A petrochemical terminal digital twin management method and system

Abstract

The present invention discloses a petrochemical terminal digital twin management method and system, which relates to the technical field of digital twin management, including obtaining multi-dimensional data of the petrochemical terminal and preprocessing the data; using empirical mode decomposition technology and noise processing methods to further decompose and process the preprocessed data to obtain multi-dimensional signals; constructing a digital twin model based on the multi-dimensional signals to identify whether the terminal equipment is operating normally; and monitoring whether the equipment is operating normally in real time through the digital twin model and issuing early warnings. The present invention uses empirical mode decomposition technology to decompose complex multi-dimensional signals into multiple mode components, effectively extract signal characteristics of different frequencies and energy distributions, and further optimize the signal quality through noise processing methods, thereby improving the resolution of the signal. At the same time, it can accurately identify minor anomalies in the operation of the equipment and enhance the sensitivity of the digital twin model to the operating status of the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of digital twin management, and in particular to a petrochemical terminal digital twin management method and system. Background Art

[0002] As an important hub for the transportation of petroleum and chemical products, the normal operation and safety of equipment at petrochemical terminals are of vital importance. Digital twin technology has gradually become an important tool for equipment management and maintenance. Digital twin technology creates a virtual model that is synchronized with the physical equipment in real time to reflect whether the equipment is operating normally, and uses multi-dimensional data collected by sensors to monitor, analyze and predict the equipment. In petrochemical terminals, the complexity of equipment operation and the variability of environmental conditions make real-time monitoring and status prediction an extremely challenging task.

[0003] There are still some shortcomings in the application of existing digital twin technology in petrochemical terminals. Petrochemical terminals have a wide variety of equipment, and the collected signals often have complex characteristics such as multi-dimensionality, nonlinearity, and non-stationarity. Traditional signal processing methods are difficult to effectively extract useful information, resulting in insufficient accuracy and reliability of digital twin models. Secondly, existing technologies are not yet perfect in noise processing. Factors such as environmental noise and sensor errors will affect data quality, and thus affect the accuracy of digital twin models. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a petrochemical terminal digital twin management method and system to solve the problem that traditional signal processing methods are difficult to effectively extract useful information from oil terminal equipment, resulting in insufficient accuracy and reliability of the digital twin model.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a petrochemical terminal digital twin management method, which includes:

[0008] Obtain multi-dimensional data of petrochemical terminals and pre-process them;

[0009] The empirical mode decomposition technology and noise processing method are used to further decompose and process the pre-processed data to obtain multi-dimensional signals;

[0010] Build a digital twin model based on multi-dimensional signals to identify whether terminal equipment is operating normally;

[0011] Use digital twin models to monitor the normal operation of terminal equipment in real time and issue early warnings;

[0012] The digital twin model is constructed based on multi-dimensional signals to identify whether the terminal equipment is operating normally, including:

[0013] Use physical equations to build basic models and create terminal equipment geometry;

[0014] For the parts that are difficult to accurately describe with the physical model, a data-driven model is introduced to compensate and enhance them, combining the physical model with the data-driven model to form a unified hybrid model;

[0015] Map multidimensional signals to the input parameters of the hybrid model to build a digital twin model;

[0016] Monitor the data of actual terminal equipment in real time, map the monitored data information of actual terminal equipment into the digital twin model, ensure that the status of the digital twin model is consistent with that of the actual terminal equipment, and the equipment status changes in the digital twin model are also fed back to the actual terminal equipment.

[0017] As a preferred solution of the digital twin management method for petrochemical terminals described in the present invention, the preprocessing includes outlier processing, noise filtering and standardization.

[0018] As a preferred solution of the digital twin management method for petrochemical terminals of the present invention, the empirical mode decomposition technology and the noise processing method are used to further decompose and process the pre-processed data, including:

[0019] Define the standardized time series data of each sensor dimension as the signal to be processed, traverse each sampling point of the signal to be processed, and detect its local maximum and local minimum;

[0020] All points that meet the maximum and minimum conditions are recorded as the maximum point set and the minimum point set respectively;

[0021] Perform linear interpolation on the detected maximum and minimum points to obtain preliminary upper and lower envelopes;

[0022] Based on the preliminary upper envelope and lower envelope, the envelope is optimized twice using adaptive regularized least squares interpolation;

[0023] Calculate the average value of the upper envelope and the lower envelope to get the mean line;

[0024] Calculate the difference between the signal value to be processed and the mean line to obtain the preliminary mode component;

[0025] Frequency distribution optimization and energy distribution optimization are used to optimize and adjust the preliminary mode components;

[0026] The optimized mode components are effective mode components of the signal to be processed.

