A multi-point packet loss data automatic reconstruction method for intelligent fusion terminal

Abstract

The application relates to the technical field of intelligent fusion terminal data processing, and discloses a multi-point packet loss data automatic reconstruction method for an intelligent fusion terminal, which comprises the following steps: real-time data stream monitoring, adjacent device data synchronous acquisition, basic environment parameter acquisition, benchmark data feature extraction, data compensation and interpolation repair module construction, dynamic data reconstruction model establishment, and optimized data sequence generation; through combination of adjacent device data compensation and interpolation repair, the method can effectively deal with the data loss problem under the multi-point communication interruption scene, fully utilizes the correlation between adjacent devices and the change trend of environment parameters, and improves the accuracy and reliability of data reconstruction; the design discards the limitations of the traditional single compensation mechanism, avoids system decision deviation caused by data loss, enhances the adaptability and robustness of the system, and is suitable for large-scale industrial scene application.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent fusion terminal data processing, in particular to a multi-point packet loss data automatic reconstruction method for intelligent fusion terminals. BACKGROUND

[0002] Intelligent fusion terminals are key devices for collecting, processing and transmitting data, widely used in industrial automation, energy management and smart cities. Its core function is to obtain real-time operation state data of the target area through distributed sensors and communication networks, and provide support for decision-making. However, in actual application, due to unstable communication links or external interference, packet loss may occur during data collection, resulting in the loss of some key data. This data loss will affect subsequent data analysis and system decision-making, thereby reducing the reliability and efficiency of the system.

[0003] In the prior art, the processing method for data packet loss mainly relies on a single compensation mechanism, such as trend prediction based on historical data or simple mean filling. These methods can alleviate the impact of data loss to some extent, but in complex multi-point communication environment, their adaptability and accuracy are limited. Especially when multiple nodes lose data at the same time, traditional methods are difficult to fully consider the correlation between adjacent devices, resulting in insufficient accuracy of reconstructed data. In addition, existing interpolation repair techniques are usually only suitable for short-term data loss, and the recovery effect is often not ideal for longer data interruption.

[0004] Combining adjacent device data compensation and interpolation repair, the related technology for reconstructing missing collected data due to communication interruption has not been explicitly recorded. Current technical solutions mostly focus on single-point data recovery, lacking overall optimization strategies for multi-point packet loss scenarios. Therefore, how to make full use of the data correlation of adjacent devices and combine efficient interpolation algorithms to achieve high-precision automatic data reconstruction in the case of multi-point communication interruption has become a technical problem to be solved. SUMMARY

[0005] The purpose of the present application is to provide a multi-point packet loss data automatic reconstruction method for intelligent fusion terminals, which solves the problem that the prior art fails to fully consider the correlation between adjacent devices and the interpolation repair algorithm lacks adaptability when data is lost due to communication interruption. This deficiency will lead to a decrease in data reconstruction accuracy, thereby affecting the reliability and efficiency of the system.

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

[0007] A multi-point packet loss data automatic reconstruction method for intelligent fusion terminals, comprising the following steps:

[0008] real-time data stream monitoring on a plurality of collection nodes within a target area;

[0009] synchronous acquisition of proximate device data of the collection nodes within the target area;

[0010] collection of basic environmental parameters of the collection nodes within the target area;

[0011] extraction of the collected data and processing of reference data features of the nodes;

[0012] construction of a data compensation and interpolation repair module;

[0013] obtaining of a predicted compensation value in a future time window according to the constructed data compensation and interpolation repair module;

[0014] establishment of a dynamic data reconstruction model based on the predicted compensation value in the future time window and the reference data features of the nodes;

[0015] generation of an optimized data sequence through the dynamic data reconstruction model;

[0016] data update and synchronization of the collection nodes within the target area through the optimized data sequence.

[0017] Preferably, in the real-time data stream monitoring, sampling frequency parameters, sampling period parameters and sampling data integrity indicators of the collection nodes within the target area are included. The sampling frequency parameters are used to define the number of data collection times per unit time, the sampling period parameters are used to define the time interval of each collection, and the sampling data integrity indicators are used to evaluate the missing proportion of the current collection data.

