Transparent transmission and self-adaptive packaging method, system and equipment for sensing data of icing monitoring terminal and medium

Abstract

The invention discloses a transparent transmission and adaptive packaging method, system and device for sensing data of an icing monitoring terminal and a medium, and belongs to the technical field of power system data transmission, and the method comprises the steps: obtaining the sensing data of the icing monitoring terminal; constructing a data feature vector containing a data type, a data scale, a volatility index and a priority index; according to the data feature vector, the network state parameter and the terminal energy state, a packaging strategy is determined through a decision function; packaging the sensing data according to the packaging strategy to generate a data packet containing a metadata head and a data load; and calculating a scheduling weight of the data packet based on the data feature vector, and determining a transmission sequence for sending according to the scheduling weight and a priority queue mechanism. According to the method, the problem of long-term retention of low-change-rate data is solved by quantifying data feature differences and introducing time enhancement factors, network condition changes are adapted by dynamically selecting a packaging strategy, and transmission resource allocation is optimized by a weighted scheduling mechanism.

Description

Technical Field

[0001] This invention relates to the field of power system data transmission technology, specifically to a transparent transmission and adaptive encapsulation method, system, device, and medium for icing monitoring terminal sensor data. Background Technology

[0002] The transmission line icing monitoring system needs to collect data from various sensors, such as tension, tilt angle, micro-meteorology, and images, in real time, and transmit the data to the monitoring center via wireless network.

[0003] In existing technologies, sensor data typically employs a uniform encapsulation and transmission method, failing to consider the differences in characteristics between different data types and the dynamic changes in network conditions. Existing solutions suffer from the following problems: First, using the same transmission strategy for all data results in high-value data not being prioritized for transmission, while low-change-rate data consumes excessive bandwidth; second, the fixed data encapsulation method cannot adapt to fluctuations in network conditions, leading to the transmission of large amounts of redundant data during network congestion, causing delays or even loss of critical data transmission; third, the lack of quantitative assessment of data timeliness means that data not transmitted for extended periods is easily overlooked, affecting the monitoring system's complete recording of the icing state evolution process. Therefore, A technical solution is needed that can adaptively adjust encapsulation and transmission strategies based on data characteristics and network conditions. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention provides a method, system, device and medium for transparent transmission and adaptive encapsulation of sensor data of icing monitoring terminal.

[0005] Therefore, the technical problems solved by this invention are: the failure to prioritize the transmission of high-value data and the excessive bandwidth consumption of low-change-rate data due to the use of a unified transmission strategy for all sensor data in the prior art; the inability of fixed encapsulation methods to adapt to network condition fluctuations and the transmission of a large amount of redundant data during network congestion, resulting in delays or even loss of critical data transmission; and the lack of quantitative assessment of data timeliness, which leads to the neglect of data that has not been transmitted for a long time, affecting the monitoring system's complete recording of the icing state evolution process.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for transparent transmission and adaptive encapsulation of sensor data for icing monitoring terminals, comprising, Acquire sensor data from the icing monitoring terminal; Construct a data feature vector corresponding to the sensor data, wherein the data feature vector includes data type, data size, volatility index and priority index; Based on the data feature vector, network state parameters, and terminal energy state, the encapsulation strategy corresponding to the sensing data is determined through a decision function. The sensor data is encapsulated according to the encapsulation strategy to generate a data packet containing metadata header and data payload; The scheduling weight of the data packet is calculated based on the data feature vector, and the transmission order of the data packet is determined and sent according to the scheduling weight and the preset priority queue mechanism.

[0007] As a preferred embodiment of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal described in this invention, wherein: the construction of the data feature vector corresponding to the sensor data includes: The sensor data is type-identified to determine the data type; The original size of the sensor data is calculated to determine the data scale. The relative change between the current sampled value and the historical sampled value is obtained, and the volatility index is determined by combining the time interval since the last transmission. Determine whether the sensor data triggers a preset alarm condition, and determine the priority indicator by combining the volatility indicator and the similarity of historical data.

[0008] As a preferred embodiment of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal described in this invention, wherein determining the volatility index includes: The difference between the current sampled value and the historical sampled value is calculated and normalized based on the sensor's measurement range to obtain the relative change. Calculate the time enhancement factor based on the time interval and the preset time decay coefficient; The volatility index is obtained by combining the relative change with the time enhancement factor.

[0009] The beneficial effects of this preferred technical solution are as follows: After introducing the time enhancement factor, data that has not been transmitted for a long time receives a higher volatility score during scheduling, solving the problem of low-change-rate data being stuck for a long time due to low priority. In network fluctuation or congestion scenarios, this mechanism balances the contradiction between data freshness and transmission priority, ensuring that monitoring data with small changes but important timeliness is not continuously ignored, thereby fully capturing the evolution process of the icing state and avoiding the omission of key information.

[0010] As a preferred embodiment of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal described in this invention, wherein: determining the encapsulation strategy corresponding to the sensor data through a decision function includes: The transmission quality parameters of the current network are collected as the network status parameters; The current energy balance of the terminal is obtained as the energy state of the terminal; The data feature vector, network state parameters, and terminal energy state are input into a preset decision matrix; The encapsulation strategy is determined based on the decision results output by the decision matrix.

[0011] As a preferred embodiment of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal described in this invention, wherein determining the priority index includes: Determine the alarm indication factor corresponding to the sensor data; Calculate the redundancy estimate of the sensor data based on historical transmission data; The priority index is obtained by weighting the alarm indicator factor, the volatility index, and the redundancy estimate with their respective weighting coefficients.

[0012] As a preferred embodiment of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal according to the present invention, wherein: the encapsulation processing of the sensor data according to the encapsulation strategy includes: The dynamic compression ratio is determined based on the network status parameters. Select a compression algorithm that matches the data type; The sensor data is compressed according to the dynamic compression ratio and the compression algorithm to generate compressed data. A compression algorithm identifier is added to the compressed data to form the data payload.

[0013] As a preferred embodiment of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal described in this invention, wherein: determining the dynamic compression rate based on the network state parameters includes: The network state parameters are normalized to obtain network quality indicators; The dynamic compression ratio is calculated based on the preset base compression ratio, the preset adjustment coefficient, and the network quality index.

[0014] The beneficial effects of this preferred technical solution are as follows: It dynamically adjusts the compression ratio based on network quality indicators, automatically increasing the compression ratio to reduce the amount of transmitted data and lower the risk of packet loss when network conditions deteriorate; and decreasing the compression ratio to retain more data details when network conditions are good. This adaptive mechanism overcomes the problem of transmission failure or data quality loss that can easily occur with a fixed compression ratio in different network environments, adapts to the dynamic changes in network conditions in complex field environments, and enhances transmission reliability.

[0015] This invention provides a transparent transmission and adaptive encapsulation system for sensor data of an icing monitoring terminal.

[0016] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a transparent transmission and adaptive encapsulation system for sensor data of an icing monitoring terminal, comprising: The data acquisition module is used to acquire sensor data from the icing monitoring terminal; The feature analysis module is used to construct a data feature vector corresponding to the sensing data. The data feature vector includes data type, data size, volatility index and priority index. The encapsulation decision module is used to determine the encapsulation strategy corresponding to the sensing data through a decision function based on the data feature vector, network state parameters and terminal energy state, and to encapsulate the sensing data according to the encapsulation strategy to generate a data packet containing metadata header and data payload. The transmission scheduling module is used to calculate the scheduling weight of the data packet based on the data feature vector, and to determine the transmission order of the data packet according to the scheduling weight and a preset priority queue mechanism for sending.