[0027] As a preferred solution of the digital twin management method for petrochemical terminals of the present invention, wherein: the empirical mode decomposition technology and noise processing method are used to further decompose and process the pre-processed data to obtain a multi-dimensional signal, and also includes:

[0028] Calculate the total energy of the signal to be processed, the expression is,

[0029] ;

[0030] in, is the total energy, is the signal to be processed, is the frequency domain signal, is the total time length of the signal, is the frequency range, For time, is the frequency;

[0031] Calculate the first The energy of the effective mode components is expressed as,

[0032] ;

[0033] in, is extracted from the signal to be processed. The energy of the effective mode components, is extracted from the signal to be processed. The effective mode components, is the total time length of the signal, For time;

[0034] When decomposed into When there are valid mode components, the residual signal energy is calculated and the energy threshold is set. When the residual signal energy is higher than the set energy threshold, the residual signal is used as the input for the next mode component extraction and the decomposition is continued. When the ratio of the residual energy to the total energy is lower than the set energy threshold, the decomposition is stopped.

[0035] For each effective mode component, its frequency component is extracted by fast Fourier transform, and its energy in the corresponding frequency band is calculated to obtain the frequency band energy;

[0036] For each adjacent effective mode component, calculate the energy ratio of the two frequency bands;

[0037] Set the frequency band energy ratio threshold, when the frequency band energy ratio is less than the threshold, stop decomposition;

[0038] When any stopping condition is met, the decomposition process is terminated and all valid mode components are output;

[0039] For each effective mode component, the noise intensity is calculated according to its energy distribution and frequency characteristics, and a noise adjustment factor is assigned to each effective mode component according to the noise intensity to perform denoising;

[0040] The signal is reconstructed based on the effective mode components and the residual signal energy after denoising to obtain a multi-dimensional signal.

[0041] As a preferred solution of the petrochemical terminal digital twin management method of the present invention, the real-time monitoring of whether the terminal equipment is operating normally and early warning through the digital twin model includes:

[0042] Set signal thresholds for different terminal equipment signals, which are divided into primary threshold, secondary threshold, tertiary threshold and quaternary threshold;

[0043] When the device signal continues to exceed the first-level threshold, it is set to a slight abnormal state and continues to operate, but the terminal equipment is monitored and adjusted;

[0044] When the device signal continues to exceed the secondary threshold, it is set to a moderate abnormal state and preventive measures are taken to avoid further deterioration;

[0045] When the device signal exceeds the third level threshold continuously, it is set to a serious abnormal state and immediate action is required;

[0046] When the device signal exceeds the level 4 threshold continuously, it is set to a critical abnormal state, requiring emergency shutdown and equipment repair.

[0047] In a second aspect, the present invention provides a petrochemical terminal digital twin management system, comprising:

[0048] The data acquisition and preprocessing module acquires multi-dimensional data of the petrochemical terminal and preprocesses it;

[0049] The signal decomposition and noise processing module uses empirical mode decomposition technology and noise processing methods to further decompose and process the pre-processed data to obtain multi-dimensional signals;

[0050] Digital twin model building module, which builds a digital twin model based on multi-dimensional signals to identify whether the terminal equipment is operating normally;

[0051] The real-time monitoring and early warning module uses the digital twin model to monitor whether the equipment is operating normally in real time and issue early warnings.

[0052] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the digital twin management method for a petrochemical terminal as described in the first aspect of the present invention is implemented.

[0053] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the digital twin management method for a petrochemical terminal as described in the first aspect of the present invention is implemented.

[0054] The beneficial effects of the present invention are as follows: through the empirical mode decomposition technology, complex multidimensional signals can be decomposed into multiple mode components, and signal characteristics of different frequencies and energy distributions can be effectively extracted. The signal quality is further optimized through the noise processing method, thereby improving the signal resolution. At the same time, it can accurately identify minor anomalies in equipment operation, enhance the sensitivity of the digital twin model to the equipment operation status, and ensure accurate identification and real-time monitoring of the equipment status. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0056] Figure 1 This is a flow chart of the digital twin management method for the petrochemical terminal in Example 1.