[0018] Preferably, in the proximate device data acquisition, the proximate devices at least include first-type devices with adjacent physical positions, second-type devices with similar functions and third-type devices with similar network topology structures; the data parameters of the first-type devices include the mean and variance of their historical collection data; the data parameters of the second-type devices include their functional response time and data fluctuation range; and the data parameters of the third-type devices include their network delay and data transmission rate.

[0019] Preferably, in the basic environmental parameter collection, the basic environmental parameters at least include temperature distribution data, humidity distribution data and electromagnetic interference intensity data within the target area. The temperature distribution data are used to reflect the thermal environment change trend within the target area, the humidity distribution data are used to analyze the possible impact on device operation, and the electromagnetic interference intensity data are used to evaluate the stability of the communication link.

[0020] Preferably, the reference data features of each node are obtained by processing, and the processing specifically comprises:

[0021] extracting real-time data stream monitoring results of each collection node in the target area;

[0022] collecting sampling data integrity indicators of each collection node in the target area, and adding an adaptive floating threshold to obtain data integrity features;

[0023] extracting adjacent device data of each collection node in the target area;

[0024] collecting the mean and variance of the historical collection data of the first type of device, and adding an adaptive floating threshold to obtain adjacent device data features;

[0025] collecting the function response time and data fluctuation range of the second type of device, and adding an adaptive floating threshold to obtain function correlation features;

[0026] extracting basic environmental parameters of each collection node in the target area;

[0027] setting a plurality of time window thresholds, and extracting temperature distribution data, humidity distribution data, and electromagnetic interference intensity data using the set time window thresholds;

[0028] collecting environmental parameter data within at least three time window thresholds to obtain environmental features within each time window.

[0029] Preferably, the data compensation and interpolation repair module comprises a data compensation unit, an interpolation repair unit, and a verification unit; the data compensation unit and the interpolation repair unit are connected through a high-speed communication interface; and the interpolation repair unit and the verification unit are connected through a low-delay data channel.

[0030] The data compensation unit is configured to collect adjacent device data change trends, function correlation change trends, and environmental feature change trends within future time windows;

[0031] The interpolation repair unit is configured to analyze data missing conditions in the target area according to the change trends collected by the data compensation unit, and generate predicted compensation values;

[0032] The verification unit is configured to verify the predicted compensation values generated by the interpolation repair unit, and the verification condition is that the change trend data collected by the data compensation unit is taken as a reference, and the historical data change trends collected in the past are taken as a reference, and the verification formula is as follows:

[0033] Pmax·C ≤ Pmax ≤ Pmax·D;

[0034] C is the minimum lower threshold value in the verification formula, and D is the maximum upper threshold value; if the verification formula is established, the prediction compensation value is confirmed, otherwise the prediction compensation value is invalid.

[0035] Preferably, the dynamic data reconstruction model is built by the following steps:

[0036] First, the prediction compensation value, real-time data stream monitoring results in the target area, adjacent device data and basic environmental parameters are received and substituted into the dynamic data reconstruction model;

[0037] The output data sequence in at least five time windows is extracted as a reference;

[0038] The extracted output data sequence is substituted into the dynamic data reconstruction model;

[0039] The demand data characteristics are confirmed and substituted into the dynamic data reconstruction model for dynamic learning;

[0040] Through the dynamic data reconstruction model, data analysis and pattern recognition are performed to generate an optimized data sequence;

[0041] The generated optimized data sequence is subjected to credibility evaluation.

[0042] In the credibility evaluation of the optimized data sequence, the upper layer credibility Tw and the lower layer credibility Tg are confirmed by combining analysis of two kinds of data, and the upper layer credibility Tw is calculated according to the following formula:

[0043] Tw = NUMt×M + (RSSI1+ RSSI 2 +......)× N;

[0044] In the formula, NUMt represents the data integrity characteristics in the target area, RSSI represents the functional correlation characteristics of several adjacent devices, M and N are both predetermined weight values, and the sum of the weight values of M and N is 1;

[0045] The lower layer credibility Tg is calculated according to the following formula:

[0046] Tg = CN1 × K + Pmax× L;

[0047] In the formula, CN1 represents the basic environmental parameter, Pmax represents the prediction compensation value, K and L are both predetermined weight values, and the sum of the weight values of K and L is 1;

[0048] If the verification formulas of the upper layer credibility Tw and the lower layer credibility Tg are both established, the final credibility Ep is determined according to the first credibility Tw and the second credibility Tg according to the following formula:

[0049] Ep = Tg / (Tg + Tw).

