A data acquisition and processing method and system based on digital twin platform

Abstract

The present invention relates to the technical field of data acquisition and processing, and specifically to a data acquisition and processing method and system based on a digital twin platform, comprising the following steps: S1: the digital twin platform determines the location of the sensor based on the coordinate system and creates a specific model for marking; S2: real-time acquisition of multi-source data in the actual environment; S3: preliminary processing of the collected data through edge computing equipment; S4: data fusion and enhancement using a preset multimodal data fusion algorithm; S5: dynamic simulation and prediction of the actual environment; S6: generation of optimization control instructions; S7: regular model calibration and parameter adjustment, and continuous optimization of the accuracy of the digital twin model. By integrating edge computing and multimodal data fusion technology, the efficiency and accuracy of data processing are achieved, while physical simulation and machine learning are combined to improve prediction accuracy, significantly improving the response speed and decision-making quality of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of data acquisition and processing, and in particular to a data acquisition and processing method and system based on a digital twin platform. Background Art

[0002] In today's pharmaceutical industry, ensuring drug quality, improving production efficiency and meeting strict regulatory requirements are the core tasks facing pharmaceutical companies. The pharmaceutical environment of the production workshop has a vital impact on pharmaceutical production. In the existing technology, HVAC systems are usually used to detect temperature, humidity, pressure difference and air cleanliness in the pharmaceutical environment; however, HVAC systems rely on centralized data processing centers when detecting multiple environmental parameters, which often show problems of high processing delay and poor scalability when processing large-scale, multi-source data, and lack of intelligent control and real-time adjustment, which exacerbates energy waste. Summary of the invention

[0003] Based on the above objectives, the present invention provides a data acquisition and processing method and system based on a digital twin platform.

[0004] A data acquisition and processing method based on a digital twin platform comprises the following steps:

[0005] S1: Obtain the location information of the sensor in the actual environment. The digital twin platform determines the location of the sensor based on the coordinate system and creates a specific model at the location of the sensor for marking.

[0006] S2: Deploy corresponding sensors and data acquisition equipment at the location of the sensor, including temperature sensors, humidity sensors, differential pressure gauges, position sensors, and video monitoring equipment, to collect multi-source data in the actual environment in real time;

[0007] S3: Perform preliminary processing on the collected data through edge computing devices, including data filtering, outlier detection, and preliminary aggregation processing;

[0008] S4: The preliminarily processed data is transmitted to the digital twin platform, and the data is fused and enhanced using the preset multimodal data fusion algorithm;

[0009] S5: Based on the fused data, the preset physical simulation algorithm and machine learning model are used to dynamically simulate and predict the actual environment on the digital twin platform;

[0010] S6: Generate optimized control instructions based on simulation and prediction results, and adjust equipment and system parameters in the actual environment in real time through feedback mechanisms;

[0011] S7: Continuously monitor the data collected and simulation results throughout the process, regularly perform model calibration and parameter adjustments, and continuously optimize the accuracy of the digital twin model.

[0012] Furthermore, the S1 specifically includes:

[0013] S11: Use wireless sensor network positioning technology to measure sensor location information in the actual environment;

[0014] S12: inputting the measured sensor position information and the measurement origin into the computer-aided design software, and generating a preliminary virtual environment model through data processing and model building functions;

[0015] S13: Introduce physical property data in the actual environment into the virtual environment model, including material properties, temperature, humidity, and pressure difference, and enhance the realism and accuracy of the model through parameter setting and attribute assignment;

[0016] S14: Use real-time synchronization technology to obtain dynamic data in the actual environment through IoT devices in real time, and synchronize and update the data with the virtual model in real time to ensure that the virtual model can dynamically reflect changes in the actual environment;

[0017] S15: Through multiple rounds of calibration and verification steps, the virtual model is calibrated for accuracy and error correction using known calibration points and measurement data;

[0018] S16: Integrate data visualization tools on the digital twin platform to display and interact with the created virtual model in three dimensions.

[0019] Furthermore, the S2 specifically includes:

[0020] S21: Based on the monitoring requirements and layout design of the actual environment, various sensors and data acquisition equipment are reasonably distributed in the monitoring area to ensure that all predetermined locations are covered;

[0021] S22: Use high-precision temperature sensors installed in areas of the environment susceptible to temperature changes to measure ambient temperature in real time;

[0022] S23: Deploy humidity sensors in the transition area between the humid environment and the dry environment to monitor the changes in environmental humidity in real time;

[0023] S24: Install differential pressure gauges at stress-bearing structures and areas where differential pressure changes, to collect differential pressure data in the environment in real time;

[0024] S25: Use high-precision position sensors to locate dynamic equipment and personnel to ensure real-time acquisition of position data. The position sensor achieves high-precision positioning by combining GNSS modules and indoor positioning technology;

[0025] S26: Install video surveillance equipment in the monitoring area and collect environmental video data in real time through cameras;

[0026] S2: Integrate data acquisition devices of various sensors to build a unified data acquisition network to ensure that the data of each sensor can be transmitted to the edge computing device in real time for preliminary processing and storage.