[0057] Figure 2 This is a flow chart of the empirical mode decomposition stopping principle in Example 1. DETAILED DESCRIPTION

[0058] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0059] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0060] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0061] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, provides a petrochemical terminal digital twin management method, comprising the following steps:

[0062] S1. Obtain multi-dimensional data of petrochemical terminals and pre-process them;

[0063] S101. Specifically, deploy tank level sensors, pipeline pressure sensors, temperature sensors, leak detection sensors, and weather station sensors at terminal equipment and environmental points to collect tank level data, pipeline pressure data, temperature data, leak detection data, and weather data in real time;

[0064] S102, performing outlier processing and noise filtering on the collected data, performing time alignment using a linear interpolation method, and then performing standardization processing on the data;

[0065] S103, after completing the outlier processing, noise filtering and standardization of the data, integrating the data of all dimensions according to the timestamps to form multidimensional time series data;

[0066] S2, using empirical mode decomposition technology and noise processing methods to further decompose and process the pre-processed data to obtain a multi-dimensional signal;

[0067] S201, defining the standardized time series data of each sensor dimension as a signal to be processed, traversing each sampling point of the signal to be processed, and detecting its local maximum and local minimum;

[0068] S202, recording all points that meet the maximum and minimum conditions as a maximum point set and a minimum point set respectively;

[0069] S203, linear interpolation is performed on the detected maximum and minimum points respectively to obtain preliminary upper and lower envelopes, which are expressed as follows:

[0070] ;

[0071] in, is the envelope line, is the interpolation coefficient, is the spline basis function, is the number of basis functions, Number the basis functions;

[0072] Use the maximum points and the corresponding signal values ​​to be processed to construct a cubic spline interpolation equation, and solve the equation to obtain the interpolation coefficients to generate the upper envelope; use the minimum points and the corresponding signal values ​​to be processed to construct a cubic spline interpolation equation, and solve the equation to obtain the interpolation coefficients to generate the lower envelope. The upper envelope needs to be above the signal, and the lower envelope needs to be below the signal to ensure that the envelope covers the entire signal;

[0073] The envelope generated by spline interpolation is smooth to ensure that the subsequently generated intrinsic mode function has good smoothness. The generation of the envelope ensures that it passes through all local maximum and minimum points.

[0074] S204, based on the preliminary upper envelope and lower envelope, using adaptive regularized least squares interpolation to perform secondary optimization on the envelope;

[0075] S2041, calculating the average values ​​of the upper envelope line and the lower envelope line to obtain a mean line;

[0076] S2042, calculating the difference between the signal value to be processed and the mean line to obtain a preliminary mode component;

[0077] S2043, optimizing and adjusting the preliminary mode components by using frequency distribution optimization and energy distribution optimization;

[0078] S205. The optimized mode components are effective mode components of the signal to be processed.

[0079] S206. Calculate the total energy of the signal to be processed, expressed as:

[0080] ;

[0081] in, is the total energy, is the signal to be processed, is the frequency domain signal, is the total time length of the signal, is the frequency range;

[0082] S2061, calculate the first The energy of the effective mode components is expressed as,

[0083] ;

[0084] in, is extracted from the signal to be processed. The energy of the effective mode components, is extracted from the signal to be processed. The effective mode components, is the total time length of the signal;

[0085] S2062, when decomposed to When there are effective mode components, calculate the remaining signal energy,

[0086] ;

[0087] in, is the residual signal energy, is the total energy, For the The energy of a component, is the index of the component, is the total number of components obtained after signal decomposition;

[0088] Set an energy threshold. When the energy of the residual signal is higher than the set energy threshold, use the residual signal as the input for the next step of pattern component extraction and continue to decompose. When the ratio of the residual energy to the total energy is lower than the set energy threshold, stop decomposition.

