WEB end data processing method and device based on digital twinning and medium

Abstract

The invention discloses a WEB end data processing method and device based on digital twinning and a medium, and the method comprises the steps: determining a physical end system which carries out data interaction with a WEB end, collecting a state data set, and carrying out the preprocessing; calculating a quality coefficient of the state data set, performing interpolation completion on deleted abnormal values, and outputting a complete state data set; screening a communication protocol for interaction between the edge gateway and the WEB end according to the service quality coefficient; generating a random number key; and the WEB end and the edge gateway carry out matching according to the session key, the state data in the complete state data set updates the digital twin model, and the state data before updating is sent to the edge gateway for storage. The device comprises at least one processor, a sensing layer is built, and the processor is connected with an edge gateway. The medium stores a computer program for calculating a method. The pressure of WEB end data processing is reduced, and the updating quality of the digital twin model is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of WEB end digital twin application, and particularly relates to a WEB end data processing method based on digital twin, equipment and medium. BACKGROUND

[0002] Digital twin technology realizes real-time monitoring, simulation analysis and prediction optimization of the state of physical entities by constructing a virtual mapping model of the physical entities, and has been widely applied in industrial manufacturing, smart city and other fields. As an important interactive entrance of the digital twin system, the WEB end needs to receive multi-source heterogeneous data from sensors, PLCs (programmable logic controllers), historical databases, third-party systems and the like, and complete data processing, model driving and visual presentation.

[0003] The existing WEB end digital twin model has the following defects in operation:

[0004] Multi-source heterogeneous data processing is difficult, different data types have high requirements for the data processing capability of the WEB end, the WEB end data processing pressure is too large, which leads to delay in updating the WEB end digital twin model, and even errors;

[0005] WEB end resource consumption is too high: using full data transmission and caching strategy, when the digital twin model contains 1000 + monitoring points, the WEB end bandwidth occupancy is more than 50Mbps, and the browser memory occupancy is more than 2GB, which easily leads to page lag or crash;

[0006] Data quality lacks dynamic control: only through static threshold to filter abnormal data, it is impossible to real-time check data integrity and consistency, and abnormal data repair depends on manual completion, which leads to the disconnection between the digital twin model and the physical entity.

[0007] The WEB end data interaction is low in security and cannot cope with the security data interaction and transmission of multiple types and large number of physical end system entities. SUMMARY

[0008] In view of the above deficiencies in the prior art, the present application provides a WEB end data processing method based on digital twin, equipment and medium, which processes the data before it enters the WEB end, ensures the accuracy and effectiveness of the state data entering the WEB end, and makes the data interaction between the WEB end and the physical end system safe enough.

[0009] In order to achieve the above application purpose, the technical scheme adopted by the present application is:

[0010] A WEB end data processing method based on digital twin is provided, which comprises:

[0011] Step S1: Determine the physical end system for data interaction with the WEB end, collect the state data of the physical end system entity using the perception layer, generate a state data set, and preprocess the state data using the edge gateway constructed by the physical end system, interpolate and fill in the missing values in the state data set, and delete abnormal values;

[0012] Step S2: Calculate the integrity index, consistency index and deviation rate of the state data set, and calculate the quality coefficient of the state data set, evaluate whether the state data set meets the requirements, interpolate and fill in the deleted abnormal values, and output the complete state data set;

[0013] Step S3: The edge gateway sends a connection request to the WEB end for building a digital twin model, and calculates the service quality coefficient of the historical communication protocol, and selects the communication protocol for interaction between the edge gateway and the WEB end according to the service quality coefficient;

[0014] Step S4: The WEB end dynamically updates the key according to the interaction risk of the selected communication protocol, and generates a random number key, and the edge gateway calculates the session key of the edge gateway end according to the random number key; The edge gateway also sends the random number key to the WEB end to calculate the session key of the WEB end;

[0015] Step S5: The WEB end and the edge gateway match according to the session key, after successful matching, the edge gateway sends the complete state data set to the WEB end, updates the digital twin model according to the state data in the complete state data set, and sends the state data before updating to the edge gateway for storage.