[0027] Furthermore, the S3 specifically includes:

[0028] S31: Deploy a data filtering module on the edge computing device to pre-process the raw data collected by the sensor, remove noise and invalid data, and ensure data clarity;

[0029] S32: Using a preset outlier detection algorithm, perform outlier detection on the filtered data on the edge computing device to identify abnormal data points that are significantly different from normal data distribution;

[0030] S33: Mark and record the detected abnormal data on the edge computing device, and store the abnormal data separately from the normal data to facilitate subsequent analysis and processing;

[0031] S34: performing preliminary aggregation processing on the filtered and outlier-detected data, grouping and counting the data based on the time window and data type, and generating preliminary aggregated data;

[0032] S35: Perform preliminary analysis on the aggregated data on the edge computing device and calculate statistical indicators, including the average, maximum, minimum, and standard deviation;

[0033] S36: Transmit the preliminarily processed data to the digital twin platform through edge computing devices to ensure the timeliness and continuity of the data.

[0034] Furthermore, the S4 specifically includes:

[0035] S41: First, the preliminarily processed data is transmitted to the digital twin platform through a secure network protocol;

[0036] S42: Deploy a multimodal data fusion module on the digital twin platform to receive data from different sensors and perform time alignment and spatial alignment processing on them to ensure the consistency of the data in time and space;

[0037] S43: Apply weighted average fusion algorithm to perform preliminary fusion of multi-source data and set weights for different sensor data , calculate the weighted average , the formula is: ,in, For the The data of the sensors, is the corresponding weight, is the number of sensors;

[0038] S44: Use the principal component analysis algorithm to perform dimensionality reduction processing on the initially fused data to extract features. The specific calculation formula is: ,in, is the input data matrix, is the eigenvector matrix, is the data after dimension reduction;

[0039] S45: Apply data enhancement technology based on generative adversarial networks to enhance the reduced-dimensional data to generate representative data samples. The formula for generating adversarial networks is:

[0040] ,in, For the generator, is the discriminator, is the real data distribution, is the noise distribution;

[0041] S46: Perform consistency check on the enhanced data to ensure consistency and accuracy during data fusion and enhancement. The specific use is the correlation coefficient To test, the formula is: ,in, and are the sample values ​​of the two data sets respectively, and are the means of the two data sets respectively;

[0042] S47: Finally, the fused and enhanced data is stored in the database of the digital twin platform for subsequent simulation, prediction and analysis. At the same time, a data structure based on the sensor location information is created for the device. The data in the data structure is the data obtained by the sensor near the device and the corresponding standard value range.

[0043] Furthermore, the S5 specifically includes:

[0044] S51: Deploy the physical simulation module on the digital twin platform, receive the fused data, build the physical simulation model based on the physical characteristics and operating conditions of the actual environment, and set the initial conditions and boundary conditions;

[0045] S52: Apply the finite element analysis method to simulate the physical simulation model and calculate the state of environmental changes within a predetermined time step. The calculation formula of the finite element analysis is: ,in, is the stiffness matrix, is the displacement vector, is the force vector;

[0046] S53: storing and analyzing the results of each time step during the simulation process, and adjusting the parameters of the simulation model to improve the simulation accuracy by comparing the simulation results with the actual data;

[0047] S54: Deploy the machine learning model on the digital twin platform, select the long short-term memory network algorithm, and train and predict the fused time series data;

[0048] S55: Use the trained machine learning model to predict the future environmental state and generate prediction results, including the future trend of temperature, humidity, and pressure difference parameters;

[0049] S56: Compare and integrate the results of the physical simulation and machine learning models, combine the accuracy of the physical simulation model and the prediction ability of the machine learning model, and generate the final dynamic simulation and prediction results;

[0050] S57: Visualize the final simulation and prediction results, and display the dynamic changes of the environment through the visualization tools of the digital twin platform.

[0051] Furthermore, the S56 specifically includes:

[0052] S561: Get simulation results within the same time step from the physical simulation module and the machine learning model and prediction results ;

[0053] S562: Calculate the weights of the physical simulation results and the machine learning prediction results, adjust based on the accuracy of historical data and the reliability of the model, and set the weights of the physical simulation model and machine learning model weights , the weight must be satisfied ;

[0054] S563: Apply the weighted average fusion algorithm to fuse the physical simulation results and the machine learning prediction results to calculate the final dynamic simulation and prediction results , the formula is:

[0055] ,in, is the time step The physical simulation results are is the time step The machine learning prediction results, and are the weights of physical simulation results and machine learning results respectively;

[0056] S564: Perform error analysis on the weighted fusion result and calculate the error between the fusion result and the actual data , the error calculation formula is: ,in, is the time step actual data; is the number of data samples;

[0057] S565: According to the results of the error analysis, the weights of the physical simulation model and the machine learning model are adjusted to optimize the weight allocation strategy;

[0058] S566: Repeat the above weighted fusion and error analysis steps, iteratively optimize, and finally generate high-precision dynamic simulation and prediction results.