[0017] The present invention provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal.

[0018] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal.

[0019] The beneficial effects of this invention are as follows: By constructing a data feature vector that includes data type, scale, volatility index, and priority index, the characteristic differences of different sensor data are quantified. The volatility quantification method, which introduces a time-enhancing factor, solves the problem of long-term retention of low-rate-of-change data and fully records the evolution process of icing conditions. Based on a multi-dimensional decision-making mechanism of data feature vectors, network status, and terminal energy, a dynamic encapsulation strategy is selected to reduce data volume when network conditions deteriorate and retain data details when the network is good. The improved weighted scheduling algorithm allocates transmission weights according to data priority and volatility, giving high-value data priority transmission opportunities while ensuring basic bandwidth for low-priority data. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a general flowchart of a method for transparent transmission and adaptive encapsulation of sensor data for an icing monitoring terminal, provided as an embodiment of the present invention. Detailed Implementation

[0022] To make the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0023] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a method for transparent transmission and adaptive encapsulation of sensor data from an icing monitoring terminal, comprising: Step 1: Acquire sensor data from the icing monitoring terminal; Step 2: Construct the data feature vector corresponding to the sensor data. The data feature vector includes data type, data size, volatility index, and priority index. Step 3: Based on the data feature vector, network state parameters, and terminal energy state, determine the encapsulation strategy corresponding to the sensing data through a decision function; Step 4: Encapsulate the sensor data according to the encapsulation strategy to generate a data packet containing metadata header and data payload; Step 5: Calculate the scheduling weight of the data packet based on the data feature vector, and determine the transmission order of the data packet according to the scheduling weight and the preset priority queue mechanism for sending.

[0024] Existing icing monitoring systems face several technical challenges in data transmission: First, data from various sensors, such as tension sensors, tilt sensors, micro-meteorological sensors, and image sensors, differ significantly in type, scale, and frequency of change, making it impossible to address them all with a uniform transmission strategy. Second, wireless network quality fluctuates frequently in field monitoring environments, with drastic changes in signal strength and bandwidth availability. Fixed data encapsulation methods, while transmitting complete data during network congestion, lead to delays in critical information transmission. Third, although sensor sampling intervals are fixed, data change rates are uneven. Slow-changing data remains untransmitted for extended periods but is ignored by the scheduling system due to a lack of timeliness assessment mechanisms, resulting in incomplete monitoring records. Finally, different data have fundamentally different importance; alarm data requires immediate transmission while routine data can be processed with a delay. However, existing solutions lack multi-dimensional priority quantification methods, leading to unreasonable allocation of transmission resources.

[0025] This embodiment establishes a data feature vector quantization mechanism to transform the type, scale, volatility, and priority of sensor data into calculable feature parameters, providing a decision-making basis for subsequent adaptive processing. By introducing a volatility calculation method using a time enhancement factor, the timeliness assessment problem of low-rate-of-change data is addressed. A multi-dimensional decision matrix dynamically selects encapsulation strategies based on data characteristics, network status, and terminal energy. Compression or sampling encapsulation is used to reduce data volume under poor network conditions, while complete encapsulation is used to preserve data details under good network conditions. A weighted scheduling algorithm assigns transmission weights to each data packet, combined with a multi-level priority queue mechanism. High-value data receives priority transmission opportunities while low-priority data still retains basic transmission bandwidth, thereby optimizing transmission efficiency and resource allocation.

[0026] Example 2, an embodiment of the present invention, provides a transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal, based on the previous embodiment, comprising: Step 2: Constructing the data feature vector corresponding to the sensing data includes the following steps A1-A4: A1: Perform type identification on the sensor data to determine the data type; A2: Calculate the original size of the sensor data to determine the data scale; A3: Obtain the relative change between the current sampled value and the historical sampled value, and determine the volatility index by combining the time interval since the last transmission; A4: Determine whether the sensor data triggers a preset alarm condition, and determine the priority indicator by combining the volatility index and the similarity of historical data.

[0027] In this embodiment, in step 2, the data feature vector is obtained by: parsing the frame header identifier field of the sensor data to determine the data type; counting the number of bytes of the original data to determine the data size; calculating the difference between the current sampled value and the previous sampled value and dividing it by the sensor range to normalize and obtain the relative change; multiplying the relative change by a time enhancement factor calculated based on the time interval from the last transmission to obtain a volatility index; determining whether the data exceeds a preset threshold to obtain an alarm factor; calculating the similarity with historical data to obtain a redundancy estimate; and weighting and summing the alarm factor, volatility index, and redundancy estimate with preset weight coefficients to obtain a priority index. The above four dimensions are combined to form the data feature vector.

[0028] In an optional implementation, in step 2, the data feature vector can be formed by: extracting the data type and counting the number of bytes as the scale; collecting the most recent ten sampling points and calculating their arithmetic mean as the mean feature; calculating the difference between the maximum and minimum values ​​among the ten sampling points as the range feature; and counting the percentage of points exceeding the mean among the ten sampling points as the skewness feature. The data type, scale, range feature, and skewness feature are combined to form the feature vector. This method reflects the data change characteristics through statistical features within a sliding window.

[0029] In another optional implementation, in step 2, the data feature vector can also be formed by: identifying the data type and statistical scale, using the edge detection operator to count the proportion of edge pixels to the total number of pixels as image features for image data, calculating the sum of the absolute values ​​of the first-order differences of three consecutive samples as change features for numerical data, assigning basic weight values ​​according to the sensor's position in the monitoring system, and combining the data type, scale, content features and position weights to form a feature vector.

[0030] In this embodiment of the application, in step A3, the volatility index is obtained by: calculating the difference between the current sampled value and the historical sampled value, and normalizing it based on the sensor range; Calculate the time enhancement factor based on the time interval and the preset time decay coefficient; The volatility index is obtained by combining the relative change with the time enhancement factor.

[0031] In an optional implementation, in step A3, the volatility index can be obtained by: performing sliding window statistical analysis on the sampled sequence, with the window length set to 8 sampling periods, calculating the root mean square deviation of the sampled sequence within the window, specifically by calculating the sum of squares of the deviations of each sampled value within the window from the mean of the sequence, dividing the sum of squares by the window length and taking the square root to obtain the standard deviation, and normalizing the standard deviation according to the sensor range as the volatility index.

[0032] In another optional implementation, in step A3, the volatility index can also be obtained by: performing a first-order difference operation on the sampled sequence to obtain a difference sequence, calculating the energy characteristic value of the difference sequence, specifically by performing a square operation on each difference value in the difference sequence and summing the results to obtain the total energy, normalizing the total energy according to the length of the difference sequence and taking the square root, and then normalizing it according to the sensor range to obtain the volatility index.

[0033] In this embodiment of the application, in step A4, the priority indicator is determined by: determining the alarm indication factor corresponding to the sensing data; Calculate the redundancy estimate of the sensor data based on historical transmission data; The priority index is obtained by weighting the alarm indicator factor, the volatility index, and the redundancy estimate with their respective weight coefficients.