[0016] Further, step S1 includes:

[0017] Step S11: Determine the physical end system for data interaction with the WEB end, build a perception layer according to the physical end system, collect state data for building a digital twin model using the perception layer at a fixed frequency, and obtain a state data set of the physical end system in a set period , K is the amount of state data collected in the set period, is the state data with timestamp ;

[0018] Step S12: Build an edge gateway in the physical end system, preprocess the state data collected by the perception layer using the edge gateway, calibrate the collected state data using the network time protocol, and linearly interpolate and fill in the state data with missing timestamps;

[0019] ;

[0020] Where, t is the missing timestamp, is the filled state data, two known time stamps adjacent to the missing time stamp t, respectively, are known time stamps corresponding known state data;

[0021] Step S13: obtain the interpolated and completed state data set A, and according to the state data in the set time period a threshold value, screen the abnormal values in the state data set A, delete the abnormal values in the state data set A, and retain the normal values;

[0022] When the state data in the state data set A satisfies , it is determined that the state data is an abnormal value, otherwise, it is determined that the state data is a normal value; is an error allowed value of the state data.

[0023] Further, step S2 comprises:

[0024] Step S21: according to the number u of normal values in the state data set A, calculate the integrity index of the state data collected in the set time period , i is the type of state data;

[0025] Step S22: according to the normal values in the state data set A calculate the consistency index of the state data collected in the set time period ;

[0026] ;

[0027] Step S23: according to the average value of the state data collected in each set time period in the history calculate the deviation rate of the state data collected in the set time period ;

[0028] ;

[0029] Step S24: according to the integrity index , the consistency index and the deviation rate , calculate the quality coefficient of the state data collected in the set time period ;

[0030] ;

[0031] wherein, respectively, are the influence weights of integrity, consistency and deviation rate on the quality of collected state data; w is the set time period number of historical collected state data, and W is the set time period number of historical collected state data, The integrity index, the consistency index and the deviation rate corresponding to the state data collected in each historical setting period respectively;

[0032] Step S25: setting a quality coefficient threshold of the state coefficient ;

[0033] If , it is determined that the state data in the state data set A meets the requirements, the abnormal values deleted in the state data set A are filled by linear interpolation, and the complete state data set is output ;

[0034] Specifically:

[0035] Step S251: outputting the state data set A with deleted abnormal values, and filling the deleted abnormal values by linear interpolation to output the complete state data set ;

[0036] ;

[0037] Wherein, is the state data replacing the abnormal value, are normal values on both sides of the abnormal value respectively;

[0038] Otherwise, it is determined that the state data in the state data set A does not meet the requirements, a prediction model is constructed by using the state data collected in the historical setting period, the state data is predicted, and the normal values in the state data set A are compared and replaced, and the complete state data set is output .

[0039] Further, a prediction model is constructed by using the state data collected in the historical setting period, the state data is predicted, and the normal values in the state data set A are compared and replaced, and the complete state data set is output , and specifically includes the following steps:

[0040] Step S27: constructing a prediction model for predicting state data based on the state data collected in the continuous historical setting period;

[0041] ;

[0042] Wherein, is a constant term, is an error term, is an autoregressive coefficient, is the number of continuous historical setting periods, is the state data corresponding to the timestamp in the wth historical setting period, is the state data corresponding to the timestamp in the predicted setting period;

[0043] Step S28: Divide the W historical set time periods into several consecutive historical set time periods, and take each consecutive historical set time period as a fitting data period. Each fitting data period includes... Each historical time period was defined; timestamps were extracted from each historical time period within the fitted data period. The corresponding state data is used as a set of fitting data, and several fitting data periods correspond to several sets of fitting data.