[0059] Furthermore, the S6 specifically includes:

[0060] S61: Obtaining dynamic simulation and prediction results , parameter changes in the analysis results, including temperature, humidity, and pressure difference;

[0061] S62: Generate an optimized control instruction using a preset control algorithm according to the parameter change. The control algorithm formula is: ,in, is the time step control instructions, is the error between the expected value and the actual value, , and are proportional, integral and differential control parameters respectively;

[0062] S63: Sending the generated optimization control instructions to the devices and systems in the actual environment through a secure communication protocol;

[0063] S64: After receiving the optimization control instruction, the equipment and system adjust corresponding operating parameters according to the instruction, including adjusting parameters of temperature, humidity, and pressure difference, so as to achieve environmental optimization;

[0064] S65: Real-time monitoring of equipment and system parameter adjustment results, and feedback of real-time data through sensors to the digital twin platform;

[0065] S66: Real-time feedback data on the digital twin platform Analyze and judge the effect of equipment and system parameter adjustment to see whether the expected optimization effect is achieved; if the expected optimization effect is not achieved, adjust the control algorithm parameters , and , and regenerate optimization control instructions , iterative optimization is carried out through the feedback mechanism until the expected effect is achieved.

[0066] Furthermore, the S7 specifically includes:

[0067] S71: Continuously collect real-time data from sensors and data acquisition equipment, monitor environmental changes and equipment operating status in real time, including temperature, humidity, pressure difference and location data, to ensure comprehensive monitoring of the environment;

[0068] S72: Regularly compare the collected real-time data with the simulation results, use data comparison tools and statistical analysis methods to evaluate the accuracy and consistency of the digital twin model, and identify potential deviations and errors, including comparing the trends, fluctuation ranges and characteristics of the actual data and the simulated data;

[0069] S73: Based on the evaluation results, use historical data and real-time data to adjust the parameters of the digital twin model, specifically adjusting and updating the physical properties of the model, including material properties, fluid mechanics parameters, initial conditions and boundary conditions to reflect the latest environmental status;

[0070] S74: Deploy an automatic calibration module on the digital twin platform to calibrate the model through an iterative optimization algorithm. The algorithms used in the calibration process include particle swarm optimization and genetic algorithm. The deviation between the model and the actual environment is gradually reduced through multiple iterations.

[0071] S75: Verify the calibrated model by comparing the simulation results and actual data before and after calibration to ensure the effectiveness and accuracy of the calibration process. The verification process includes using an independent test data set to simulate and observe the match between the model response and the actual data.

[0072] S76: Based on the verification results, optimize the calibration and parameter adjustment strategies to ensure the high accuracy and high reliability of the model in a dynamic environment. The optimization strategies include adjusting the parameters of the optimization algorithm and increasing the frequency of data collection.

[0073] A data acquisition and processing system based on a digital twin platform, used to implement the above-mentioned data acquisition and processing method based on a digital twin platform, includes the following modules:

[0074] Data acquisition module: It is equipped with various types of sensors and data acquisition devices, including temperature sensors, humidity sensors, differential pressure gauges, particle counters, position sensors, RFID sensors, and video surveillance equipment to collect multi-source data in the actual environment in real time and transmit it to the edge computing module through wireless or wired networks;

[0075] Edge computing module: receives data from the data acquisition module and performs preliminary processing, including data filtering, outlier detection, and aggregation processing;

[0076] Data fusion module: deployed on the digital twin platform, receives processed data from the edge computing module, and uses a multimodal data fusion algorithm to fuse and enhance the data to ensure data consistency and integrity;

[0077] Simulation and prediction module: Using the fused data, the preset physical simulation algorithm and machine learning model are run on the digital twin platform to dynamically simulate and predict the actual environment, and generate detailed environmental status reports and future trend forecasts;

[0078] Optimization control module: According to the results generated by the simulation prediction module, the optimization control instructions are formulated, and the equipment and system parameters in the actual environment are adjusted in real time through the feedback mechanism to achieve optimization and control of the environment;

[0079] Monitoring and calibration module: Continuously monitor the operating status and data accuracy of the entire system, regularly calibrate the model and adjust parameters to ensure the accuracy of the digital twin model and the long-term stable operation of the system.

[0080] The beneficial effects of the present invention are as follows: by integrating multimodal data fusion and edge computing technology, the real-time and accuracy of data processing are significantly improved. The edge computing module is used for preliminary processing and filtering of data, which reduces the interference of invalid data and noise. At the same time, the amount of data transmitted to the central processing module is reduced, and the processing speed and efficiency of the entire system are improved. The multimodal data fusion algorithm further enhances the consistency and integrity of different data sources, ensuring the high reliability of data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0082] Figure 1 A schematic diagram of a data acquisition and processing method according to an embodiment of the present invention;

[0083] Figure 2 Schematic diagram of a data acquisition and processing system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0084] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0085] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0086] like Figure 1 As shown, a data acquisition and processing method based on a digital twin platform includes the following steps:

[0087] S1: Obtain the location information of the sensor in the actual environment. The digital twin platform determines the location of the sensor based on the coordinate system and creates a specific model at the location of the sensor for marking.

