Visual synchronous updating method and system for real-time data stream of Internet of Things

Abstract

The invention discloses a visual synchronous updating method and system for real-time data streams of the Internet of Things, and the method comprises the steps: obtaining basic features, and calculating the comprehensive importance score of each data stream through employing an analytic hierarchy process; establishing a dynamic weight distribution table, determining a priority and a display resource distribution proportion by utilizing piecewise function mapping, and setting a differential updating frequency; dynamically adjusting the visual size of each data stream based on the priority and the updating frequency to form a hierarchical interface layout; establishing a synchronous updating scheduling queue, managing updating tasks with different priorities, and triggering interface refreshing according to a preset frequency; executing load balancing control, and dynamically adjusting the actual updating frequency of each priority data stream; and continuously tracking the importance change of the data flow, recalculating the comprehensive importance score when abnormal fluctuation is detected, and updating the weight distribution table and the priority. According to the method, accurate distribution of display resources, self-optimization of interface layout and efficient synchronization of updating are realized.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of Internet of Things information, in particular to a real-time data stream visualization synchronous updating method and system for Internet of Things. BACKGROUND

[0002] The rapid development of Internet of Things technology enables massive sensor devices to continuously generate real-time data streams. The visualization of these data streams has become a core technical requirement in key application fields such as industrial monitoring, smart cities, and environmental monitoring. Real-time data stream visualization systems need to convert complex multi-source heterogeneous data into intuitive graphical interfaces to provide timely and accurate information support for decision-makers.

[0003] Chinese Patent, Publication Number: CN114124717B, Publication Date: March 8, 2024, discloses a "distribution network line topology visualization monitoring method based on Internet of Things technology", which includes the following steps: data updating for obtaining the superior and inferior network connection relationship information of known data terminals; connection updating for completing data terminal initialization; state updating for setting confirmation data packets and starting data transmission modules; hierarchical updating for constructing data terminal topology connection relationships or directed graphs in the initial state of the service platform, thereby completing the visualization monitoring of the data terminal line topology.

[0004] The above technical solution has the following disadvantages: a unified updating strategy and display mode are adopted, and the same refresh frequency and interface layout are used for all data sources. This processing method ignores the significant differences in business value and urgency of different data streams, leading to unreasonable allocation of system resources, and key information may be overwhelmed by a large amount of secondary data, affecting users' timely access to important information. SUMMARY

[0005] To solve the problems of the prior art, the purpose of the present application is to provide a real-time data stream visualization synchronous updating method and system for Internet of Things, which can establish a dynamic weight distribution mechanism based on data importance, and realize differentiated display and updating of data streams with different importance.

[0006] To achieve the above purpose, the technical solution adopted by the present application is as follows: The present application provides a real-time data stream visualization synchronous updating method for Internet of Things, which comprises the following steps: S101, a multi-dimensional feature extraction algorithm is used to extract multi-dimensional features from the data stream, and different dimensions of features are fused to obtain basic features; S102, according to the basic features, the hierarchical analysis method is used to calculate the weight coefficients of each dimension, evaluate the standardized feature vectors, obtain the scores of each dimension, and calculate the comprehensive importance score; S103, according to the comprehensive importance score, using a piecewise function mapping mechanism to correspond different priority levels to each score interval, and determining the display resource allocation proportion of each data stream through the priority level; S104, setting a differentiated update frequency according to the priority level, and judging the refresh timing of each data stream through a time stamp comparison mechanism; S105, dynamically adjusting the visualization interface size of each data stream according to the display resource allocation proportion of each data stream, and obtaining a hierarchical interface layout scheme; S106, establishing a synchronous update scheduling queue according to the update frequency of each data stream and the interface layout scheme, managing data stream update tasks of different priority levels, and triggering interface refresh operations of corresponding data streams according to a preset update frequency through a time polling mechanism; S107, performing load balancing control according to the synchronous update scheduling queue, and dynamically adjusting the actual update frequency of data streams of each priority level; S108, continuously tracking the importance changes of each data stream, and if abnormal data fluctuation is detected, recalculating the comprehensive importance score, and updating the weight coefficient and the priority level according to the comprehensive importance score.

[0007] As a preferred technical solution, in S101, the multi-dimensional feature extraction of the data stream includes: according to a predefined rule base, assigning a service tag to the data stream, outputting a tag vector of the data stream, the tag vector being [service tag 1, service tag 2 tag, …, service tag m], if the data in the data stream belongs to a certain type of service tag, then the type of service tag is set to 1, otherwise it is set to 0; calculating the change rate of each type of data in the data stream, and recording the abnormal frequency of the data in a period, setting a change rate weight coefficient and an abnormal frequency weight coefficient, and using a weighted summation algorithm to obtain the dynamic characteristics of each type of data.

[0008] As a preferred technical solution, before the multi-dimensional feature extraction of the data stream in S101, the data stream is preprocessed and cleaned, including: using a sliding window algorithm to calculate the mean and standard deviation of the data in the window, and eliminating abnormal values that are more than several times the standard deviation of the mean; using linear interpolation or LSTM prediction to fill in missing values, and using a lightweight model to fill in missing values at the edge; each data stream synchronizes the time stamp through the PTP protocol to align the sampling frequency.

[0009] As a preferred technical solution, in S102, the analytic hierarchy process is used to calculate the weight coefficient of each dimension, including: constructing a judgment matrix, the elements in the judgment matrix representing the relative importance proportion of each data stream type; solving the maximum eigenvalue of the judgment matrix and the corresponding eigenvector, then normalizing the eigenvector to obtain the weight coefficient; calculating the consistency ratio to determine whether the consistency requirement is met.