[0034] In an optional implementation, in step A4, the priority index can be obtained by: calculating the Shannon entropy of the data sequence based on information theory methods as a measure of information value; performing symbolization processing on the data sequence and counting the frequency of each symbol; calculating the negative logarithm weighted sum according to the definition of Shannon entropy to obtain the entropy value; normalizing the entropy value to a unit interval as the information value component; combining the data timeliness decay function to calculate the timeliness component, which decays linearly with the time span from the collection time to the current time; and linearly weighting and fusing the information value component and the timeliness component to obtain the priority index.

[0035] In another optional implementation, in step A4, the priority index can also be obtained by: establishing a sensor importance classification system, assigning differentiated basic weights based on the sensor's functional positioning in the monitoring system, assigning higher basic weights to structural safety sensors, including tension sensors and tilt sensors, and assigning lower basic weights to environmental monitoring sensors, including temperature and humidity sensors, superimposing a data activity correction term on the basic weights (the activity correction term is obtained by calculating the normalized deviation of the current sampled value from the recent sampled mean), and then superimposing a transmission frequency correction term (the frequency correction term is obtained by quantifying the deviation between the actual transmission frequency and the preset expected frequency), and then linearly weighting the basic weights and the two correction terms to obtain the priority index.

[0036] Step 3: Based on the data feature vector, network state parameters, and terminal energy state, the encapsulation strategy corresponding to the sensing data is determined through a decision function, including the following steps B1-B4: B1: Collect the current network transmission quality parameters as the network status parameters; B2: Obtain the current energy balance of the terminal as the energy status of the terminal; B3: Input the data feature vector, network state parameters, and terminal energy state into a preset decision matrix; B4: Determine the encapsulation strategy based on the decision results output by the decision matrix.

[0037] In this embodiment, in step 3, the decision function selects the encapsulation strategy by establishing a multi-dimensional decision matrix. The decision matrix contains four parallel judgment rules: Rule 1 determines whether the volatility index is lower than a preset volatility threshold and whether the priority index is low priority. When the conditions are met, the difference encapsulation strategy is output, which only transmits the change between the current value and the previous transmitted value. Rule 2 determines whether the data size exceeds a preset size threshold. When the conditions are met, the compression encapsulation strategy is output, and the compression rate is dynamically determined based on the network state parameters. Rule 3 determines whether the network state parameters are lower than a preset network threshold and whether the priority index is low priority. When the conditions are met, the sampling encapsulation strategy is output, which reduces the transmission frequency of this type of data. Rule 4 determines whether the priority index is high priority. When the conditions are met, the complete encapsulation strategy is output, and the highest network transmission priority is assigned. The system takes the data feature vector, network state parameters, and terminal energy state as input to the decision matrix, matches the four rules in sequence, and outputs the encapsulation strategy corresponding to the first rule that meets the conditions.

[0038] In an optional implementation, in step 3, the decision function can achieve strategy selection by: constructing a fuzzy logic-based decision reasoning system; performing fuzzification processing on the input parameters; dividing continuous parameters such as volatility indicators, data scale, and network quality parameters into three fuzzy levels (low, medium, and high) according to a preset boundary threshold; establishing a fuzzy rule base to store encapsulation strategies corresponding to different combinations of fuzzy levels; associating each rule with a confidence weight to characterize the credibility of the rule; matching the fuzzy level of the current input parameter with all rules in the rule base; calculating the activation strength of each rule, i.e., the product of the membership degree of the input parameter and the confidence degree of the rule; and selecting the encapsulation strategy corresponding to the rule with the highest activation strength as the decision output.

[0039] In another optional implementation, in step 3, the decision function can also achieve strategy selection by constructing a hierarchical decision tree structure. The decision tree adopts a multi-level node segmentation method. The root node performs binary judgment on the priority index. When the judgment result is high priority, it directly outputs the complete encapsulation strategy as the leaf node result. When the judgment result is not high priority, it enters the second-level node. The second-level node performs binary judgment on the data size. When the judgment result is large-scale, it outputs the compression encapsulation strategy. When the judgment result is not large-scale, it enters the third-level node. The third-level node performs binary judgment on the network state parameters. When the judgment result is poor network quality, it outputs the sampling encapsulation strategy. When the judgment result is not poor network quality, it outputs the difference encapsulation strategy. The system judges the input parameters layer by layer along the decision tree path until it reaches the leaf node to obtain the encapsulation strategy output.

[0040] Step 4: The sensor data is encapsulated according to the encapsulation strategy to generate a data packet containing metadata header and data payload, including the following steps C1-C4: C1: Determine the dynamic compression ratio based on the network state parameters; C2: Select a compression algorithm that matches the data type; C3: Compress the sensor data according to the dynamic compression ratio and the compression algorithm to generate compressed data; C4: Add a compression algorithm identifier to the compressed data to form the data payload.

[0041] Furthermore, determining the dynamic compression ratio based on the network state parameters includes: The network state parameters are normalized to obtain network quality indicators; The dynamic compression ratio is calculated based on the preset base compression ratio, the preset adjustment coefficient, and the network quality index.

[0042] Furthermore, in the specific implementation of step A3, the volatility index The calculation employs a method that integrates relative change and time weighting. The current sample value is obtained. and the previous sample value Calculate the absolute value of the difference between the two. Extract the upper range limit from the sensor calibration parameters. and lower limit of range The normalized relative change is obtained by dividing the absolute value of the difference by the measurement range, thus eliminating the influence of different sensor dimensions. The time interval is obtained by taking the difference between the current time and the last time the sensor data was successfully transmitted. Set the time decay coefficient Calculate the time enhancement factor Volatility Indicators The calculation formula is: Time decay coefficient The value of is determined through historical transmission data analysis and fitting. Before system deployment, data change sequences under typical environments are collected. The impact of transmission delay on monitoring accuracy is used as the optimization objective, and the result is obtained using gradient descent fitting. .

[0043] In the specific implementation of step A4, the priority indicator is calculated using a multi-factor decision model. An alarm indicator factor is constructed. Determine if the current sensor data exceeds the preset alarm interval value; if it exceeds the threshold... ,otherwise The redundancy estimate is obtained by calculating the similarity with historical data. This value is calculated based on the similarity of historical data. The formula for calculating the priority indicator is: Weighting coefficient satisfy The weighted fitting process is as follows: collect historical transmission records, label the actual importance of each data point, and use logistic regression or support vector machine (SVM) for fitting, with the importance label as the dependent variable. As the independent variable, the result obtained after training This model comprehensively considers three dimensions: security alerts, data volatility, and information redundancy.

[0044] In the specific implementation of steps B3 and B4, the decision matrix contains four parallel decision rules to implement the encapsulation strategy selection. The decision function is... ,in For data feature vectors, Network status metrics include signal strength and bandwidth availability. This refers to the terminal energy state.

[0045] Rule 1 applies to the encapsulation of differences, when and Priority At that time, the output difference encapsulation strategy only transmits the change between the current value and the previous transmitted value.

[0046] Rule 2 applies to compressed packaging, when Size At that time, output compression and encapsulation strategies, with the compression ratio based on the current network status. Dynamic adjustment.

[0047] Rule 3 applies to sampling encapsulation, when and Priority When the frequency is low, the output sampling and encapsulation strategy actively reduces the transmission frequency of this type of data to the reference frequency. .

[0048] Rule 4 applies to complete encapsulation, when Priority During the High phase, a full encapsulation strategy is output, using full encapsulation and assigning the highest network transmission priority regardless of volatility, size, or network status.