[0044] Step S29: Extract the timestamps from each set of fitted data. The corresponding state data is input into the prediction model, and the autoregressive coefficients are fitted using the least squares method. Then, the error term and constant term are calculated to obtain the fitted prediction model;

[0045] Step S210: Set the time period Previous consecutive Timestamps collected during a historical time period The corresponding state data is input into the fitted prediction model, and the output is a timestamp. Predicted state data ;

[0046] Step S211: Transfer the normal values ​​from the state dataset A With predicted state data Perform the difference, if Then the normal value Replace with predicted state data Otherwise, retain the normal value. ;

[0047] Step S212: Output the state dataset after replacing the state data. And execute step S251 to transfer the state dataset. The deleted outliers are filled in using linear interpolation, and the complete state dataset is output. .

[0048] Further, step S3 includes:

[0049] Step S31: The edge gateway of the physical system obtains the complete state dataset. Then, a connection request is sent to the web client that builds the digital twin model, and the historical communication protocol types of the interaction between the historical edge gateway and the web client are retrieved.

[0050] Step S32: Calculate the quality of service coefficient of each communication protocol based on historical network status data for each protocol. ;

[0051] ;

[0052] wherein e is the type of the communication protocol, is the number of data packet loss of the interaction process, is the total number of data packets, is the delay of data transmission, is the available bandwidth, is the total bandwidth of the communication protocol, is the weight of the packet loss rate, the delay and the available bandwidth of the data transmission on the quality of service, respectively;

[0053] Step S33: obtaining the quality of service coefficient corresponding to each communication protocol in history, and taking the communication protocol X corresponding to the maximum value of the quality of service coefficient as the communication protocol for the interaction between the edge gateway and the WEB end this time.

[0054] Further, step S4 comprises:

[0055] Step S41: the edge gateway requests the WEB end to interact by using the communication protocol X, and the WEB end dynamically updates the key according to the risk of the interaction with the edge gateway, and the period of dynamically updating the key is;

[0056] ;

[0057] wherein, is the basic update period of the key, is the risk value of the interaction between the edge gateway and the WEB end by using the communication protocol X, is the risk control coefficient, is the number of abnormal requests for requesting the WEB end to interact by using the communication protocol X, is the total number of historical requests for interacting with the WEB end by using the communication protocol X;

[0058] Step S42: the WEB end dynamically updates the random number key m1 according to the period and sends it to the edge gateway, and the edge gateway combines the random number key m to calculate the session key ; is the random number key generated by the edge gateway, is the round function, is the round function operator, is the preset root key, and the session key is the session key of the edge gateway end;

[0059] Step S43: the edge gateway sends the random number key to the WEB end, and the WEB end also calculates the session key as the session key of the WEB end.

[0060] Further, step S5 comprises:

[0061] Step S51: session key with session key is matched, if session key is same, then WEB end and edge gateway are matched successfully, and step S52 is executed; otherwise, the matching is unsuccessful, and WEB end rejects the connection request of edge gateway, and feedback instruction exception;

[0062] Step S52: edge gateway sends complete state data set to WEB end, and updates digital twin model according to state data in state data set , and sends state data before updating to edge gateway for storage.

[0063] An electronic device is provided, comprising:

[0064] at least one processor, a perception layer is built in the processor, the perception layer is connected with a physical end system entity, and state data of the physical end system entity is sent to the perception layer;

[0065] the processor is connected with an edge gateway, and the perception layer sends the state data to the edge gateway;

[0066] a memory connected in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the above-mentioned WEB end data processing method based on digital twin is run.

[0067] A computer readable storage medium is provided, and the computer readable storage medium stores a computer program, and when the computer program is run by a processor, the above-mentioned WEB end data processing method based on digital twin is executed.

[0068] The beneficial effects of the present application are:

[0069] The present application builds a data preprocessing scheme in the front end of the WEB end, combines a prediction model to correct the state data of each data source, and can realize multi-dimensional data, ensures that the state data has sufficient quality before entering the digital twin model, guarantees the completeness of the state data input into the WEB end digital twin model, reduces the pressure of WEB end data processing, increases the smoothness of digital twin model operation, and constructs an end-to-end WEB end digital twin data processing method.