[0089] S207, for each effective mode component, extract its frequency component by fast Fourier transform, and calculate its energy in the corresponding frequency band to obtain the frequency band energy;

[0090] S2071. For each adjacent effective mode component, calculate the frequency band energy ratio between the two components;

[0091] S2072, setting a frequency band energy ratio threshold, and stopping decomposition when the frequency band energy ratio is less than the threshold;

[0092] S208, when any stop condition is met, the decomposition process is terminated and all valid mode components are output;

[0093] S209, for each effective mode component, calculate its noise intensity according to its energy distribution and frequency characteristics, and assign a noise adjustment factor to each effective mode component according to the noise intensity to perform denoising;

[0094] The noise intensity calculation expression is:

[0095] ;

[0096] in, is the noise intensity, is the noise frequency band, is the spectrum of the effective mode component, is the frequency,

[0097] The adaptive determination expression of the adjustment factor is:

[0098] ;

[0099] in, is the noise adjustment factor, is the noise intensity,

[0100] S210 , reconstructing the signal based on the effective mode components after denoising and the residual signal energy to obtain a multi-dimensional signal.

[0101] By using two stopping conditions at the same time, the empirical mode decomposition can be more effectively controlled, thereby avoiding over-decomposition or under-decomposition.

[0102] By combining empirical mode decomposition technology and noise processing methods, the decomposition of effective mode components is more efficient and accurate, and useful information in the signal can be more effectively extracted, noise can be removed, and high-quality signal input can be provided for subsequent digital twin management.

[0103] S3. Building a digital twin model based on multi-dimensional signals to identify whether the terminal equipment is operating normally, including the following steps;

[0104] S301, using physical equations to establish a physical model and create a geometric structure of the terminal equipment;

[0105] The physical model includes an input layer, an equation layer, and a boundary condition layer; S3011, the input layer inputs the tank level data, pipeline pressure data, temperature data, leakage detection data, and meteorological data collected by the tank level sensor, pipeline pressure sensor, temperature sensor, leakage detection sensor, and meteorological station sensor;

[0106] The equation layer uses PINNs (Physics-Informed Neural Networks), which enables the network to follow known physical laws while learning data distribution. Specifically, it includes the embedding of physical equations, neural network approximation solutions, and automatic differentiation.

[0107] The embedded equations include Newton's second law, heat conduction equation, continuity equation, Bernoulli equation, energy conservation equation and heat conduction equation;

[0108] Neural networks are designed to approximate the solutions to physical equations. By minimizing the loss function, neural networks can learn the solutions to physical systems and make predictions.

[0109] PINNs can efficiently calculate the derivative of the output of the neural network with respect to the input by using automatic differentiation technology, which can be used to solve partial derivatives in physical equations.

[0110] The boundary condition layer includes the spatial constraints of the equipment, external pressure, and the interaction between the equipment and the environment. The specific constraints are set according to the needs and equipment types. The boundary conditions are mainly used to ensure that the physical model conforms to the operating environment of the actual equipment. S3012. The output of the physical model is the displacement, velocity, acceleration and other physical states of the equipment. This output represents the ideal behavior of the equipment under normal operating conditions.

[0111] S302: For the parts that are difficult to accurately describe with the physical model, a data-driven model is introduced to compensate and enhance them, and the physical model is combined with the data-driven model to form a unified hybrid model;

[0112] The data-driven model is built based on the support vector machine, including the input layer, model structure layer, training and optimization layer, and output layer.

[0113] The input layer is used to input multidimensional data.

[0114] The model structure layer uses support vector machine for deep learning.

[0115] The training and optimization layer uses historical data to train the model, adjusts the model parameters, and uses cross-validation and hyperparameter tuning during the optimization process.

[0116] The output layer is a quantitative assessment of the current health status of the device, judging the normal and abnormal status of the device;

[0117] S303, mapping the multidimensional signal to the input parameters of the hybrid model to construct a digital twin model;

[0118] The digital twin model includes data layer, model layer and service layer;

[0119] The data layer is responsible for real-time data collection, transmission, storage and preprocessing to ensure high quality and real-time performance of the data;

[0120] The model layer, which includes physical models and data-driven models, is responsible for simulating and analyzing the status of the equipment;

[0121] The service layer provides model management, computing resource management, data management, and security management services to ensure efficient, reliable, and secure operation of the model.

[0122] S3031. Acquire multi-dimensional signals in real time and map them to the input layer of the digital twin model;

[0123] S3032. Through the input of real-time multi-dimensional signals, the digital twin model can automatically update its internal state to ensure that the equipment operation state is consistent with the actual terminal equipment. The parameters in the physical model and data-driven model are adjusted according to the real-time multi-dimensional signals to maintain high-precision simulation of the terminal equipment.