[0050] Preferably, the optimized data sequence is generated by the dynamic data reconstruction model, so that the data integrity of each collection node in the target area is updated according to the generated optimized data sequence, and the functional correlation of the adjacent device is adjusted according to the optimized data sequence.

[0051] Compared with the prior art, the present application has the following beneficial effects:

[0052] 1. By combining adjacent device data compensation and interpolation repair, the data missing problem in the multi-point communication interruption scene can be effectively dealt with, and the correlation between adjacent devices and the change trend of environmental parameters are fully utilized to improve the accuracy and reliability of data reconstruction. This design discards the limitations of traditional single compensation mechanism, avoids the system decision deviation caused by data missing, and enhances the adaptability and robustness of the system, which is suitable for large-scale industrial scene application.

[0053] 2. In use, based on real-time data flow monitoring results, adjacent device data and basic environmental parameters, combined with the constructed data compensation and interpolation repair module, the predicted compensation value is generated and verified, so as to realize the data management of each collection node in the target area. This management method has strong predictability, and a large amount of historical data is verified to avoid system instability caused by data missing. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 is a schematic diagram of the overall process of the method of the present application;

[0055] Figure 2 is a flowchart of the data compensation and interpolation repair module of the present application;

[0056] Figure 3 is a construction flowchart of the dynamic data reconstruction model of the present application;

[0057] Figure 4 is a calculation logic diagram of the credibility evaluation of the present application. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0059] This invention provides a method for automatic reconstruction of multi-point packet loss data in intelligent converged terminals, the specific implementation of which is described in conjunction with the appendix. Figure 1 To be continued Figure 4 Please provide a detailed explanation. For example... Figure 1 As shown, the overall process of this method includes multiple modules and steps, among which the real-time data stream monitoring module, the nearby device data acquisition module, the basic environmental parameter acquisition module, the data compensation unit, the interpolation repair unit, the verification unit, the dynamic data reconstruction model, the optimized data sequence generation module, and the credibility assessment module work together to complete the entire data reconstruction process.

[0060] In practical applications, the first step is to monitor the real-time data streams of multiple acquisition nodes within the target area using a real-time data stream monitoring module. This process requires defining the sampling frequency parameters, sampling period parameters, and data integrity indicators for each acquisition node. The sampling frequency parameter defines the number of times data is collected per unit time, the sampling period parameter defines the time interval between each collection, and the data integrity indicator is used to assess the proportion of missing data in the current collection. The settings of these parameters need to be determined based on the specific requirements of the actual application scenario. For example, in an industrial scenario, if the target area is a large production workshop, a reasonable sampling frequency and period can be set according to the operating characteristics of the equipment within the workshop. Simultaneously, the calculation of the data integrity indicator needs to incorporate historical data; a reasonable threshold range is derived through statistical analysis of historical data to ensure the accuracy of subsequent processing.

[0061] The proximity device data acquisition module is responsible for synchronously acquiring data from neighboring devices of each acquisition node within the target area. Neighboring devices include at least three categories: physically adjacent devices (Category 1), functionally similar devices (Category 2), and devices with similar network topologies (Category 3). Data parameters for Category 1 devices include the mean and variance of their historical data; for Category 2 devices, they include their response time and data fluctuation range; and for Category 3 devices, they include network latency and data transmission rate. In practice, the selection of neighboring devices must be based on the specific layout and device distribution of the target area. For example, in a smart grid system, Category 1 devices might be other sensors within the same substation, Category 2 devices might be measuring instruments with similar functions, and Category 3 devices might be other nodes in the same communication network. The data acquired by the proximity device data acquisition module will be transmitted to subsequent processing modules as the basis for data compensation and interpolation repair.