[0049] The system takes the data feature vector, network state parameters, and terminal energy state as input to the decision matrix, matches four rules in sequence, and outputs the encapsulation strategy corresponding to the first rule that meets the conditions.

[0050] In the specific implementation of step C1, the dynamic compression ratio The determination adopts a network adaptive adjustment mechanism. Network quality indicators are obtained by normalizing the network state parameters. The normalized value range is Normalized maximum value The base compression ratio is typically 1. Query the base compression ratio based on the data type. Set the value to 0.5 for image data and 0.3 for numerical data. (Set the adjustment factor.) The impact of network quality on compression ratio is controlled. The dynamic compression ratio calculation formula is: Adjustment coefficient The method for determining this is as follows: Simulate different network states in a laboratory environment. Up to version 1.0, with the dual objectives of compressed data restoration accuracy and transmission time, Pareto front analysis was used to determine... This mechanism allows the system to automatically increase the compression ratio to reduce data volume when the network is poor, and appropriately reduce the compression ratio to retain more data details when the network is good.

[0051] In the specific implementation of steps C3 and C4, the sensor data is compressed according to the dynamic compression ratio and compression algorithm to generate compressed data. A compression algorithm identifier is added to the compressed data to form a data payload. The data payload exhibits different structures depending on the encapsulation strategy; when using differential encapsulation, the payload contains a data identifier and a change value; when using compressed encapsulation, the payload contains a compression algorithm identifier and compressed data blocks.

[0052] The data packet uses a TLV structure, containing a fixed-length metadata header and a variable-length data payload. The metadata header contains four core data types: data acquisition timestamp accurate to milliseconds, unique identifier for the terminal device, sensor type identifier, and data quality confidence level. ; Data quality confidence The calculation formula is: in Indicates the first The magnitude of the various error sources, including inherent accuracy errors of the sensor, data acquisition timing jitter errors, and errors caused by environmental interference, is obtained through equipment calibration and historical statistics. This is the proportion of such errors in the overall quality assessment. The determination was made using the Analytic Hierarchy Process (AHP): multiple domain experts were invited to conduct pairwise comparisons of the importance of various error sources, construct a judgment matrix, and calculate the consistency ratio. Finally, a weight vector is obtained, such as a sensor accuracy error weight of 0.4 and a timing error weight of 0.3.

[0053] In the specific implementation of step 5, the packet scheduling weight The weight of each data packet is calculated using a modified Weighted Fair Queuing (WFQ) scheduling algorithm. The calculation formula is: in The base weight is set to 1.0 as a benchmark, and Priority is the data priority. It's data volatility. as well as This refers to the adjustment coefficient. The fitting method is as follows: Simulate multi-priority data flows in a simulation environment, with the optimization objective being the fairness of the average latency of high-priority data and the throughput of low-priority data. A grid search method is then used to... Optimize within the range and finally determine This weighting calculation method allows high-priority, high-rate-of-change data packets to receive greater scheduling weights and higher priority transmission opportunities. At the same time, the basic weights and formula structure ensure that low-priority data can still obtain basic transmission bandwidth when the system load is high.

[0054] Data packets are allocated to a multi-level priority queue mechanism based on priority, dividing them into four levels: urgent, high, medium, and low. Different quality of service parameters are assigned to each level, such as maximum allowable latency, minimum bandwidth guarantee, and acceptable packet loss rate. Within each queue, packets are sorted from highest to lowest scheduling weight to determine the transmission order. During transmission, data packets are preferentially selected from the high-priority queue for sending, while ensuring that the low-priority queue receives the minimum bandwidth guarantee.

[0055] This embodiment integrates network status awareness, continuously monitoring core network parameters such as link latency, jitter, and packet loss rate. When a network quality degradation is detected, it automatically triggers a change in transmission strategy: further reducing the transmission frequency of low-priority data, or enabling forward error correction (FEC) coding for high-priority core data to enhance the transmission performance of these core data under adverse channel conditions.

[0056] Furthermore, the dimensional analysis of the volatility index calculation formula is as follows: The data volatility calculation formula V=(|Vcurrent-Vprevious|) / (Vmax-Vmin)×(1+α·t) is divided into two parts. The first part (|Vcurrent-Vprevious|) / (Vmax-Vmin) calculates the relative rate of change. The numerator is the absolute value of the difference between the current sample value and the previous sample value, and its dimension is consistent with the physical quantity of the sensor. The denominator is the difference between the upper limit and the lower limit of the range, and its dimension is consistent with the physical quantity of the sensor. The numerator and denominator have the same dimension, and dividing them yields a dimensionless scalar. In the second part (1+α·t), t is the time interval in seconds (s). The dimension of the time decay coefficient α is the reciprocal of the second, i.e., s⁻¹. The dimension of α·t is s⁻¹ multiplied by s, which equals dimensionless. 1 plus a dimensionless quantity still equals dimensionless. Multiplying the dimensionless scalar in the first half by the dimensionless quantity in the second half results in a dimensionless final product, representing an enhanced relative rate of change. This formula eliminates the risk of dimensional confusion.

[0057] Furthermore, the physical meaning of the volatility index calculation formula is as follows: the data volatility formula integrates two dimensions, instantaneous change and time decay, to form an information value density or data novelty index. The first part, the relative rate of change, quantifies the degree of instantaneous change in the data. The greater the difference between the current sampled value and the previous sampled value, the greater the relative rate of change, indicating that the data contains more new information. The second part, the time enhancement factor, quantifies the time decay of the data. The longer the time interval t since the last transmission, the greater the enhancement factor. Even data with a small relative rate of change that has not been transmitted for a long time will obtain a higher volatility score due to the effect of the time enhancement factor, thus gaining a transmission opportunity. This design reflects the innovation of the present invention and solves the problem of incomplete monitoring records caused by the long-term retention of low-rate-of-change data in traditional schemes.

[0058] Furthermore, the dimensional analysis of the volatility index calculation formula is as follows: The data volatility calculation formula V=(|Vcurrent-Vprevious|) / (Vmax-Vmin)×(1+α·t) is divided into two parts. The first part (|Vcurrent-Vprevious|) / (Vmax-Vmin) calculates the relative rate of change. The numerator is the absolute value of the difference between the current sample value and the previous sample value, and its dimension is consistent with the physical quantity of the sensor. The denominator is the difference between the upper limit and the lower limit of the range, and its dimension is consistent with the physical quantity of the sensor. After dividing the numerator and denominator, which have the same dimensions, a dimensionless scalar is obtained. In the second part (1+α·t), t is the time interval in seconds (s), and the dimension of the time decay coefficient α is the reciprocal of the second (s). -1The dimensionless quantity of α·t is 1, and adding a dimensionless quantity to 1 also results in a dimensionless quantity. Multiplying the first part of the dimensionless scalar by the second part of the dimensionless quantity yields a dimensionless result, representing an enhanced relative rate of change. This formula eliminates the risk of dimensional confusion.

[0059] Furthermore, the physical meaning of the volatility index calculation formula is as follows: the data volatility formula integrates two dimensions, instantaneous change and time decay, to form an information value density or data novelty index. The first part, the relative rate of change, quantifies the degree of instantaneous change in the data. The greater the difference between the current sampled value and the previous sampled value, the greater the relative rate of change, indicating that the data contains more new information. The second part, the time enhancement factor, quantifies the time decay of the data. The longer the time interval t since the last transmission, the greater the enhancement factor. Even data with a small relative rate of change that has not been transmitted for a long time will obtain a higher volatility score due to the effect of the time enhancement factor, thus gaining a transmission opportunity. This design reflects the innovation of the present invention and solves the problem of incomplete monitoring records caused by the long-term retention of low-rate-of-change data in traditional schemes.