[0070] ​The scheme avoids the WEB end and the edge gateway from interacting with each other by using a traditional fixed protocol, sets a network dynamic selection optimal protocol, effectively increases the quality of data interaction, ensures sufficient data interaction security, reduces the pressure of WEB end data processing, and improves the quality of digital twin model updating. BRIEF DESCRIPTION OF DRAWINGS

[0071] Figure 1 A flowchart of the WEB end data processing method based on digital twin. DETAILED DESCRIPTION

[0072] The specific embodiments of the application are described below to facilitate those skilled in the art to understand the application, but it should be clear that the application is not limited to the scope of the specific embodiments, and for those skilled in the art, as long as various changes are within the spirit and scope of the application defined and determined by the appended claims, these changes are obvious, and all applications utilizing the concept of the application are within the scope of protection.

[0073] As shown in Figure 1 , a WEB end data processing method based on digital twin includes:

[0074] Step S1: determining a physical end system for data interaction with the WEB end, collecting state data of the physical end system entity by using a perception layer, generating a state data set, and pre-processing the state data by using an edge gateway constructed by the physical end system, interpolating and filling in missing values in the state data set, and deleting abnormal values.

[0075] Step S1 specifically includes the following steps:

[0076] Step S11: determining a physical end system for data interaction with the WEB end, building a perception layer according to the physical end system, collecting state data for building a digital twin model by using the perception layer at a fixed frequency, and obtaining a state data set of the physical end system in a set period , K is the amount of state data collected in the set period, is the state data of the timestamp ;

[0077] The perception layer of this embodiment is a multi-modal sensor cluster (such as a vibration sensor, a temperature sensor, a current sensor) of the physical end system, a PLC controller, and an actuator (such as a motor, a valve); state data (such as bearing vibration value of industrial fan, stator temperature) of the physical end system entity is collected, and control instructions issued by the WEB end are received and executed;

[0078] Step S12: An edge gateway is constructed in a physical end system, state data collected by a perception layer is preprocessed by using the edge gateway, collected state data is time-stamped calibrated by using a network time protocol, and missing time-stamped state data is linearly interpolated and filled in;

[0079] ;

[0080] wherein t is a missing time stamp, is filled-in state data, are two known time stamps adjacent to the missing time stamp t, are known state data corresponding to the known time stamps , respectively.

[0081] Step S13: Obtaining the interpolated and filled-in state data set A, and screening abnormal values in the state data set A according to a threshold value of state data in a set period, deleting the abnormal values in the state data set A, and retaining normal values.

[0082] When the state data in the state data set A satisfies , the state data is determined as an abnormal value, otherwise, the state data is determined as a normal value. is an error allowable value of the state data.

[0083] Step S2: Calculating a completeness index, a consistency index and a deviation rate of the state data set, calculating a quality coefficient of the state data set, evaluating whether the state data set meets requirements, interpolating and filling in the deleted abnormal values, and outputting a complete state data set.

[0084] Step S2 specifically includes the following steps:

[0085] Step S21: According to the number u of normal values in the state data set A, calculating a completeness index of state data collected in a set period, i being a type of the state data.

[0086] Step S22: According to the normal values in the state data set A, calculating a consistency index of state data collected in the set period.

[0087] ;

[0088] Step S23: According to average values of state data collected in each set period in history, calculating a deviation rate of state data collected in the set period.

[0089] ;

[0090] Step S24: Calculate the quality coefficient of the state data collected in the set period according to the integrity index , the consistency index , and the deviation rate ; ;

[0091] ;

[0092] Wherein, are the influence weights of the integrity, consistency, and deviation rate on the quality of the collected state data, respectively, and ; w is the set period number of the historical collected state data, and W is the set period number of the historical collected state data, are the integrity index, consistency index, and deviation rate corresponding to the state data collected in each set period of the historical state data, respectively;

[0093] Step S25: Set the quality coefficient threshold of the state coefficient ;

[0094] If , it is determined that the state data in the state data set A meets the requirements, the abnormal values deleted in the state data set A are linearly interpolated and filled, and the complete state data set is output ;

[0095] Specifically,

[0096] Step S251: Output the state data set A with deleted abnormal values, and fill the deleted abnormal values by linear interpolation to output the complete state data set ;

[0097] ;