[0062] The basic environment parameter acquisition module is used to collect temperature distribution data, humidity distribution data and electromagnetic interference intensity data in the target area. The temperature distribution data reflects the thermal environment change trend in the target area, the humidity distribution data analyzes the possible impact on the equipment operation, and the electromagnetic interference intensity data evaluates the stability of the communication link. The collection of these environmental parameters needs to be realized through various sensors arranged in the target area, such as temperature sensors, humidity sensors and electromagnetic interference detectors. The collected environmental parameter data will be divided into several time windows, and extracted through the set time window threshold, so as to obtain the environmental characteristics in each time window. This process needs to be combined with the characteristics of the actual application scene, for example, in a high temperature and high humidity environment, the influence of humidity on the performance of the equipment needs to be focused on, and in a strong electromagnetic interference environment, the influence of electromagnetic interference on the stability of the communication link needs to be analyzed.

[0063] The data compensation unit, the interpolation repair unit and the verification unit jointly constitute a data compensation and interpolation repair module, as shown in Figure 2 The data compensation unit is connected with the interpolation repair unit through a high-speed communication interface, and the interpolation repair unit is connected with the verification unit through a low-delay data channel. The function of the data compensation unit is to collect the change trend of the adjacent equipment data, the change trend of the functional correlation and the change trend of the environmental characteristics in the future time window. The collection of these change trends needs to be comprehensively analyzed in combination with historical data and real-time data, for example, by comparing the change law of the adjacent equipment data in the historical data, the change trend in the future time window is predicted. The interpolation repair unit analyzes the data missing situation in the target area according to the change trend collected by the data compensation unit, and generates a predicted compensation value. The verification unit verifies the predicted compensation value generated by the interpolation repair unit, and the verification condition is to take the change trend data collected by the data compensation unit as the reference, and to take the historical data change trend collected in the past as the reference. The verification formula is Pmax·C ≤ Pmax ≤ Pmax·D, wherein C is the minimum lower threshold and D is the maximum upper threshold. If the verification formula is established, the predicted compensation value is confirmed, otherwise the predicted compensation value is invalid. This process needs to ensure that the data transmission between units is efficient and accurate to ensure the reliability of the predicted compensation value.

[0064] The construction of the dynamic data reconstruction model is shown in Figure 3 First, the predicted compensation value and the real-time data stream monitoring results, the adjacent equipment data and the basic environment parameters in the target area are received and substituted into the dynamic data reconstruction model. The output data sequence in at least five time windows is extracted as a reference, the extracted output data sequence is substituted into the dynamic data reconstruction model, the required data characteristics are confirmed and dynamic learning is performed. Through the dynamic data reconstruction model, data analysis and pattern recognition are performed to generate an optimized data sequence. The optimized data sequence needs to be evaluated by a credibility evaluation module, as shown in Figure 4The credibility evaluation module analyzes by combining the upper layer credibility Tw and the lower layer credibility Tg, the upper layer credibility Tw is calculated by Tw = NUMt x M + (RSSI1+ RSSI2+...) x N, wherein NUMt represents the data integrity feature in the target area, RSSI represents the function correlation feature of several adjacent devices, M and N are both specified weight values, and the sum of the weight values of M and N is 1. The lower layer credibility Tg is calculated by Tg = CN1 x K + Pmax x L, wherein CN1 represents the basic environment parameter, Pmax represents the prediction compensation value, K and L are both specified weight values, and the sum of the weight values of K and L is 1. If the upper layer credibility Tw and the lower layer credibility Tg verification formula are both established, the credibility Ep is finally determined according to the first credibility Tw and the second credibility Tg, and the specific formula is Ep = Tg / (Tg + Tw). This process needs to ensure the data transmission and processing logic between modules to ensure the credibility of the optimized data sequence.

[0065] The optimized data sequence generation module updates the data integrity of each collection node in the target area according to the generated optimized data sequence, and adjusts the function correlation of adjacent devices according to the optimized data sequence. This process needs to combine the characteristics of the actual application scene, for example, in the intelligent traffic system, the control strategy of the traffic signal lamp can be adjusted through the optimized data sequence to improve the management efficiency of the traffic flow. At the same time, the optimized data sequence can also be used to improve the overall operation efficiency and reliability of the system, for example, in the smart grid system, the power dispatching strategy can be optimized through the optimized data sequence to reduce the decision deviation caused by data loss.