[0060] Before calculating the Priority index using a multi-factor decision model in the specific implementation of step A4, add the following: Furthermore, the physical meaning of the priority index calculation formula is derived from the fact that the data volatility index V and the data priority index Priority together form a multi-factor decision model. This model simulates the working logic of an intelligent scheduler. Alarm data has the highest processing priority to ensure real-time transmission of security alarm information. Data with significant changes contains more information and should receive a higher transmission priority. Duplicate data has lower utility, and its transmission priority is reduced through redundancy estimation. The weighted summation of the data priority formula Priority = β1·Ialarm + β2·V + β3·(1-Rredundancy) quantifies the above three decision rules. The alarm indicator factor Ialarm has the largest weight coefficient β1 (0.5), reflecting the absolute priority of alarm data. The volatility index V has the second largest weight coefficient β2 (0.3), reflecting the importance of data novelty. The inverse indicator of redundancy estimation (1-Rredundancy) has a relatively small weight coefficient β3 (0.2), reflecting the inhibitory effect of information redundancy on priority. The three weighting coefficients satisfy the normalization constraint β1+β2+β3=1, ensuring that the value of the priority index is between 0 and 1.

[0061] Furthermore, the innovation of the multi-factor decision-making model lies in the fact that it represents an intelligent data transmission decision-making model with adaptive capabilities, optimizability, and clear physical meaning. The model comprehensively considers the instantaneous value and timeliness of data changes, quantifying the information value density of data by combining the relative rate of change with a time decay factor through a volatility index V. It quantifies the data transmission priority by weighting and fusing three dimensions: alarm urgency, data novelty, and information redundancy through a priority index Priority. Using this model, the system shifts from rule-based transmission to value-based transmission, ensuring both the real-time nature of alarm data and maximizing the information transmission efficiency of limited channel capacity.

[0062] It should be noted that the design of the time enhancement factor for the volatility index V is the key innovation of this invention. In traditional schemes, data scheduling is based solely on the instantaneous changes in data. Data with small changes has low priority and may not receive a transmission opportunity for a long time, resulting in the monitoring system being unable to fully record the evolution of the icing state. This invention introduces a time enhancement factor (1+α·t), so that even data with small changes that has not been transmitted for a long time will receive a higher volatility score due to the increased time interval t, thus gaining a transmission opportunity during scheduling. This mechanism balances the contradiction between data freshness and transmission priority, ensuring that monitoring data with small changes but important timeliness are not continuously ignored in network fluctuation or congestion scenarios, and fully capturing the evolution of the icing state to avoid missing key information.

[0063] Furthermore, the physical meaning of the time decay coefficient α is as follows: the time decay coefficient α controls the intensity of the influence of time factors on volatility indicators. The larger the value of α, the faster the time enhancement factor (1+α·t) grows, the more significant the improvement in volatility scores after a long period of data not being transmitted, and the higher the system's emphasis on data timeliness. The smaller the value of α, the slower the time enhancement factor grows, and the system focuses more on the instantaneous rate of change of data. The dimension of α is s⁻¹ to ensure that α·t is a dimensionless quantity, making the time enhancement factor a purely numerical gain coefficient. The value of α is determined through historical transmission data analysis and fitting. Taking the impact of transmission delay on monitoring accuracy as the optimization objective, the data change sequence collected under typical conditions is fitted using the gradient descent method to obtain α equal to 0.1. This value achieves an optimal balance between ensuring data freshness and transmission efficiency.

[0064] Furthermore, the optimization method for the weight coefficients involves the following: the weight coefficients β1, β2, and β3 in the priority index calculation formula are not only typical values, but are obtained through training with historical data or optimization using an expert system. The weight coefficient determination steps involve collecting historical transmission records and labeling each data point with its actual importance tag. These tags are assigned by domain experts based on the data's impact on icing monitoring decisions: critical alarm data is labeled as high importance, routine monitoring data as medium importance, and redundant data as low importance. A fitting process is performed, using the importance tag as the dependent variable, the alarm indicator factor Ialarm, the volatility index V, and the redundancy estimate Rredundancy as independent variables, resulting in β1 equal to 0.5, β2 equal to 0.3, and β3 equal to 0.2. This model comprehensively considers three dimensions: safety alarms, data volatility, and information redundancy. During system operation, the weight coefficients are adjusted online based on new transmission records and importance tags to adapt to changes in the network environment and monitoring needs.

[0065] Furthermore, the application effect of the multi-factor decision model is demonstrated by the following: Using this model, the system can intelligently identify the transmission value of different data. For alarm data, the alarm indicator factor Ialarm is 1, the priority index is at least 0.5, and other factors are greater than or equal to 0.5, obtaining the highest transmission priority and ensuring the real-time nature of security alarms. For highly volatile data, a larger volatility index V corresponds to a higher priority index, resulting in timely transmission of data containing new information. For data with low change rates but long periods without transmission, a time enhancement factor increases the volatility index V, and the priority index increases accordingly, preventing such data from remaining indefinitely and ensuring the integrity of monitoring records. For data with high redundancy, a larger redundancy estimate Rredundancy and a smaller reverse index (1-Rredundancy) reduce the priority index, decreasing redundant data transmission and saving bandwidth resources. This model achieves a shift from rule-based transmission to value-based transmission, maximizing the information transmission efficiency of limited channel capacity.

[0066] In an optional implementation, the weighting coefficients can be determined by: collecting N historical transmission data samples, each sample containing three feature values: an alarm indicator Ialarm, a volatility index V, and a redundancy estimate Rredundancy, along with corresponding actual importance labels. The actual importance labels are converted to numerical values: high importance is labeled 1, medium importance 0.5, and low importance 0. A linear regression model is constructed, where the predicted value of the priority index is equal to β1 multiplied by Ialarm, β2 multiplied by V, and β3 multiplied by (1-Rredundancy). The weighting coefficients are solved using the least squares method, calculating the sum of squares of the differences between the predicted values ​​and actual label values ​​of the N samples as the loss function. Partial derivatives of the loss function with respect to β1, β2, and β3 are taken and set to zero, resulting in a system of three linear equations. The initial values ​​of the weighting coefficients are obtained by solving the system of equations. The weighting coefficients are then normalized so that β1 plus β2 plus β3 equals 1. The normalization method is to divide each weighting coefficient by the sum of the three weighting coefficients. The model performance was tested on an independent validation set, and the mean squared error (MSE) and correlation coefficient (R²) between the predicted importance labels and the actual importance labels were calculated. If the MSE was greater than a preset threshold or the R² was less than a preset threshold, the model was retrained after adjusting the sample data or performing feature engineering. The final values ​​were β1 = 0.5, β2 = 0.3, and β3 = 0.2.