[0124] S304: Monitor the tank level data, pipeline pressure data, temperature data, leakage detection data and meteorological data of the actual terminal in real time, and map the monitored terminal equipment data information to the digital twin model to ensure that the digital twin model is consistent with the status of the actual terminal equipment. The equipment status changes in the digital twin model are also fed back to the actual terminal.

[0125] S3041. The output of the digital twin model includes the displacement, velocity, acceleration, temperature, pressure and other states of the equipment. The normal state of the equipment is obtained through historical data or the ideal output of the physical model. The error between the current state and the normal state is used to determine whether the equipment is operating normally.

[0126] S4. Use the digital twin model to monitor whether the equipment is operating normally in real time and issue early warnings, including the following steps:

[0127] S401, setting thresholds for different terminal equipment signals, which are divided into primary threshold, secondary threshold, tertiary threshold and quaternary threshold;

[0128] S401, when the device signal continues to exceed the first-level threshold, it is set to a slightly abnormal state and continues to operate, but the terminal equipment is monitored and adjusted;

[0129] S402: When the device signal continues to exceed the secondary threshold, it is set to a moderate abnormal state and preventive measures are taken to avoid further deterioration;

[0130] S403: When the device signal exceeds the third level threshold continuously, it is set to a serious abnormal state and immediate action is required;

[0131] S404: When the device signal exceeds the level 4 threshold continuously, it is set to a critical abnormal state, requiring emergency shutdown and equipment repair.

[0132] This embodiment also provides a petrochemical terminal digital twin management system, including:

[0133] The data acquisition and preprocessing module acquires multi-dimensional data of the petrochemical terminal and preprocesses it;

[0134] The signal decomposition and noise processing module uses empirical mode decomposition technology and noise processing methods to further decompose and process the pre-processed data to obtain multi-dimensional signals;

[0135] Digital twin model building module, which builds a digital twin model based on multi-dimensional signals to identify whether the terminal equipment is operating normally;

[0136] The real-time monitoring and early warning module uses the digital twin model to monitor whether the equipment is operating normally in real time and issue early warnings.

[0137] In summary, the present invention can decompose complex multidimensional signals into multiple mode components through empirical mode decomposition technology, effectively extract signal characteristics of different frequencies and energy distributions, and further optimize the signal quality through noise processing methods, thereby improving the resolution of the signal. At the same time, it can accurately identify minor anomalies in equipment operation, enhance the sensitivity of the digital twin model to the equipment operation status, and ensure accurate identification and real-time monitoring of the equipment status.

[0138] This embodiment also provides a computer device, which is applicable to the digital twin management method and system of petrochemical terminals, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the digital twin management method of petrochemical terminals proposed in the above embodiment.

[0139] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0140] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the digital twin management method and system for a petrochemical terminal proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage device, a flash memory, a disk or an optical disk.

[0141] Example 2, referring to Table 1, is the second example of the present invention. In order to further verify the technical solution of the present invention, experimental simulation data of the digital twin management method of the petrochemical terminal are provided.

[0142] In this embodiment, a digital twin management method for a petrochemical terminal was tested in a petrochemical terminal scenario. The test objects included three key equipment: storage tanks, oil pumps, and pipelines. In order to verify the superiority of the present invention over traditional equipment monitoring methods, each device was installed with a variety of sensors to collect multi-dimensional data, including vibration, temperature, and pressure. During the test, the data collection frequency was 100 times per second, and the operating status of the equipment under different loads was continuously monitored.

[0143] Multi-dimensional data of the equipment is obtained through sensors, including vibration frequency, temperature, and pressure. The preprocessing steps include outlier processing, noise filtering, and data standardization of the collected data. The outlier processing method is used to eliminate abnormal data caused by sensor failure or external factors to ensure data accuracy. In terms of noise filtering, low-pass filters and time domain smoothing technology are combined to remove interference from environmental noise and obtain pure data.