[0066] In the entire implementation process, the connection relationship and position relationship between modules need to be clear and reasonable. For example, the real-time data flow monitoring module, the adjacent device data collection module and the basic environment parameter collection module need to be arranged at appropriate positions in the target area to ensure the comprehensiveness and accuracy of data collection. The data compensation unit, the interpolation repair unit and the verification unit need to realize efficient data transmission through high-speed communication interface and low-delay data channel to ensure the timeliness and reliability of the prediction compensation value. The dynamic data reconstruction model and the credibility evaluation module need to be designed through reasonable algorithm and parameter setting to ensure the generation and evaluation process of the optimized data sequence to be scientific and effective. Through the cooperative work of the above modules, the data management of each collection node in the target area can be realized, so as to solve the problem of data loss caused by communication interruption and improve the adaptability and robustness of the system.

[0067] In order to better enable the relevant persons in the technical field to fully understand and implement the present application, the specific implementation principles of the present application are further supplemented in combination with a specific application scene.

[0068] In the energy management system of a certain intelligent industrial park, a plurality of sensor nodes are distributed in the workshop for real-time monitoring of equipment operating status and energy consumption data. Due to the complex workshop environment and the communication link susceptible to electromagnetic interference, data loss often occurs in some nodes. In order to solve this problem, the multi-point packet loss data automatic reconstruction method provided by the application is used for data management.

[0069] Firstly, a real-time data flow monitoring module is arranged in the target workshop. The module monitors the data flow of each collection node in real time by setting the sampling frequency parameter, the sampling period parameter and the sampling data integrity index. The sampling frequency parameter is set to collect 10 data per second, the sampling period parameter is set to 100 milliseconds, and the threshold range of the sampling data integrity index is 95% to 100% according to historical data analysis. When the data integrity of a node is less than 95%, the system determines that the node has data loss and triggers the subsequent compensation and repair process.

[0070] Secondly, the adjacent equipment data collection module starts to work and synchronously acquires the data of the first type of equipment with adjacent physical position, the second type of equipment with similar functions and the third type of equipment with similar network topology structure. For example, the first type of equipment includes other temperature sensors in the same workshop, the historical collection data mean value of which is 25°C and the variance is 0.5°C; the second type of equipment includes pressure sensors with the same function, the functional response time of which is 50 milliseconds and the data fluctuation range is ±2%; the third type of equipment includes other nodes in the same communication network, the network delay of which is 30 milliseconds and the data transmission rate is 1Mbps. These data are transmitted to the data compensation unit through a high-speed communication interface, providing a basis for the generation of subsequent predicted compensation values.

[0071] At the same time, the basic environment parameter collection module collects the temperature distribution data, humidity distribution data and electromagnetic interference intensity data in the workshop. The temperature sensor detects that the average temperature in the current workshop is 30°C, the humidity sensor detects that the relative humidity is 70%, and the electromagnetic interference detector records that the current electromagnetic interference intensity is 5dB. These environmental parameters are divided into several time windows, and the length of each time window is set to 1 minute. Through the set analysis of the data in the three time windows, the current environmental characteristics are obtained, such as the temperature change trend is slowly rising, the humidity tends to be stable, and the electromagnetic interference intensity presents periodic fluctuation.

[0072] Subsequently, the data compensation unit combines historical data and real-time data to predict the change trends of neighboring device data, functional correlation change trends, and environmental feature change trends within the future time window. For example, based on historical data, it is found that when the workshop temperature rises, the data fluctuation range of the pressure sensor will slightly increase, so it is predicted that the data fluctuation range of the pressure sensor within the future time window may expand to ±2.5%. The interpolation repair unit analyzes the data loss in the target area according to these change trends and generates a predicted compensation value. Assuming that a node loses 10 seconds of data due to communication interruption, the interpolation repair unit generates a predicted compensation value for these 10 seconds by combining neighboring device data through a linear interpolation algorithm.

[0073] The verification unit verifies the predicted compensation value generated by the interpolation repair unit. The verification condition is based on the change trend data collected by the data compensation unit and references historical data change trends. For example, if the generated predicted compensation value is 28°C for the temperature data of a certain node, the verification formula is Pmax·C ≤ Pmax ≤ Pmax·D, where C is set to 0.9 and D is set to 1.1. After calculation, 28°C falls within the range of the verification formula, so the predicted compensation value is confirmed to be valid.