[0067] In another optional implementation, the weight coefficients can be determined by: collecting historical transmission data samples and labeling them with actual importance tags; binarizing the importance tags, labeling high and medium importance as 1 to indicate priority transmission, and labeling low importance as 0 to indicate delayed transmission; constructing a logistic regression classification model, using the alarm indicator Ialarm, volatility index V, and (1-Rredundancy) as input features; and outputting the logistic regression model as the probability P that data needs priority transmission, calculated as P equals 1 divided by (1 plus e to the power of negative z), where z equals β1 multiplied by Ialarm plus β2 multiplied by V plus β3 multiplied by (1-Rredundancy). The model parameters are trained using gradient descent, initializing the weight coefficients β1, β2, and β3 to random values, calculating the cross-entropy loss between the predicted probability P and the actual label for each sample, calculating the gradient of the loss function with respect to β1, β2, and β3 respectively, updating the weight coefficients according to the gradient descent direction, iterating until the loss function converges. The trained weight coefficients are then normalized so that β1 plus β2 plus β3 equals 1. The classification accuracy, precision, and recall were tested on the validation set, and the weight coefficients were fine-tuned based on the test results. The final values ​​were β1 = 0.5, β2 = 0.3, and β3 = 0.2.

[0068] Example 3 is an embodiment of the present invention, which provides a transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.

[0069] This embodiment uses an icing monitoring terminal on tower #47 of a 220kV transmission line in a high-altitude area. The terminal is equipped with various sensors, including a tension sensor with a range of 0 to 50kN, a dual-axis tilt sensor, a micro-weather station to measure wind speed, temperature, and humidity parameters, and a 1080P high-definition camera to collect image data. The terminal connects to a monitoring center 50 kilometers away via a 4G network, where the local network signal is unstable, and the terminal is powered by solar energy, resulting in limited power supply.

[0070] Traditional transmission methods suffer from three problems. First, the data formats of various sensors are inconsistent, requiring the monitoring center to maintain multiple parsing interfaces. Second, large-capacity data transmission, such as image data, is prone to failure when the network signal is poor. Third, periodic full transmission generates a large amount of redundant data, wasting both bandwidth and power. This embodiment illustrates how the solution of this invention can solve these problems.

[0071] In the overall system architecture, data flows from bottom to top, passing through the sensor array, processing units, and finally being uploaded to the monitoring center via the communication module. The system comprises four core processing layers: data acquisition, feature recognition, adaptive encapsulation, metadata association, and communication scheduling, progressively completing intelligent data processing. The data acquisition module is responsible for acquiring raw data from heterogeneous sensors. The feature analysis module, the core computing layer, calculates data feature vectors, including data type, data size, volatility indicators, and priority indicators. The encapsulation decision module, the intelligent decision-making layer, selects the optimal encapsulation strategy based on the data feature vector and network state parameters. The metadata association unit, the data enhancement layer, adds source information and quality indices to data packets. The transmission scheduling module, the output control layer, manages the sending queue to ensure high-priority data is sent first. The monitoring center only interfaces with one standardized data format, simplifying system integration.

[0072] After system startup, parameters are initialized, and the feature analysis module loads sensor feature parameters. The tensile sensor's range is 0 to 50 kN, the tilt sensor's range is -30 to 30 degrees, the temperature range is -40 to 60℃, and the wind speed range is 0 to 60 m / s. The encapsulation decision module determines the decision thresholds: the volatility threshold Vlow is 0.01, the data size threshold Shigh is 100 KB, and the network quality threshold Nthreshold is 0.3. The weighting coefficients are configured as follows: volatility calculation time decay coefficient α is 0.1, priority decision weights β1, β2, and β3 are 0.5, β3 is 0.3, β4 is 0.2, compression ratio adjustment coefficient γ is 0.2, and scheduling weight coefficients δ and ε are 0.3 and ε, respectively. The metadata association unit pre-sets terminal information: device ID is T047, GPS coordinates are E118.25°N36.42°, and altitude is 1250 meters. The transmission scheduling module initializes four priority queues, corresponding to urgent, high, medium, and low priority data, respectively.

[0073] Under normal weather conditions, the system operates in normal monitoring mode. During data acquisition from 8:00 AM to 8:30 AM on a certain morning, the data acquisition module acquired sensor data. The tension sensor collected a value of 15.21 kN at 8:00:00, 15.21 kN at 8:00:30, and 15.22 kN at 8:01:00. The tilt sensor's X-axis tilt angle was 2.1 degrees, and its Y-axis tilt angle was 1.8 degrees. The micro-weather station's temperature parameters were -2℃, humidity parameters were 65%, and wind speed was 3.2 m / s. The network status parameter N of 0.8 indicates good network quality.

[0074] The feature analysis module performs feature analysis on the tension data, constructing a data feature vector for the tension value of 15.22 kN at 8:01:00. The difference between the current sampled value of 15.22 kN and the historical sampled value of 15.21 kN is calculated to be 0.01 kN. Normalization based on the sensor's range of 50 kN to 0 kN yields a relative change of 0.01 divided by 50, which equals 0.0002. The time interval since the last transmission is 60 seconds. Based on the time interval and a preset time attenuation coefficient α of 0.1, a time enhancement factor is calculated as 1 plus 0.1 multiplied by 60, which equals 7. Combining the relative change with the time enhancement factor, the volatility index V is calculated as 0.0002 multiplied by 7, which equals 0.0014.

[0075] To determine if the sensor data triggered the preset alarm conditions, the tensile force of 15.22 kN did not exceed the alarm threshold of 45 kN, and the alarm indicator factor Ialarm was 0. Based on historical transmission data, the estimated redundancy of the sensor data, Rredundancy, was calculated to be 0.9. The alarm indicator factor Ialarm, volatility index V, and redundancy estimate Rredundancy were weighted by their respective weighting coefficients to obtain the priority index: 0.5 x 0 + 0.3 x 0.0014 + 0.2 x (1) minus 0.9 equals 0.03042, which is low-priority data.

[0076] The data type is identified as floating-point (Float), and the original size of the statistical sensor data is 4 bytes. The constructed data feature vector is: data type Float, data size 4 bytes, volatility index 0.0014, and priority index 0.03042.

[0077] The encapsulation decision module collects the current network transmission quality parameters as network state parameters (N = 0.8). It obtains the terminal's current energy margin as the terminal energy state. The data feature vector, network state parameters, and terminal energy state are input into a preset decision matrix. The decision matrix contains four parallel decision rules. Rule 1 determines that if the volatility index V = 0.0014, which is less than the volatility threshold Vlow = 0.01, and the priority index is low priority, then the differential encapsulation strategy is output. Based on the decision results output by the decision matrix, the encapsulation strategy is determined to be differential encapsulation.

[0078] The sensor data is encapsulated using a differential encapsulation strategy, only encapsulating the change amount +0.01kN and timestamp information. The metadata association unit adds a metadata header to the encapsulated data. The metadata header includes the device ID T047, the timestamp 2024-01-15T08:01:00.000Z, the location coordinates E118.25°N36.42°, the data quality confidence score 0.98, and the encapsulation strategy identifier delta_encoding. A data packet containing the metadata header and data payload is generated.

[0079] The transmission scheduling module calculates the scheduling weight of data packets based on data feature vectors. Data packets are placed in a low-priority queue according to a priority index of 0.03042 and sent according to the regular scheduling cycle. During this stage, the amount of tension data transmitted is reduced from 4 bytes of complete data per transmission to 1 byte of variation, achieving a bandwidth saving rate of 75%.