[0088] S2: Deploy corresponding sensors and data acquisition equipment at the location of the sensor, including temperature sensors, humidity sensors, differential pressure gauges, position sensors, and video monitoring equipment, to collect multi-source data in the actual environment in real time;

[0089] S3: Perform preliminary processing on the collected data through edge computing devices, including data filtering, outlier detection, and preliminary aggregation processing;

[0090] S4: The preliminarily processed data is transmitted to the digital twin platform, and the data is fused and enhanced using the preset multimodal data fusion algorithm;

[0091] S5: Based on the fused data, the preset physical simulation algorithm and machine learning model are used to dynamically simulate and predict the actual environment on the digital twin platform;

[0092] S6: Generate optimized control instructions based on simulation and prediction results, and adjust equipment and system parameters in the actual environment in real time through feedback mechanisms;

[0093] S7: Continuously monitor the data collected and simulation results throughout the process, regularly perform model calibration and parameter adjustments, and continuously optimize the accuracy of the digital twin model.

[0094] Furthermore, the S1 specifically includes:

[0095] S11: Use wireless sensor network positioning technology to measure sensor location information in the actual environment;

[0096] The wireless sensor network positioning technology measures the distance or angle information between nodes through a ranging positioning algorithm, and determines the position information of the sensor according to the distance or angle information between the nodes, one of the nodes is the origin, and the other is the fixed point position of the sensor.

[0097] The wireless sensor network positioning technology includes a module for measuring arrival time, a module for measuring arrival time difference, a module for measuring arrival angle, a module for receiving signal strength and a module for calculating position. The module for measuring arrival time measures the time required for a signal to propagate from one node to another, and then calculates the distance according to the signal propagation speed (usually the speed of light or the speed of sound).

[0098] The arrival time difference measurement module measures the time difference of signals arriving at different nodes, and calculates the position of the unknown node by combining the position information of the known node with the hyperbolic positioning principle.

[0099] The arrival angle measurement module uses an antenna array or other angle measurement equipment to measure the angle of arrival of the signal, and determines the position of the unknown node in combination with the position information of the known node.

[0100] The received signal strength module estimates the distance between nodes based on the relationship between the strength of the received signal and the distance, and measures the distance or angle information between nodes through various technical means.

[0101] The position calculation module calculates the position of the unknown node according to the measured distance or angle information using a specific mathematical algorithm, which includes triangulation, triangulation and maximum likelihood estimation.

[0102] S12: inputting the measured sensor position information and the measurement origin into the computer-aided design software, and generating a preliminary virtual environment model through data processing and model building functions;

[0103] S13: Introduce physical property data in the actual environment into the virtual environment model, including material properties, temperature, humidity, pressure difference and other parameters, and enhance the realism and accuracy of the model through parameter setting and attribute assignment;

[0104] S14: Use real-time synchronization technology to obtain dynamic data in the actual environment through IoT devices in real time, and synchronize and update the data with the virtual model in real time to ensure that the virtual model can dynamically reflect changes in the actual environment;

[0105] S15: Through multiple rounds of calibration and verification steps, the virtual model is calibrated for accuracy and error correction using known calibration points and measurement data;

[0106] S16: Integrate data visualization tools on the digital twin platform to display and interact with the created virtual model in three dimensions.

[0107] Furthermore, the S2 specifically includes:

[0108] S21: Based on the monitoring requirements and layout design of the actual environment, various sensors and data acquisition equipment are reasonably distributed in the monitoring area to ensure that all predetermined locations are covered;

[0109] S22: Use high-precision temperature sensors installed in areas of the environment susceptible to temperature changes to measure ambient temperature in real time;

[0110] S23: Deploy humidity sensors in the transition area between the humid environment and the dry environment to monitor the changes in environmental humidity in real time;

[0111] S24: Install differential pressure gauges at stress-bearing structures and areas where differential pressure changes, to collect differential pressure data in the environment in real time;

[0112] S25: Use high-precision position sensors to locate dynamic equipment and personnel to ensure real-time acquisition of position data. The position sensor achieves high-precision positioning by combining GNSS modules and indoor positioning technology;

[0113] S26: Install video surveillance equipment in the monitoring area and collect environmental video data in real time through cameras;

[0114] S27: Integrate data acquisition devices of various sensors to build a unified data acquisition network to ensure that the data of each sensor can be transmitted to the edge computing device in real time for preliminary processing and storage.

[0115] Furthermore, the S3 specifically includes:

[0116] S31: Deploy a data filtering module on the edge computing device to pre-process the raw data collected by the sensor, remove noise and invalid data, and ensure data clarity;

[0117] S32: Using a preset outlier detection algorithm, perform outlier detection on the filtered data on the edge computing device to identify abnormal data points that are significantly different from normal data distribution;

[0118] S33: Mark and record the detected abnormal data on the edge computing device, and store the abnormal data separately from the normal data to facilitate subsequent analysis and processing;

[0119] S34: performing preliminary aggregation processing on the filtered and outlier-detected data, grouping and counting the data based on the time window and data type, and generating preliminary aggregated data;

[0120] S35: Perform preliminary analysis on the aggregated data on the edge computing device and calculate statistical indicators, including the average, maximum, minimum, and standard deviation;

[0121] S36: Transmit the preliminarily processed data to the digital twin platform through edge computing devices to ensure the timeliness and continuity of the data.