[0074] Next, the dynamic data reconstruction model receives the predicted compensation value and the real-time data stream monitoring results, neighboring device data, and basic environmental parameters within the target area. It extracts output data sequences within at least five time windows as a reference and substitutes these data into the dynamic data reconstruction model for dynamic learning. For example, it extracts temperature data sequences within the past 5 minutes, confirms the required data features through pattern recognition algorithms, and generates an optimized data sequence. The optimized data sequence is evaluated by the credibility evaluation module, and the upper-layer credibility Tw and the lower-layer credibility Tg are calculated as follows: Tw = NUMt×M +(RSSI1+ RSSI 2)× N, where NUMt is the data integrity feature, RSSI is the functional correlation feature of neighboring devices, and M and N are 0.6 and 0.4, respectively; Tg = CN1 × K + Pmax× L, where CN1 is the basic environmental parameter, Pmax is the predicted compensation value, and K and L are 0.5 and 0.5, respectively. The final credibility Ep is calculated by the formula Ep = Tg / (Tg + Tw), and if Ep is greater than the preset threshold, the optimized data sequence is confirmed to be reliable.

[0075] Finally, the optimized data sequence generation module updates the data integrity of each collection node in the target area according to the optimized data sequence. For example, after optimization, the missing part of the temperature data of a certain node is filled in, thereby ensuring that the data integrity of the node reaches 98%. In addition, the optimized data sequence is also used to adjust the functional relevance of adjacent equipment. For example, in the workshop, the response time of the pressure sensor is recalibrated through the optimized data sequence to make it more consistent with the actual operating state.

[0076] Throughout the process, the connection relationship and position relationship between the modules are clear and reasonable. The real-time data flow monitoring module, the adjacent equipment data acquisition module, and the basic environment parameter acquisition module are arranged at key positions in the workshop to ensure the comprehensiveness and accuracy of data acquisition. The data compensation unit, the interpolation repair unit, and the verification unit realize efficient data transmission through high-speed communication interfaces and low-delay data channels. The dynamic data reconstruction model and the credibility evaluation module ensure that the optimized data sequence generation process is scientific and effective through reasonable algorithm design and parameter setting.

[0077] Through the above steps, the present application realizes the data management of each collection node in the target area, solves the problem of data loss caused by communication interruption, and improves the adaptability and robustness of the system. This scheme is not only suitable for the energy management system of the intelligent industrial park, but also can be popularized to smart cities, industrial automation and other fields, and has wide application prospect.

[0078] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for automatic reconstruction of multi-point packet loss data for intelligent fusion terminals, characterized in that: The method comprises the following steps: S1: Real-time data stream monitoring is performed on a plurality of collection nodes in a target area; S2: Synchronous acquisition of adjacent device data of the collection nodes in the target area is performed; S3: Basic environment parameters of the collection nodes in the target area are collected; S4: The data collected in S1 to S3 are extracted, and reference data features of the nodes are obtained through processing; S5: A data compensation and interpolation repair module is constructed; S6: A predicted compensation value in a future time window is obtained according to the constructed data compensation and interpolation repair module; S7: A dynamic data reconstruction model is established based on the predicted compensation value in the future time window and the reference data features of the nodes; S8: An optimized data sequence is generated through the dynamic data reconstruction model; S9: Data updating and synchronization are performed on the collection nodes in the target area through the optimized data sequence; In S7, the establishment of the dynamic data reconstruction model specifically comprises the following steps: S7.1: The predicted compensation value and the real-time data stream monitoring results, the adjacent device data and the basic environment parameters in the target area are received and substituted into the dynamic data reconstruction model; S7.2: Output data sequences in at least five time windows are extracted as references; S7.3: The extracted output data sequences are substituted into the dynamic data reconstruction model; S7.4: Demand data features are confirmed and substituted into the dynamic data reconstruction model for dynamic learning; S7.5: Data analysis and pattern recognition are performed through the dynamic data reconstruction model to generate an optimized data sequence; S7.6: The generated optimized data sequence is evaluated for credibility.

2. The method of claim 1, wherein the method is a method for reconstructing multi-point packet loss data for an intelligent converged terminal. In S1, the real-time data stream monitoring comprises sampling frequency parameters, sampling period parameters and sampling data integrity indicators of the collection nodes in the target area, wherein the sampling frequency parameters define the number of collected data per unit time, the sampling period parameters define the time interval of each collection, and the sampling data integrity indicators evaluate the missing proportion of the current collected data.