[0080] That afternoon, weather conditions deteriorated, and the system detected an anomaly. At 14:30:00, the data acquisition module collected sensor data, showing a significant increase in tensile force to 25.63 kN, a continuous decrease in temperature to -5℃, a significant increase in humidity to 92%, and an increase in wind speed to 12.5 m / s.

[0081] The feature analysis module recalculates the data feature vector. The difference between the current sampled value of 25.63 kN and the historical sampled value of 18.45 kN is calculated to be 7.18 kN. Normalization based on the sensor's measurement range yields a relative change of 7.18 divided by 50, which equals 0.1436. The time interval since the last transmission is 30 seconds, and the time enhancement factor is calculated as 1 plus 0.1 multiplied by 30, which equals 4. Combining the relative change with the time enhancement factor, the volatility index V is calculated to be 0.1436 multiplied by 4, which equals 0.5744.

[0082] The system determines whether the sensor data triggers the preset alarm conditions. The tensile force value of 25.63 kN does not exceed the alarm threshold, so the alarm indicator factor is 0. The calculated redundancy estimate is 0.2. The priority index is 0.5 multiplied by 0 plus 0.3 multiplied by 0.5744 plus 0.2 multiplied by 1 (in parentheses) minus 0.2, which equals 0.33232, classifying it as medium-high priority.

[0083] Simultaneously, the module status monitoring unit detected a deterioration in network status, with the network status parameter N dropping to 0.4. The encapsulation decision module, based on the new data feature vector data type (Float), data size (4 bytes), volatility index (0.5744), priority index (0.33232), and network status parameter N of 0.4, determines the encapsulation strategy through a decision matrix. Rule four determines that if the priority index is deemed medium to high priority, the complete encapsulation strategy is output, ensuring the complete transmission of all core data.

[0084] As the icing situation worsened, the system entered an emergency state. At 16:45:00, the data acquisition module collected key data: the tension value was 42.56kN, close to the alarm value of 45kN; the X-axis tilt angle was 15.6 degrees; and the Y-axis tilt angle was 12.3 degrees, showing a significant shift. The image sensor captured a 450KB image of obvious icing.

[0085] The volatility index is calculated using the tensile data characteristic analysis module. The difference between the current sampled value of 42.56 kN and the historical sampled value of 35.21 kN is 7.35 kN. Normalization yields a relative change of 7.35 divided by 50, which equals 0.147. The time interval is 5 seconds, and the time enhancement factor is 1 plus 0.1 multiplied by 5, which equals 1.5. The volatility index V is 0.147 multiplied by 1.5, which equals 0.2205.

[0086] Since the value is close to the alarm threshold of 45kN, the system automatically increases its priority. The alarm indicator factor Ialarm is set to 1, and the priority index Priority is 0.5 multiplied by 1 plus 0.3 multiplied by 0.2205 plus 0.2 multiplied by parentheses 1 minus 0.1 equals 0.5 plus 0.06615 plus 0.18 equals 0.74615.

[0087] For image data with a statistical data size of 450KB, decision matrix rule two determines that the data size of 450KB is greater than the data size threshold Shigh of 100KB, and outputs a compression and encapsulation strategy. The network state parameter N is normalized to 0.2 to obtain the network quality index. Based on the preset base compression ratio λbase of 0.5, the preset adjustment coefficient γ of 0.2, and the network quality index, the dynamic compression ratio λ is calculated according to the formula λ=λbase+γ×(1-N / Nmax). λ is calculated as 0.5 + 0.2 multiplied by parentheses 1 minus 0.2 divided by 1 equals 0.5 + 0.16 equals 0.66.

[0088] The JPEG compression algorithm, which matches the image data type, is selected. The image data is compressed using a dynamic compression ratio of 0.66 and the JPEG compression algorithm, resulting in approximately 300KB of compressed data, saving 150KB of transmission space. The JPEG compression algorithm identifier is added to the compressed data to form the data payload.

[0089] The metadata association unit calculates the data quality confidence score Q. According to the formula Q=1-Σ(wi·ei), the inherent accuracy error of the sensor e1 is 0.05 with a weight of w1 of 0.1, the data acquisition timing jitter error e2 is 0.1 with a weight of w2 of 0.2, and the calculated Q is 1 minus 0.1 multiplied by 0.05 plus 0.2 multiplied by 0.1, which equals 1 minus 0.025, which equals 0.975, indicating that the data quality is relatively high.

[0090] The transmission scheduling module calculates the scheduling weight of data packets based on data feature vectors. For the volatility index V of the tension alarm data, which is 0.85, the scheduling weight W is calculated using the formula W=wbase×(1+δ·Priority)×(1+ε·V). This weight is calculated as 1.0 multiplied by (1 + 0.3)×0.746× (1 + 0.2)×0.85, which equals 1.0 multiplied by 1.2238 multiplied by 1.17, equaling 1.432. The highest transmission priority is assigned, and the data packet is placed in the emergency priority queue and immediately sent according to the scheduling weight and the preset priority queue mechanism.

[0091] When the network continues to deteriorate and the network state parameter N drops to 0.1, the system activates a deep power-saving mode. For low-priority meteorological data, a sampling and encapsulation strategy is used. Decision matrix rule three determines that the network state parameter is below the network quality threshold and the priority index is low, and outputs a sampling and encapsulation strategy to reduce the transmission frequency of this type of data from once per minute to once every 5 minutes. For medium-priority data, the compression ratio λ is increased to 0.8 to reduce the data volume.

[0092] The system achieves 45% bandwidth savings and 38% energy reduction within 24 hours through intelligent scheduling. Differential encapsulation saves 60% of numerical data traffic, adaptive compression reduces image data volume by 35%, and intelligent scheduling avoids 15% of data retransmissions.

[0093] After receiving the data, the monitoring center reconstructs it using metadata. For data with differential encapsulation, a complete data curve is reconstructed based on the baseline value and change sequence in the metadata. For compressed data, decompression is performed based on the compression algorithm identifier in the metadata. The monitoring center quickly filters high-quality data for accurate analysis using the Q-score, marking data with a Q-score below 0.9 as data requiring verification to make the analysis results more reliable.

[0094] To verify the advantages of the present invention over traditional methods, a comparative experiment was conducted. Traditional methods employ a uniform transmission strategy for all sensor data, transmitting complete data packets each time, without considering differences in data characteristics or changes in network status, lacking a data priority mechanism, and using a fixed encapsulation method without compression processing.

[0095] During routine monitoring when the change in tensile force data is small, the traditional method transmits 4 bytes of the complete tensile force value each time, with 1440 transmissions in 24 hours (calculated at 1 transmission per minute), resulting in a total data transmission volume of 5760 bytes. The proposed solution uses an interpolation encapsulation strategy to transmit 1 byte of the change value each time, with 1440 transmissions in 24 hours, resulting in a total data transmission volume of 1440 bytes. The proposed solution achieves a bandwidth saving rate of 75%.

[0096] When network quality degrades to N = 0.4 during the anomaly detection phase, traditional solutions still use full encapsulation to transmit all data. This leads to increased transmission failure rates during network congestion, with critical data transmission delays averaging 5 to 10 seconds. The solution of this invention determines a full encapsulation strategy based on the network state parameter N = 0.4 and the priority index 0.33232, and increases transmission priority, reducing critical data transmission delays to 1 to 2 seconds and improving transmission success rate by 30%.