[0122] Furthermore, the S4 specifically includes:

[0123] S41: First, the preliminarily processed data is transmitted to the digital twin platform through a secure network protocol;

[0124] S42: Deploy a multimodal data fusion module on the digital twin platform to receive data from different sensors and perform time alignment and spatial alignment processing on them to ensure the consistency of the data in time and space;

[0125] S43: Apply weighted average fusion algorithm to perform preliminary fusion of multi-source data and set weights for different sensor data , calculate the weighted average , the formula is: ,in, For the The data of the sensors, is the corresponding weight, is the number of sensors;

[0126] S44: Use the principal component analysis algorithm to perform dimensionality reduction processing on the initially fused data to extract features. The specific calculation formula is: ,in, is the input data matrix, is the eigenvector matrix, is the data after dimension reduction;

[0127] S45: Apply data enhancement technology based on generative adversarial networks to enhance the reduced-dimensional data to generate representative data samples. The formula for generating adversarial networks is:

[0128] ,in, For the generator, is the discriminator, is the real data distribution, is the noise distribution;

[0129] S46: Perform consistency check on the enhanced data to ensure consistency and accuracy during data fusion and enhancement. The specific use is the correlation coefficient To test, the formula is: ,in, and are the sample values ​​of the two data sets respectively, and are the means of the two data sets respectively;

[0130] S47: Finally, the fused and enhanced data is stored in the database of the digital twin platform for subsequent simulation, prediction and analysis. At the same time, a data structure based on the sensor location information is created for the device. The data in the data structure is the data obtained by the sensor near the device and the corresponding standard value range.

[0131] Furthermore, the S5 specifically includes:

[0132] S51: Deploy the physical simulation module on the digital twin platform, receive the fused data, build the physical simulation model based on the physical characteristics and operating conditions of the actual environment, and set the initial conditions and boundary conditions;

[0133] S52: Apply the finite element analysis method to simulate the physical simulation model and calculate the state of environmental changes within a predetermined time step. The calculation formula of the finite element analysis is: ,in, is the stiffness matrix, is the displacement vector, is the force vector;

[0134] S53: storing and analyzing the results of each time step during the simulation process, and adjusting the parameters of the simulation model to improve the simulation accuracy by comparing the simulation results with the actual data;

[0135] S54: Deploy the machine learning model on the digital twin platform, select the long short-term memory network algorithm, and train and predict the fused time series data;

[0136] S55: Use the trained machine learning model to predict the future environmental state and generate prediction results, including the future trend of temperature, humidity, and pressure difference parameters;

[0137] S56: Compare and integrate the results of the physical simulation and machine learning models, combine the accuracy of the physical simulation model and the prediction ability of the machine learning model, and generate the final dynamic simulation and prediction results;

[0138] S57: Visualize the final simulation and prediction results, and display the dynamic changes of the environment through the visualization tools of the digital twin platform.

[0139] Furthermore, the S56 specifically includes:

[0140] S561: Get simulation results within the same time step from the physical simulation module and the machine learning model and prediction results ;

[0141] S562: Calculate the weights of the physical simulation results and the machine learning prediction results, adjust based on the accuracy of historical data and the reliability of the model, and set the weights of the physical simulation model and machine learning model weights , the weight must be satisfied ;

[0142] S563: Apply the weighted average fusion algorithm to fuse the physical simulation results and the machine learning prediction results to calculate the final dynamic simulation and prediction results , the formula is:

[0143] ,in, is the time step The physical simulation results are is the time step The machine learning prediction results, and are the weights of physical simulation results and machine learning results respectively;

[0144] S564: Perform error analysis on the weighted fusion result and calculate the error between the fusion result and the actual data , the error calculation formula is: ,in, is the time step actual data; is the number of data samples;

[0145] S565: According to the results of the error analysis, the weights of the physical simulation model and the machine learning model are adjusted to optimize the weight allocation strategy;

[0146] S566: Repeat the above weighted fusion and error analysis steps, iteratively optimize, and finally generate high-precision dynamic simulation and prediction results.

[0147] Furthermore, the S6 specifically includes:

[0148] S61: Obtaining dynamic simulation and prediction results , parameter changes in the analysis results, including temperature, humidity, and pressure difference;

[0149] S62: Generate an optimized control instruction using a preset control algorithm according to the parameter change. The control algorithm formula is: ,in, is the time step control instructions, is the error between the expected value and the actual value, , and are proportional, integral and differential control parameters respectively;

[0150] S63: Sending the generated optimization control instructions to the devices and systems in the actual environment through a secure communication protocol;

[0151] S64: After receiving the optimization control instruction, the equipment and system adjust corresponding operating parameters according to the instruction, including adjusting parameters of temperature, humidity, and pressure difference, so as to achieve environmental optimization;

[0152] S65: Real-time monitoring of equipment and system parameter adjustment results, and feedback of real-time data through sensors to the digital twin platform;

[0153] S66: Real-time feedback data on the digital twin platform Analyze and judge the effect of equipment and system parameter adjustment to see whether the expected optimization effect is achieved; if the expected optimization effect is not achieved, adjust the control algorithm parameters , and , and regenerate optimization control instructions , iterative optimization is carried out through the feedback mechanism until the expected effect is achieved.

