of C is set by the staff.

6. A three-dimensional visualization early warning method for generator failure according to claim 5, characterized in that: The specific method of identifying the generator fault condition described in step S3 is: The generator digital twin model is divided according to the total number of equipment components, and u equipment components are obtained after division, which are recorded as , where u is a count index, indicating the total number of equipment components, For u equipment components: Any one of , and v is also a counting index, v starts from 1, and the maximum value of v is u; The digital twin model of the generator is simulated and runs a simulation operation cycle; Extraction of equipment parts At any time during the simulation cycle The time series running data: ,in is a data vector, Representative parts At the moment The values ​​of all running data at the time, there are q different running data in total; Based on j moments, we get the equipment components The j-time series running data is For example, Associated datasets , calculate the mean of the data set and standard deviation , through the normalization formula: The dataset All data in are normalized to the interval of 0 and 1 to obtain a normalized data set , and similarly, we get the normalized The respective associated normalized datasets; Put the normalized running data into the time series running data corresponding to each time, and get the time Normalized time series running data: Similarly, we obtain j normalized time series operation data corresponding to j moments, and fit each normalized time series operation data into the equipment component The equipment status characteristics at that moment, and draw the equipment component operation status change curve on the two-dimensional coordinate system and obtain the fluctuation degree of the curve; The simulation is run for several simulation cycles to obtain the fluctuation degree of several curves; The fluctuation degree of several curves is compared with the fluctuation degree threshold set by the staff. If the fluctuation degree of any curve exceeds the fluctuation degree threshold set by the staff, it is determined that the equipment component in the simulation operation cycle A fault condition occurs; According to the above method, fault identification processing is performed on all equipment components in the actual generator, the component equipment with fault conditions is identified, and the simulated operation cycle of the component equipment with fault conditions is obtained.

7. A three-dimensional visualization early warning method for generator failure according to claim 6, characterized in that: The specific method of mapping the identified generator fault condition with the generator digital twin model in real time and issuing early warning reminders to the staff is: For equipment components that have faults, the generator digital twin model uses flashing highlight blocks of different colors from other equipment components to inform staff when the equipment component will fail, and provides an early warning to remind staff to repair or replace the equipment component in time before it fails.

8. A method for constructing three-dimensional visualization of generator faults, characterized in that: The method includes the following: Obtain generator equipment parameters, which include equipment components and the three-dimensional shapes and positional relationships between equipment components, and build a generator digital twin model based on the three-dimensional modeling software combined with the acquired equipment parameters; Acquire generator operation data, including voltage, current, speed, vibration, stator temperature, and rotor temperature, and map them to the generator digital twin model in real time through digital twin technology to achieve simulated operation; Environmental parameters are obtained, including ambient temperature, ambient humidity, ambient air pressure, and ambient electric field strength, and are fitted in real time into the operating environment of the generator digital twin model simulation operation through digital twin technology.

Images