[0097] In emergency situations where image data size is 450KB, traditional methods directly transmit the 450KB raw image data. With poor network quality, transmission times can reach 60 to 90 seconds, and the failure rate can be as high as 40%. This invention addresses this by determining a compression and encapsulation strategy based on the data size being greater than a 100KB threshold and the network status parameter N being 0.2. With a dynamic compression ratio λ of 0.66, the compressed data size is 300KB, reducing transmission time to 30 to 45 seconds, lowering the failure rate to 10%, improving transmission efficiency by 50%, and increasing the success rate by 75%.

[0098] Regarding data priority scheduling, traditional solutions lack a priority mechanism and send all data in chronological order. High-value alarm data may queue up and wait for low-priority data to complete, resulting in an average alarm data delay of 8 to 15 seconds. The solution of this invention calculates a priority index (Priority) and a scheduling weight (W). Alarm data with a priority index of 0.746 and a scheduling weight of 1.432 receives the highest transmission priority and is sent immediately. This reduces the average alarm data delay to 1 to 3 seconds, improving response speed by 80%.

[0099] Regarding the timeliness assurance of data that has not been transmitted for a long time, traditional solutions lack a time enhancement factor mechanism. Low-change-rate data may not be transmitted for extended periods, leading to incomplete monitoring records and a data integrity loss rate of approximately 15%. The solution of this invention uses a time enhancement factor of 1 multiplied by α (t). The volatility index V of data that has not been transmitted for a long time increases with the time interval t. Even with relatively small changes, the time enhancement factor increases the volatility index, thus securing a transmission opportunity. This reduces the data integrity loss rate to 3% and improves data integrity by 80%.

[0100] Regarding energy consumption during 24-hour operation, traditional solutions periodically transmit all data, including redundant data, resulting in a total data transmission volume of approximately 2.5GB over 24 hours, which is considered high energy consumption. The solution of this invention, through differential encapsulation, dynamic compression, and sampling encapsulation strategies, reduces the total data transmission volume to approximately 1.4GB over 24 hours, saving 45% of bandwidth and reducing energy consumption by 38%.

[0101] This embodiment demonstrates that the present invention not only solves the efficiency and reliability problems in data transmission, but also significantly improves the overall performance and economy of the system through an intelligent adaptive mechanism, providing a complete data transmission solution for icing monitoring. The solution achieves a 75% bandwidth saving rate during normal monitoring, a 30% increase in transmission success rate during abnormal situation detection, reduces critical data transmission latency from 5-10 seconds to 1-2 seconds, and improves transmission efficiency by 50%, transmission success rate by 75%, alarm data response speed by 80%, and data integrity by 80% during 24-hour operation, while also saving 45% of bandwidth and reducing energy consumption by 38%.

[0102] Example 4 is an embodiment of the present invention, which provides a transparent transmission and adaptive encapsulation system for sensor data of an icing monitoring terminal, comprising: The data acquisition module is used to acquire sensor data from the icing monitoring terminal; The feature analysis module is used to construct a data feature vector corresponding to the sensing data. The data feature vector includes data type, data size, volatility index and priority index. The encapsulation decision module is used to determine the encapsulation strategy corresponding to the sensing data through a decision function based on the data feature vector, network state parameters and terminal energy state, and to encapsulate the sensing data according to the encapsulation strategy to generate a data packet containing metadata header and data payload. The transmission scheduling module is used to calculate the scheduling weight of the data packet based on the data feature vector, and to determine the transmission order of the data packet according to the scheduling weight and a preset priority queue mechanism for sending.

[0103] This embodiment also provides an electronic device applicable to a transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as proposed in the above embodiment.

[0104] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements a transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as proposed in the above embodiments.

[0105] The storage medium proposed in this embodiment and the method for transparent transmission and adaptive encapsulation of sensor data for icing monitoring terminal proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0106] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0107] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for transparent transmission and adaptive encapsulation of sensor data from an icing monitoring terminal, characterized in that: include, Acquire sensor data from the icing monitoring terminal; Construct a data feature vector corresponding to the sensor data, wherein the data feature vector includes data type, data size, volatility index and priority index; Based on the data feature vector, network state parameters, and terminal energy state, the encapsulation strategy corresponding to the sensing data is determined through a decision function. The sensor data is encapsulated according to the encapsulation strategy to generate a data packet containing metadata header and data payload; The scheduling weight of the data packet is calculated based on the data feature vector, and the transmission order of the data packet is determined and sent according to the scheduling weight and the preset priority queue mechanism.

2. The transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in claim 1, characterized in that: The construction of the data feature vector corresponding to the sensing data includes: The sensor data is type-identified to determine the data type; The original size of the sensor data is calculated to determine the data scale. The relative change between the current sampled value and the historical sampled value is obtained, and the volatility index is determined by combining the time interval since the last transmission. Determine whether the sensor data triggers a preset alarm condition, and determine the priority indicator by combining the volatility indicator and the similarity of historical data.

3. The transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in claim 2, characterized in that: Determining the volatility indicator includes: The difference between the current sampled value and the historical sampled value is calculated and normalized based on the sensor's measurement range to obtain the relative change. Calculate the time enhancement factor based on the time interval and the preset time decay coefficient; The volatility index is obtained by combining the relative change with the time enhancement factor.

4. The transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in claim 3, characterized in that: The step of determining the encapsulation strategy corresponding to the sensing data through a decision function includes: The transmission quality parameters of the current network are collected as the network status parameters; The current energy balance of the terminal is obtained as the energy state of the terminal; The data feature vector, network state parameters, and terminal energy state are input into a preset decision matrix; The encapsulation strategy is determined based on the decision results output by the decision matrix.

5. The transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in claim 4, characterized in that: The determination of the priority index includes: Determine the alarm indication factor corresponding to the sensor data; Calculate the redundancy estimate of the sensor data based on historical transmission data; The priority index is obtained by weighting the alarm indicator factor, the volatility index, and the redundancy estimate with their respective weight coefficients.

6. The transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in claim 5, characterized in that: The encapsulation process of the sensing data according to the encapsulation strategy includes: The dynamic compression ratio is determined based on the network status parameters. Select a compression algorithm that matches the data type; The sensor data is compressed according to the dynamic compression ratio and the compression algorithm to generate compressed data. A compression algorithm identifier is added to the compressed data to form the data payload.

7. The transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in claim 6, characterized in that: Determining the dynamic compression ratio based on the network state parameters includes: The network state parameters are normalized to obtain network quality indicators; The dynamic compression ratio is calculated based on the preset base compression ratio, the preset adjustment coefficient, and the network quality index.

8. A transparent transmission and adaptive encapsulation system for sensor data of an icing monitoring terminal, employing the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire sensor data from the icing monitoring terminal; The feature analysis module is used to construct a data feature vector corresponding to the sensing data. The data feature vector includes data type, data size, volatility index and priority index. The encapsulation decision module is used to determine the encapsulation strategy corresponding to the sensing data through a decision function based on the data feature vector, network state parameters and terminal energy state, and to encapsulate the sensing data according to the encapsulation strategy to generate a data packet containing metadata header and data payload. The transmission scheduling module is used to calculate the scheduling weight of the data packet based on the data feature vector, and to determine the transmission order of the data packet according to the scheduling weight and a preset priority queue mechanism for sending.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the transparent transmission and adaptive encapsulation method for sensor data of an icing monitoring terminal as described in any one of claims 1 to 7.

Images