[0154] Furthermore, the S7 specifically includes:

[0155] S71: Continuously collect real-time data from sensors and data acquisition equipment, monitor environmental changes and equipment operating status in real time, including temperature, humidity, pressure difference and location data, to ensure comprehensive monitoring of the environment;

[0156] S72: Regularly compare the collected real-time data with the simulation results, use data comparison tools and statistical analysis methods to evaluate the accuracy and consistency of the digital twin model, and identify potential deviations and errors, including comparing the trends, fluctuation ranges and characteristics of the actual data and the simulated data;

[0157] S73: Based on the evaluation results, use historical data and real-time data to adjust the parameters of the digital twin model, specifically adjusting and updating the physical properties of the model, including material properties, fluid mechanics parameters, initial conditions and boundary conditions to reflect the latest environmental status;

[0158] S74: Deploy an automatic calibration module on the digital twin platform to calibrate the model through an iterative optimization algorithm. The algorithms used in the calibration process include particle swarm optimization and genetic algorithm. The deviation between the model and the actual environment is gradually reduced through multiple iterations.

[0159] S75: Verify the calibrated model by comparing the simulation results and actual data before and after calibration to ensure the effectiveness and accuracy of the calibration process. The verification process includes using an independent test data set to simulate and observe the match between the model response and the actual data.

[0160] S76: Based on the verification results, optimize the calibration and parameter adjustment strategies to ensure the high accuracy and high reliability of the model in a dynamic environment. The optimization strategies include adjusting the parameters of the optimization algorithm and increasing the frequency of data acquisition.

[0161] Furthermore, the device is provided with an automatic data detection algorithm for the data structure, which compares various fused sensor data with the standard value range. When the value of a certain sensor exceeds the standard data range, the digital twin platform will mark the abnormal sensor data with a red signal according to the location information, and apply the latest adjustment strategy to the entire sensor data to uniformly adjust its value.

[0162] like Figure 2 As shown, a data acquisition and processing system based on a digital twin platform is used to implement the above-mentioned data acquisition and processing method based on a digital twin platform, and includes the following modules:

[0163] Data acquisition module: It is equipped with various types of sensors and data acquisition devices, including temperature sensors, humidity sensors, differential pressure gauges, position sensors, and video monitoring devices, to collect multi-source data in the actual environment in real time and transmit it to the edge computing module through wireless or wired networks;

[0164] Edge computing module: receives data from the data acquisition module and performs preliminary processing, including data filtering, outlier detection, and aggregation processing;

[0165] Data fusion module: deployed on the digital twin platform, receives processed data from the edge computing module, and uses a multimodal data fusion algorithm to fuse and enhance the data to ensure data consistency and integrity;

[0166] Simulation and prediction module: Using the fused data, the preset physical simulation algorithm and machine learning model are run on the digital twin platform to dynamically simulate and predict the actual environment, and generate detailed environmental status reports and future trend forecasts;

[0167] Optimization control module: According to the results generated by the simulation prediction module, the optimization control instructions are formulated, and the equipment and system parameters in the actual environment are adjusted in real time through the feedback mechanism to achieve optimization and control of the environment;

[0168] Monitoring and calibration module: Continuously monitor the operating status and data accuracy of the entire system, regularly calibrate the model and adjust parameters to ensure the accuracy of the digital twin model and the long-term stable operation of the system.

[0169] The present invention, by combining physical simulation and machine learning technology, can perform accurate dynamic simulation and trend prediction of the actual environment. This combination method not only utilizes the detailed environmental mapping capability of the physical model, but also incorporates the self-learning and adaptability of the machine learning model, making the prediction results more accurate, effectively guiding actual operations and decision-making, and greatly improving the response speed and decision-making quality.

[0170] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity.