3. The method of claim 1, wherein the method is characterized by: In S2, the adjacent devices at least include first-class devices adjacent in physical position, second-class devices similar in function and third-class devices similar in network topology structure, the data parameters of the first-class devices include the mean and variance of historical collected data, the data parameters of the second-class devices include functional response time and data fluctuation range, and the data parameters of the third-class devices include network delay and data transmission rate.

4. The method of claim 1, wherein the method is characterized by: In S3, the basic environment parameters at least include temperature distribution data, humidity distribution data and electromagnetic interference intensity data, wherein the temperature distribution data reflect the thermal environment change trend in the target area, the humidity distribution data analyze the possible impact on device operation, and the electromagnetic interference intensity data evaluate the stability of the communication link.

5. The method of claim 1, wherein the method is a method of automatically reconstructing multi-point packet loss data for a smart converged terminal. In S4, the processing to obtain the reference data features of the nodes specifically comprises: S4.1: Real-time data stream monitoring results of the collection nodes in the target area are extracted; S4.2: Collect the sampling data integrity indicators of each collection node in the target area, and add an adaptive floating threshold to obtain a data integrity feature; S4.3: Extract the adjacent device data of each collection node in the target area; S4.4: Collect the historical collection data mean and variance of the first type of device, and add an adaptive floating threshold to obtain an adjacent device data feature; S4.5: Collect the functional response time and data fluctuation range of the second type of device, and add an adaptive floating threshold to obtain a functional correlation feature; S4.6: Extract the basic environmental parameters of each collection node in the target area; S4.7: Set several time window thresholds, and extract the temperature distribution data, humidity distribution data, and electromagnetic interference intensity data according to the set time window thresholds; S4.8: Collect the environmental parameters data in at least three time window thresholds to obtain the environmental features in each time window.

6. The method of claim 1, wherein the method is a method of automatically reconstructing multi-point packet loss data for a smart converged terminal. In the S5, the data compensation and interpolation repair module includes a data compensation unit, an interpolation repair unit, and a verification unit. The data compensation unit and the interpolation repair unit are connected through a high-speed communication interface, and the interpolation repair unit and the verification unit are connected through a low-delay data channel.

7. The method of claim 6, wherein the method is a method for automatic reconstruction of multi-point packet loss data for a smart converged terminal. The data compensation unit is used to collect the adjacent device data change trend, the functional correlation change trend, and the environmental feature change trend in the future time window. The interpolation repair unit is used to analyze the data missing condition in the target area according to the change trend collected by the data compensation unit, and generate a predicted compensation value. The verification unit is used to verify the predicted compensation value generated by the interpolation repair unit. The verification condition is that the change trend data collected by the data compensation unit is used as a reference, and the historical data change trend collected in the past is used as a reference. The verification formula is Pmax·C ≤ Pmax ≤ Pmax·D, where Pmax represents the predicted compensation value, C is the minimum lower threshold, and D is the maximum upper threshold.

8. The method of claim 1, wherein the method is a method of automatic reconstruction of multi-point packet loss data for a smart converged terminal. In the optimized data sequence credibility evaluation, the upper layer credibility Tw and the lower layer credibility Tg are combined for analysis. The upper layer credibility Tw calculation formula is Tw = NUMt×M + (RSSI1+ RSSI 2 +......)×N, where NUMt represents the data integrity feature in the target area, RSSI represents the functional correlation feature of several adjacent devices, M and N are both specified weight values and the sum is 1. The lower layer credibility Tg calculation formula is Tg = CN1×K + Pmax×L, where CN1 represents the basic environmental parameter, Pmax represents the predicted compensation value, K and L are both specified weight values and the sum is 1. If the upper layer credibility Tw and the lower layer credibility Tg verification formula are both established, the final credibility Ep is determined, and the calculation formula is Ep = Tg / (Tg + Tw).

9. The method of claim 1, wherein the method is a method of automatic reconstruction of multi-point packet loss data for a smart converged terminal. In the S9, the optimized data sequence is generated by the dynamic data reconstruction model, so that the data integrity of each acquisition node in the target region is updated according to the generated optimized data sequence, and the functional correlation of the adjacent device is adjusted according to the optimized data sequence.

Images