[0098] Wherein, is the state data replacing the abnormal value, are the normal values on both sides of the abnormal value, respectively;

[0099] Otherwise, it is determined that the state data in the state data set A does not meet the requirements, a prediction model is constructed by using the state data collected in the historical set period to predict the state data, and the normal values in the state data set A are compared and replaced to output the complete state data set ;

[0100] Specifically, the following steps are included:

[0101] Step S27: Construct a prediction model for predicting state data based on the state data collected in the continuous historical set period;

[0102] ;

[0103] wherein, is a constant term, is an error term, is an auto-regressive coefficient, is a number of consecutive historical setting periods, is a timestamp corresponding state data in the wth historical setting period, is a predicted setting period timestamp corresponding state data;

[0104] Step S28: splitting the W historical setting periods into several consecutive historical setting periods, taking each consecutive historical setting period as a fitting data period, each fitting data period including historical setting periods; extracting the timestamp corresponding state data in each historical setting period in the fitting data period as a set of fitting data, and the several fitting data periods correspond to several sets of fitting data;

[0105] Step S29: inputting the timestamp corresponding state data in each set of fitting data into the prediction model, fitting the auto-regressive coefficient using the least square method, and calculating the error term and the constant term to obtain the fitted prediction model;

[0106] For example, we take the state data collected every hour in the middle of the day in two months of 60 days to fit the prediction model of the state data, set the setting period to one day, the timestamp corresponds to each whole point in the day, and the state data collected in the setting period includes 24 state data. If we take the state data collected at the whole point in the consecutive 5 days as a set of fitting data, then 12 sets of fitting data periods can be split, including days 1-5, 6-10, 11-15, …; inputting the state data collected at 12 o'clock in the 12 sets of fitting data periods into the prediction model, and using the least square method can fit the auto-regressive coefficient corresponding to the state data collected at 12 o'clock in the consecutive 5 days, and calculate the error term and the constant term to fit the prediction model of the state data collected at 12 o'clock in the future each day.

[0107] Step S210: inputting the timestamp and the timestamp corresponding state data collected in the consecutive historical setting periods before the setting period into the fitted prediction model, and outputting the predicted state data corresponding to the timestamp ;

[0108] Step S211: replace the normal value in the state data set A with the predicted state data Step S212: output the state data set after replacing the state data Step S213: if the difference between the normal value and the predicted state data is greater than a threshold value , replace the normal value with the predicted state data , otherwise, keep the normal value ; ;

[0109] Step S212: output the state data set after replacing the state data , and perform step S251 to fill in the deleted abnormal value in the state data set A using linear interpolation, and output the complete state data set .

[0110] Step S3: the edge gateway sends a connection request to the WEB side that builds the digital twin model, and calculates the service quality coefficient of the historical communication protocol, and filters the communication protocol for interaction between the edge gateway and the WEB side according to the service quality coefficient.

[0111] Step S3 specifically includes the following steps:

[0112] Step S31: the edge gateway of the physical side system obtains the complete state data set A , then sends a connection request to the WEB side that builds the digital twin model, and calls the historical communication protocol type for interaction between the edge gateway and the WEB side;

[0113] Step S32: according to the network state data of each historical communication protocol, calculate the service quality coefficient of the communication protocol ;

[0114] ;

[0115] Wherein, e is the type of communication protocol, is the number of data packet loss in the interaction process, is the total number of data packets, is the delay of data transmission, is the available bandwidth, is the total bandwidth of the communication protocol, respectively, the loss rate, delay and available bandwidth of data transmission, the influence weight of service quality, in this embodiment, ; the larger the service quality coefficient , the higher the data transmission quality for interaction using the communication protocol, and vice versa, and the exp function is used to amplify the service quality coefficient, so that the high-quality communication protocol is highlighted.

[0116] ​Step S33: Obtain the service quality coefficient corresponding to each communication protocol in history, and take the communication protocol X corresponding to the maximum value of the service quality coefficient as the communication protocol for the edge gateway to interact with the WEB end this time.