[0171] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A data acquisition and processing method based on a digital twin platform, characterized in that: The following steps are involved: S1: Obtain the location information of the sensor in the actual environment. The digital twin platform determines the location of the sensor based on the coordinate system and creates a model at the location of the sensor for marking; S2: Deploy corresponding sensors and data acquisition equipment at the location of the sensor, including temperature sensors, humidity sensors, differential pressure gauges, position sensors, and video monitoring equipment, to collect multi-source data in the actual environment in real time; S3: Perform preliminary processing on the collected data through edge computing devices, including data filtering, outlier detection, and preliminary aggregation processing; S4: The preliminarily processed data is transmitted to the digital twin platform, and the data is fused and enhanced using the preset multimodal data fusion algorithm; S5: Based on the fused data, the preset physical simulation algorithm and machine learning model are used to dynamically simulate and predict the actual environment on the digital twin platform; S6: Generate optimized control instructions based on simulation and prediction results, and adjust equipment and system parameters in the actual environment in real time through feedback mechanisms; S7: Continuously monitor the data collected and simulation results throughout the process, regularly perform model calibration and parameter adjustments, and continuously optimize the accuracy of the digital twin model; The S1 specifically includes: S11: Use wireless sensor network positioning technology to measure sensor location information in the actual environment; S12: inputting the measured sensor position information and the measurement origin into the computer-aided design software, and generating a preliminary virtual environment model through data processing and model building functions; S13: Introduce physical property data in the actual environment into the virtual environment model, including material properties, temperature, humidity, and pressure difference, and enhance the realism and accuracy of the model through parameter setting and attribute assignment; S14: Use real-time synchronization technology to obtain dynamic data in the actual environment through IoT devices in real time, and synchronize and update the data with the virtual model in real time to ensure that the virtual model can dynamically reflect changes in the actual environment; S15: Through multiple rounds of calibration and verification steps, the virtual model is calibrated for accuracy and error correction using known calibration points and measurement data; S16: Integrate data visualization tools on the digital twin platform to display and interact with the created virtual model in three dimensions; The S5 specifically includes: S51: Deploy the physical simulation module on the digital twin platform, receive the fused data, build the physical simulation model based on the physical characteristics and operating conditions of the actual environment, and set the initial conditions and boundary conditions; S52: Apply the finite element analysis method to simulate the physical simulation model and calculate the state of environmental changes within a predetermined time step. The calculation formula of the finite element analysis is: ,in, is the stiffness matrix, is the displacement vector, is the force vector; S53: storing and analyzing the results of each time step during the simulation process, and adjusting the parameters of the simulation model to improve the simulation accuracy by comparing the simulation results with the actual data; S54: Deploy the machine learning model on the digital twin platform, select the long short-term memory network algorithm, and train and predict the fused time series data; S55: Use the trained machine learning model to predict the future environmental state and generate prediction results, including the future trend of temperature, humidity, and pressure difference parameters; S56: Compare and integrate the results of the physical simulation and machine learning models, combine the accuracy of the physical simulation model and the prediction ability of the machine learning model, and generate the final dynamic simulation and prediction results; S57: Visualize the final simulation and prediction results and display the dynamic changes of the environment through the visualization tools of the digital twin platform; The S56 specifically includes: S561: Get simulation results within the same time step from the physical simulation module and the machine learning model and prediction results ; S562: Calculate the weights of the physical simulation results and the machine learning prediction results, adjust based on the accuracy of historical data and the reliability of the model, and set the weights of the physical simulation model and machine learning model weights , the weights must be satisfied ; S563: Apply the weighted average fusion algorithm to fuse the physical simulation results and the machine learning prediction results to calculate the final dynamic simulation and prediction results , the formula is: ,in, is the time step The physical simulation results are is the time step The machine learning prediction results, and are the weights of physical simulation results and machine learning results respectively; S564: Perform error analysis on the weighted fusion result and calculate the error between the fusion result and the actual data , the error calculation formula is: ,in, is the time step Actual data of is the number of data samples; S565: According to the results of the error analysis, the weights of the physical simulation model and the machine learning model are adjusted to optimize the weight allocation strategy; S566: Repeat the above weighted fusion and error analysis steps, iteratively optimize, and finally generate high-precision dynamic simulation and prediction results; The S7 specifically includes: S71: Continuously collect real-time data from sensors and data acquisition equipment, monitor environmental changes and equipment operating status in real time, including temperature, humidity, pressure difference and location data, to ensure comprehensive monitoring of the environment; S72: Regularly compare the collected real-time data with the simulation results, use data comparison tools and statistical analysis methods to evaluate the accuracy and consistency of the digital twin model, and identify potential deviations and errors, including comparing the trends, fluctuation ranges and characteristics of the actual data and the simulated data; S73: Based on the evaluation results, use historical data and real-time data to adjust the parameters of the digital twin model, specifically adjusting and updating the physical properties of the model, including material properties, fluid mechanics parameters, initial conditions and boundary conditions to reflect the latest environmental status; S74: Deploy an automatic calibration module on the digital twin platform to calibrate the model through an iterative optimization algorithm. The algorithms used in the calibration process include particle swarm optimization and genetic algorithm. The deviation between the model and the actual environment is gradually reduced through multiple iterations. S75: Verify the calibrated model by comparing the simulation results and actual data before and after calibration to ensure the effectiveness and accuracy of the calibration process. The verification process includes using an independent test data set to simulate and observe the match between the model response and the actual data. S76: Based on the verification results, optimize the calibration and parameter adjustment strategies to ensure the high accuracy and high reliability of the model in a dynamic environment. The optimization strategies include adjusting the parameters of the optimization algorithm and increasing the frequency of data acquisition.

2. The data acquisition and processing method based on the digital twin platform according to claim 1 is characterized in that: The S2 specifically includes: S21: Based on the monitoring requirements and layout design of the actual environment, various sensors and data acquisition equipment are reasonably distributed in the monitoring area to ensure that all predetermined locations are covered; S22: Use high-precision temperature sensors installed in areas of the environment susceptible to temperature changes to measure ambient temperature in real time; S23: Deploy humidity sensors in the transition area between the humid environment and the dry environment to monitor the changes in environmental humidity in real time; S24: Install differential pressure gauges at stress-bearing structures and areas where differential pressure changes, to collect differential pressure data in the environment in real time; S25: Use high-precision position sensors to locate dynamic equipment and personnel to ensure real-time acquisition of position data. The position sensor achieves high-precision positioning by combining GNSS modules and indoor positioning technology; S26: Install video surveillance equipment in the monitoring area and collect environmental video data in real time through cameras; S27: Integrate data acquisition devices of various sensors to build a unified data acquisition network to ensure that the data of each sensor can be transmitted to the edge computing device in real time for preliminary processing and storage.