[0144] The preprocessed multidimensional signal is subjected to empirical mode decomposition to extract effective mode components. A preliminary envelope is constructed by detecting maximum and minimum points, and the envelope is optimized secondary using adaptive regularized least squares interpolation to avoid the envelope overfitting problem caused by insufficient sensor sampling. The preliminary mode components are optimized and adjusted according to the frequency and energy distribution of each mode component. Frequency distribution optimization and energy distribution optimization are used to eliminate interference signals in irrelevant frequency bands to ensure that the extracted mode components are highly representative.

[0145] Based on multi-dimensional signals, combined with the physical equations of the equipment and data-driven models, a digital twin model of the petrochemical terminal was constructed. The physical equations are used to describe the geometric structure and basic operating characteristics of the equipment, while the data-driven model makes up for the shortcomings of the physical model in complex nonlinear scenarios. By mapping multi-dimensional signals to the input parameters of the model, the digital twin model can reflect in real time whether the equipment is operating normally.

[0146] Set the threshold levels of multi-dimensional signals, including minor abnormalities, moderate abnormalities, severe abnormalities, and critical abnormalities. When the equipment operation status exceeds a certain threshold, the system automatically triggers an early warning. For minor abnormalities, the system will continue to monitor the equipment status and make automatic adjustments; for severe and critical abnormalities, the system will issue a forced shutdown signal and record the equipment's operation data for subsequent analysis.

[0147] The details are shown in Table 1 below:

[0148] Table 1 Experimental data table

[0149]

[0150] As can be seen from the table, the vibration frequency, temperature and pressure changes of storage tanks, oil pumps and pipelines under different operating conditions. Traditional equipment monitoring methods can often only monitor single-dimensional data, such as temperature or pressure, and ignore the multi-dimensional characteristics of the equipment. Existing monitoring methods lack effective signal decomposition capabilities when processing complex multi-dimensional signals, and are easily disturbed by noise and changes in the external environment. They cannot accurately identify whether the terminal equipment is operating normally and potential failures.

[0151] In contrast, the present invention adopts the empirical mode decomposition technology, which can decompose the multi-dimensional signal of equipment operation into multiple mode components, and effectively extract the signal characteristics of different frequencies and energy distributions. During the operation of the storage tank, the vibration frequency increases from 32Hz to 54Hz, the temperature rises from 82℃ to 98℃, and the pressure also rises from 234kPa to 276kPa. Through the method of the present invention, the system can detect slight abnormalities and trigger a first-level warning, indicating that the equipment is in an early fault state. In traditional monitoring methods, these changes will not attract enough attention, thereby missing the opportunity to prevent faults.

[0152] In the monitoring of the oil pump, the vibration frequency reached 67Hz, the temperature rose to 109°C, and the pressure reached 432kPa. At this time, the digital twin model of the present invention promptly detected that the equipment was in a critical abnormal state and triggered a level 4 warning, requiring immediate shutdown for emergency repairs. In traditional technologies, it is difficult to achieve such efficient real-time warnings, which may cause the equipment to continue to run until a failure occurs, resulting in more serious consequences.

[0153] In addition, the present invention avoids the overfitting of the envelope and the underfitting of the signal decomposition by optimizing the frequency distribution and energy distribution of the multidimensional signal. The traditional signal processing method cannot effectively decompose complex signals, which easily leads to misjudgment or missed judgment. The present invention enhances the smoothness and accuracy of the envelope through adaptive regularized least squares interpolation technology, thereby ensuring the effectiveness of the mode components.