[0117] The communication protocol for the edge gateway to interact with the WEB end includes: WebSocket protocol (low delay, supports bidirectional real-time communication); MQTT protocol (lightweight, supports message subscription / publishing, packet retransmission); and CoAP protocol (ultra-lightweight, suitable for low-bandwidth and high-packet loss scenarios). The recent optimal communication protocol for the edge gateway to interact with the WEB end is screened through the communication protocol used in recent history, so as to adapt to the dynamically changing network state.

[0118] Step S4: The WEB end dynamically updates the key according to the interaction risk of the screened communication protocol, generates a random number key, and the edge gateway calculates the session key of the edge gateway end according to the random number key; the edge gateway also sends the random number key to the WEB end to calculate the session key of the WEB end.

[0119] Step S4 specifically includes the following steps:

[0120] Step S41: The edge gateway requests the WEB end to interact using the communication protocol X, and the WEB end dynamically updates the key according to the risk of interaction with the edge gateway, and the period of dynamically updating the key

[0121]

[0122] Among them, is the basic update period of the key, is the risk value of the edge gateway and the WEB end interacting using the communication protocol X, is a risk control coefficient, generally , is the number of abnormal requests for the WEB end to interact using the communication protocol X, is the total number of historical interactions between the WEB end using the communication protocol X;

[0123] Step S42: The WEB end dynamically updates the random number key m1 according to the period and sends it to the edge gateway, and the edge gateway combines the random number key m to calculate the session key . is the random number key generated by the edge gateway, is a round function, is a round function operator, is a preset root key, and the session key is the session key of the edge gateway end;

[0124] ​​Step S43: The edge gateway sends the random number key to the WEB end, and the WEB end also calculates the session key as the session key of the WEB end.

[0125] Step S5: The WEB end and the edge gateway match according to the session key, after successful matching, the edge gateway sends the complete state data set to the WEB end, updates the digital twin model according to the state data in the complete state data set, and sends the state data before updating to the edge gateway for storage. Step S5 specifically includes the following steps:

[0126] Step S51: The session key matches the session key , if the session key is the same as the session key , then the WEB end and the edge gateway match successfully, and step S52 is executed; otherwise, the matching is unsuccessful, the WEB end rejects the connection request of the edge gateway, and feedbacks the instruction exception;

[0127] Step S52: The edge gateway sends the complete state data set to the WEB end, updates the digital twin model according to the state data in the state data set , and sends the state data before updating to the edge gateway for storage.

[0128] An electronic device comprises:

[0129] At least one processor, a perception layer is built in the processor, the perception layer is connected with a physical end system entity, and state data of the physical end system entity is sent to the perception layer.

[0130] The processor is connected with an edge gateway, and the perception layer sends the state data to the edge gateway.

[0131] A memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to run the above-mentioned WEB end data processing method based on digital twin.

[0132] A computer readable storage medium stores a computer program, and the computer program is executed by the processor to execute the above-mentioned WEB end data processing method based on digital twin.

[0133] The application builds a data preprocessing scheme on the WEB front end, corrects the state data of each data source in combination with a prediction model, and can realize multidimensional data, ensures sufficient quality before the state data enters the digital twin model, guarantees the completeness of the state data input into the WEB digital twin model, reduces the pressure of WEB data processing, increases the smoothness of digital twin model operation, and builds an end-to-end WEB digital twin data processing method.

[0134] The application avoids the WEB end and the edge gateway to interact by using the traditional fixed protocol, sets the network to dynamically select the optimal protocol, effectively increases the quality of data interaction, and ensures sufficient data interaction security, reduces the pressure of WEB data processing, and improves the quality of digital twin model updating.