3. The data acquisition and processing method based on the digital twin platform according to claim 1 is characterized in that: The S3 specifically includes: S31: Deploy a data filtering module on the edge computing device to pre-process the raw data collected by the sensor, remove noise and invalid data, and ensure data clarity; S32: Using a preset outlier detection algorithm, perform outlier detection on the filtered data on the edge computing device to identify abnormal data points that are significantly different from normal data distribution; S33: Mark and record the detected abnormal data on the edge computing device, and store the abnormal data separately from the normal data to facilitate subsequent analysis and processing; S34: performing preliminary aggregation processing on the filtered and outlier-detected data, grouping and counting the data based on the time window and data type, and generating preliminary aggregated data; S35: Perform preliminary analysis on the aggregated data on the edge computing device and calculate statistical indicators, including the average, maximum, minimum, and standard deviation; S36: Transmit the preliminarily processed data to the digital twin platform through edge computing devices to ensure the timeliness and continuity of the data.

4. The data acquisition and processing method based on the digital twin platform according to claim 1 is characterized in that: The S4 specifically includes: S41: First, the preliminarily processed data is transmitted to the digital twin platform through a secure network protocol; S42: Deploy a multimodal data fusion module on the digital twin platform to receive data from different sensors and perform time alignment and spatial alignment processing on them to ensure the consistency of the data in time and space; S43: Apply weighted average fusion algorithm to perform preliminary fusion of multi-source data and set weights for different sensor data , calculate the weighted average , the formula is: ,in, For the The data of the sensors, is the corresponding weight, is the number of sensors; S44: Use the principal component analysis algorithm to perform dimensionality reduction processing on the initially fused data to extract features. The specific calculation formula is: ,in, is the input data matrix, is the eigenvector matrix, is the data after dimensionality reduction; S45: Apply data enhancement technology based on generative adversarial networks to enhance the reduced-dimensional data to generate representative data samples. The formula for generating adversarial networks is: ,in, For the generator, is the discriminator, is the real data distribution, is the noise distribution; S46: Perform consistency check on the enhanced data to ensure consistency and accuracy during data fusion and enhancement. The specific use is the correlation coefficient To test, the formula is: ,in, and are the sample values ​​of the two data sets respectively, and are the means of the two data sets respectively; S47: Finally, the fused and enhanced data is stored in the database of the digital twin platform for subsequent simulation, prediction and analysis. At the same time, a data structure based on the sensor location information is created for the device. The data in the data structure is the data obtained by the sensor near the device and the corresponding standard value range.

5. The data acquisition and processing method based on the digital twin platform according to claim 1 is characterized in that: The S6 specifically includes: S61: Obtaining dynamic simulation and prediction results , parameter changes in the analysis results, including temperature, humidity, and pressure difference; S62: Generate an optimized control instruction using a preset control algorithm according to the parameter change. The control algorithm formula is: ,in, is the time step control instructions, is the error between the expected value and the actual value, , and are proportional, integral and differential control parameters respectively; S63: Sending the generated optimization control instructions to the devices and systems in the actual environment through a secure communication protocol; S64: After receiving the optimization control instruction, the equipment and system adjust corresponding operating parameters according to the instruction, including adjusting parameters of temperature, humidity, and pressure difference, so as to achieve environmental optimization; S65: Real-time monitoring of equipment and system parameter adjustment results, and feedback of real-time data through sensors To the digital twin platform; S66: Real-time feedback data on the digital twin platform Analyze and judge the effect of equipment and system parameter adjustment to see whether the expected optimization effect is achieved; if the expected optimization effect is not achieved, adjust the control algorithm parameters , and , and regenerate optimization control instructions , iterative optimization is carried out through the feedback mechanism until the expected effect is achieved.

6. A data acquisition and processing system based on a digital twin platform, used to implement a data acquisition and processing method based on a digital twin platform as described in any one of claims 1 to 5, characterized in that: Includes the following modules: Data acquisition module: It is equipped with various types of sensors and data acquisition devices, including temperature sensors, humidity sensors, differential pressure gauges, position sensors, and video monitoring devices, to collect multi-source data in the actual environment in real time and transmit it to the edge computing module through wireless or wired networks; Edge computing module: receives data from the data acquisition module and performs preliminary processing, including data filtering, outlier detection, and aggregation processing; Data fusion module: deployed on the digital twin platform, receives processed data from the edge computing module, and uses a multimodal data fusion algorithm to fuse and enhance the data to ensure data consistency and integrity; Simulation and prediction module: Using the fused data, the preset physical simulation algorithm and machine learning model are run on the digital twin platform to dynamically simulate and predict the actual environment, and generate detailed environmental status reports and future trend forecasts; Optimization control module: According to the results generated by the simulation prediction module, the optimization control instructions are formulated, and the equipment and system parameters in the actual environment are adjusted in real time through the feedback mechanism to achieve optimization and control of the environment; Monitoring and calibration module: Continuously monitor the operating status and data accuracy of the entire system, regularly calibrate the model and adjust parameters to ensure the accuracy of the digital twin model and the long-term stable operation of the system.

Images