[0154] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A petrochemical terminal digital twin management method, characterized by: include, Obtain multi-dimensional data of petrochemical terminals and pre-process them; The empirical mode decomposition technology and noise processing method are used to further decompose and process the pre-processed data to obtain multi-dimensional signals; Build a digital twin model based on multi-dimensional signals to identify whether terminal equipment is operating normally; Use digital twin models to monitor the normal operation of terminal equipment in real time and issue early warnings; The digital twin model is constructed based on multi-dimensional signals to identify whether the terminal equipment is operating normally, including: Use physical equations to build basic models and create terminal equipment geometry; For the parts that are difficult to accurately describe with the physical model, a data-driven model is introduced to compensate and enhance them, combining the physical model with the data-driven model to form a unified hybrid model; Map multidimensional signals to the input parameters of the hybrid model to build a digital twin model; Monitor the data of real-world terminal equipment in real time, and map the monitored real-world terminal equipment data information to the digital twin model to ensure that the digital twin model is consistent with the status of the real-world terminal equipment. The equipment status changes in the digital twin model are also fed back to the real-world terminal equipment. The preprocessing includes outlier processing, noise filtering and standardization; The empirical mode decomposition technology and noise processing method are used to further decompose and process the pre-processed data, including: Define the standardized time series data of each sensor dimension as the signal to be processed, traverse each sampling point of the signal to be processed, and detect its local maximum and local minimum; All points that meet the maximum and minimum conditions are recorded as the maximum point set and the minimum point set respectively; Perform linear interpolation on the detected maximum and minimum points to obtain preliminary upper and lower envelopes; Based on the preliminary upper envelope and lower envelope, the envelope is optimized twice using adaptive regularized least squares interpolation; Calculate the average value of the upper envelope and the lower envelope to get the mean line; Calculate the difference between the signal value to be processed and the mean line to obtain the preliminary mode component; The frequency distribution optimization and energy distribution optimization are used to optimize and adjust the preliminary mode components; The optimized mode components are effective mode components of the signal to be processed.

2. The petrochemical terminal digital twin management method according to claim 1, characterized in that: The method further decomposes and processes the pre-processed data using the empirical mode decomposition technology and the noise processing method to obtain a multi-dimensional signal, and also includes: Calculate the total energy of the signal to be processed, the expression is, ； in, is the total energy, is the signal to be processed, is the frequency domain signal, is the total time length of the signal, is the frequency range, For time, is the frequency; Calculate the first The energy of the effective mode components is expressed as, ； in, is extracted from the signal to be processed. The energy of the effective mode components, is extracted from the signal to be processed. The effective mode components, is the total time length of the signal, For time; When decomposed into When there are valid mode components, the residual signal energy is calculated and the energy threshold is set. When the residual signal energy is higher than the set energy threshold, the residual signal is used as the input for the next mode component extraction and the decomposition is continued. When the ratio of the residual energy to the total energy is lower than the set energy threshold, the decomposition is stopped. For each effective mode component, its frequency component is extracted by fast Fourier transform, and its energy in the corresponding frequency band is calculated to obtain the frequency band energy; For each adjacent effective mode component, calculate the energy ratio of the two frequency bands; Set the frequency band energy ratio threshold, when the frequency band energy ratio is less than the threshold, stop decomposition; When any stopping condition is met, the decomposition process is terminated and all valid mode components are output; For each effective mode component, the noise intensity is calculated according to its energy distribution and frequency characteristics, and a noise adjustment factor is assigned to each effective mode component according to the noise intensity to perform denoising; The signal is reconstructed based on the effective mode components and the residual signal energy after denoising to obtain a multi-dimensional signal.

3. The petrochemical terminal digital twin management method according to claim 2, characterized in that: The digital twin model is used to monitor the normal operation of terminal equipment in real time and issue early warnings. include, Set signal thresholds for different terminal equipment signals, which are divided into primary threshold, secondary threshold, tertiary threshold and quaternary threshold; When the device signal continues to exceed the first-level threshold, it is set to a slight abnormal state and continues to operate, but the terminal equipment is monitored and adjusted; When the device signal continues to exceed the secondary threshold, it is set to a moderate abnormal state and preventive measures are taken to avoid further deterioration; When the device signal exceeds the third level threshold continuously, it is set to a serious abnormal state and immediate action is required; When the device signal exceeds the level 4 threshold continuously, it is set to a critical abnormal state, requiring emergency shutdown and equipment repair.

4. A petrochemical terminal digital twin management system, based on the petrochemical terminal digital twin management method according to any one of claims 1 to 3, characterized in that: include, The data acquisition and preprocessing module acquires multi-dimensional data of the petrochemical terminal and preprocesses it; The signal decomposition and noise processing module uses empirical mode decomposition technology and noise processing methods to further decompose and process the pre-processed data to obtain multi-dimensional signals; Digital twin model building module, which builds a digital twin model based on multi-dimensional signals to identify whether the terminal equipment is operating normally; The real-time monitoring and early warning module uses the digital twin model to monitor whether the equipment is operating normally in real time and issue early warnings.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the petrochemical terminal digital twin management method described in any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the petrochemical terminal digital twin management method described in any one of claims 1 to 3 are implemented.

Images