Claims

1. A WEB-side data processing method based on digital twinning, characterized in that, The method comprises the following steps: Step S1: determining a physical end system for data interaction with a WEB end, collecting state data of the physical end system entity by using a perception layer, generating a state data set, and pre-processing the state data by using an edge gateway constructed by the physical end system, and interpolating and filling in missing values in the state data set and deleting abnormal values; Step S2: calculating the integrity index, consistency index and deviation rate of the state data set, and calculating the quality coefficient of the state data set, evaluating whether the state data set meets the requirements, interpolating and filling in the deleted abnormal values, and outputting the complete state data set; Step S3: the edge gateway sends a connection request to the WEB end for building a digital twin model, and calculates the service quality coefficient of the historical communication protocol, and selects the communication protocol for interaction between the edge gateway and the WEB end according to the service quality coefficient; Step S4: the WEB end dynamically updates the key according to the interaction risk of the selected communication protocol, and generates a random number key, and the edge gateway calculates the session key of the edge gateway end according to the random number key; The edge gateway also sends the random number key to the WEB end to calculate the session key of the WEB end; Step S5: the WEB end and the edge gateway match according to the session key, and after successful matching, the edge gateway sends the complete state data set to the WEB end, updates the digital twin model according to the state data in the complete state data set, and sends the state data before updating to the edge gateway for storage.

2. The WEB-based data processing method based on digital twinning according to claim 1, characterized in that, The step S1 comprises: Step S11: determining a physical end system for data interaction with the WEB end, building a perception layer according to the physical end system, collecting state data of the built digital twin model at a fixed frequency by using the perception layer, and obtaining a state data set of the physical end system in a set period , K is the amount of state data collected in a set period, is the state data of the timestamp ; Step S12: constructing an edge gateway in the physical end system, pre-processing the state data collected by the perception layer by using the edge gateway, timestamp calibration of the collected state data by using the network time protocol, and linear interpolation and filling of the state data with missing timestamps; ； wherein t is a missing timestamp, is the status data to be completed, are two known timestamps adjacent to the missing timestamp t, are the known timestamps respectively, is the corresponding known status data; Step S13: Obtain the interpolated and padded state data set A, and select the state data in the set time period according to the set threshold The threshold value is used to filter the abnormal values in the state data set A, delete the abnormal values in the state data set A, and retain the normal values. When the state data in the state data set A satisfies , the state data is determined to be an abnormal value, otherwise, the state data is determined to be a normal value; is an error allowable value of the state data.

3. The WEB-based data processing method based on digital twinning according to claim 2, characterized in that, The step S2 comprises: Step S21: According to the number u of normal values in the state data set A, the integrity index of the state data collected in the setting period is calculated i is the kind of state data; Step S22: determining, according to the normal values in the state data set A calculating the consistency index of the state data collected in the set period ; ； Step S23: calculating an average value of the state data collected at each setting period from the history calculating a deviation rate of the state data collected at the setting period ; ； Step S24: calculating the integrity index based on the quality coefficient , the consistency index , and the deviation rate calculating the quality coefficient of the state data collected in the set period ; ； wherein, respectively are the influence weights of the integrity, consistency, and deviation rate on the quality of the collected state data; w is the set period number of the historical collected state data, and W is the set period quantity of the historical collected state data, respectively are the integrity index, consistency index, and deviation rate corresponding to the state data collected in each set period. Step S25: Setting a quality coefficient threshold value of the state coefficient ; If , it is determined that the state data in the state data set A meets the requirements, and the missing abnormal values in the state data set A are linearly interpolated and filled, and the complete state data set is output ; Specifically: Step S251: output the state data set A with the abnormal values deleted, and fill in the deleted abnormal values by linear interpolation, output the complete state data set ; ； wherein, the state data for replacing the outlier, are the normal values on either side of the outlier, respectively; Otherwise, it is determined that the state data in the state data set A does not meet the requirement, a prediction model is constructed by using the state data collected in the historical setting period, the state data is predicted, and the normal value in the state data set A is compared and replaced, and a complete state data set is output .

4. The WEB-based data processing method based on digital twinning according to claim 3, characterized in that, The state data collected in the historical setting period is used to construct a prediction model, predict state data, compare and replace normal values in the state data set A, and output a complete state data set , specifically comprising the following steps: Step S27: constructing a prediction model for predicting state data based on state data collected in a continuous historical setting period; ； wherein, is a constant term, is an error term, is an autoregressive coefficient, is the number of consecutive historical set periods, is a timestamp in the wth historical set period corresponding state data, is a predicted set period timestamp corresponding state data; Step S28: split the W historical setting periods into several continuous historical setting periods, and take each continuous historical setting period as a fitting data period, each fitting data period including W historical setting periods; extract the timestamp corresponding state data in each historical setting period in the fitting data period as a set of fitting data, and several fitting data periods correspond to obtain several sets of fitting data; Step S29: fitting the time stamp in each group of fitting data The corresponding state data is input into the prediction model, and the least square method is used to fit the autoregressive coefficients And the error term and the constant term are calculated to obtain the completed fitting prediction model; Step S210: set a time period previous consecutive time stamps collected in the previous consecutive corresponding state data into the fitted prediction model, output time stamps predicted state data ; Step S211: replace normal value in state data set A with predicted state data if ;​​​​​ Step S212: output the state data set after replacing the state data and perform step S251 to output the complete state data set with the missing values filled in by linear interpolation .

5. The WEB-based data processing method based on digital twinning according to claim 4, characterized in that, The step S3 comprises: Step S31: The edge gateway of the physical end system obtains a complete state data set Afterwards, a connection request is sent to the WEB end for building a digital twin model, and a historical communication protocol type for interaction between the historical edge gateway and the WEB end is called. Step S32: According to the historical network state data of each communication protocol, the service quality coefficient of the communication protocol is calculated ; ； wherein e is the type of communication protocol, is the number of data packet loss for the interaction process, is the total number of data packets, is the delay of data transmission, is the available bandwidth, is the total bandwidth of the communication protocol, are the weight of the packet loss rate, the delay and the available bandwidth of the data transmission, respectively, on the quality of service. Step S33: obtaining the service quality coefficient corresponding to each historical communication protocol, and taking the communication protocol X corresponding to the maximum value of the service quality coefficient as the communication protocol for interaction between the edge gateway and the WEB end.

6. The WEB-based data processing method based on digital twinning according to claim 5, characterized in that, The step S4 comprises: Step S41: The edge gateway requests the WEB end to interact using the communication protocol X, and the WEB end dynamically updates the key according to the risk of interaction with the edge gateway, and the period of dynamically updating the key is; ； wherein, is the base update period for the key, is the risk value for the edge gateway and the WEB end using the communication protocol X to interact, is the risk control coefficient, is the number of abnormal requests for the WEB end using the communication protocol X to interact, is the total number of historical requests for the WEB end using the communication protocol X to interact; Step S42: The WEB end updates the random number key m1 according to the period The dynamic update random number key m1, and send to edge gateway, edge gateway combines random number key m to calculate session key ; The random number key generated for the edge gateway, The round function, The round function operator, The preset root key, session key As the session key of the edge gateway end; Step S43: The edge gateway sends the random number key to the WEB end, and the WEB end also calculates the session key , as the session key of the WEB end. ​ 7. The WEB-based data processing method based on digital twinning according to claim 6, characterized in that, The step S5 comprises: Step S51: session key with the session key is matched, if the session key with the session key is the same, the WEB end and the edge gateway are matched successfully, and step S52 is executed; otherwise, the matching is unsuccessful, the WEB end rejects the connection request of the edge gateway, and feedbacks an instruction exception. Step S52: The edge gateway sends the complete state data set to the WEB terminal, updates the digital twin model according to the state data in the state data set, and sends the state data before the update to the edge gateway for storage. Step S52: The edge gateway sends the complete state data set to the WEB terminal, updates the digital twin model according to the state data in the state data set, and sends the state data before the update to the edge gateway for storage.​ 8. An electronic device, comprising: It comprises: At least one processor, wherein a perception layer is built in the processor, the perception layer is connected with a physical end system entity, and state data of the physical end system entity is sent to the perception layer; The processor is connected with an edge gateway, and the perception layer sends the state data to the edge gateway; The memory is in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to run the WEB end data processing method based on the digital twin as claimed in any one of claims 1 to 7.

9. A computer readable storage medium storing a computer program, wherein the computer program is executed by a processor to perform the WEB end data processing method based on the digital twin as claimed in any one of claims 1 to 7.

Images