[0010] As a preferred technical solution, in S104, determining the refresh timing of each data stream through a timestamp comparison mechanism includes: recording the most recent update timestamp of each data stream; comparing the current time with the sum of the most recent update timestamp and the refresh frequency according to a preset refresh frequency; if the current time >= (last update time + frequency interval), it is determined to be a refresh point; S104 also includes: obtaining the state changes of the data stream based on the refresh point to obtain the update requirement; using the update requirement, employing a scheduling algorithm to determine the actual refresh order of the data stream to obtain the execution plan; and verifying the execution plan using timestamps to determine the completion status of the refresh operation and obtain the update status.

[0011] As a preferred technical solution, in S105, the size of the visual interface of each data stream is dynamically adjusted according to the display resource allocation ratio of each data stream to obtain a hierarchical interface layout scheme. This includes: analyzing the available capacity of display resources using a resource evaluation algorithm, calculating the screen resolution and pixel distribution of the display area, and obtaining the display resource allocation ratio; dividing the interface into primary and secondary areas using a grid partitioning algorithm based on the display resource allocation ratio, and determining the size and position of each area; if the number of high-priority data streams exceeds a preset threshold, they are allocated to the primary display area, and a color mapping algorithm is used to assign a prominent color to the high-priority data streams to generate a highlighting visual effect; and further adjusting the display layout according to the priority of data streams. Based on the characteristics of the primary data stream, it allocates data to secondary display areas and generates a display effect distinct from the primary display area by adjusting transparency and grayscale values. A dynamic adjustment algorithm monitors data stream changes in real time. If the data stream priority changes, the allocation ratio is recalculated, and the sizes of the primary and secondary display areas are updated to obtain an adaptive interface layout scheme. A hierarchical layout algorithm integrates the display content of the primary and secondary display areas to generate an interface layout scheme that includes high-priority and medium-low-priority data streams. If the display effect of the interface layout scheme does not reach the preset visual hierarchy threshold, an iterative optimization algorithm is used to adjust the area boundaries and color contrast to obtain the final hierarchical interface layout scheme.

[0012] As a preferred technical solution, in S106, a synchronous update scheduling queue is established based on the update frequency and interface layout scheme of each data stream, including: assigning priorities according to the priority level of the data stream to obtain an ordered task sequence; extracting update tasks from the task sequence, obtaining the current timestamp through a time polling mechanism, and determining the task trigger time; managing data stream update tasks of different priority levels using a priority queue data structure, including: if the current timestamp matches the preset update frequency, extracting the highest priority task from the priority queue and triggering the data stream update operation; triggering the interface refresh operation of the corresponding data stream according to the preset update frequency through a time polling mechanism, including: obtaining the corresponding data stream content based on the triggered data stream update operation and generating an interface refresh instruction; updating the data stream display content through the interface refresh instruction to obtain the updated interface state; S106 also includes: extracting refresh operation feedback from the updated interface state to determine whether the task level in the priority queue needs to be adjusted; if the task level is adjusted, updating the task sequence in the priority queue and generating a new task scheduling plan.

[0013] As a preferred technical solution, in S107, load balancing control is performed based on the synchronous update scheduling queue, including: obtaining the system CPU utilization and memory usage, calculating the difference with a preset threshold, and obtaining the resource usage status; if the resource usage status shows that the CPU utilization exceeds the preset threshold, then suspending the update operation of low-priority data streams, recording the identifier of the suspended data stream, and generating an update pause list; adjusting the scheduling queue of low-priority data streams according to the update pause list, reallocating system CPU resources, and obtaining an optimized scheduling plan; comparing the difference between the memory usage and the preset threshold, determining whether the cache size of medium-priority data streams needs to be reduced, and generating a cache adjustment instruction; reducing the cache size of medium-priority data streams according to the cache adjustment instruction, releasing memory resources, and obtaining an updated cache allocation status; using a dynamic adjustment mechanism, combining the resource usage status and cache allocation status, calculating the optimal update frequency of each level of data stream, and generating frequency adjustment parameters; updating the execution frequency of each level of data stream in the scheduling queue using the frequency adjustment parameters, and obtaining the final load balancing control scheme.

[0014] As a preferred technical solution, in S108, the importance changes of each data stream are continuously tracked, including: acquiring real-time data from each data stream through a sensor network, analyzing data stream characteristics using stream processing technology, and obtaining a comprehensive importance score for each data stream; if abnormal data fluctuations are detected, the comprehensive importance score is recalculated, including: if the dynamic characteristics of a data stream exceed a preset threshold, it is judged as an abnormal fluctuation, triggering a recalculation of the comprehensive importance score to obtain an updated comprehensive importance score; the weight coefficients and priority levels are updated according to the changes in the comprehensive importance score, including: adjusting the weight allocation table using a weighted average algorithm based on the updated comprehensive importance score to determine new weight coefficients; updating the priority level identifier using the adjusted weight coefficients to generate a new priority sequence; S108 also includes: reallocating system resources using the updated priority sequence to obtain an optimized resource configuration scheme; adjusting the data stream processing order using the optimized resource configuration scheme to maintain the system's adaptive capability; and continuously monitoring system performance based on the changing trend of the data stream processing order to obtain real-time performance feedback data.

[0015] This application also provides a real-time data stream visualization and synchronous update system for the Internet of Things (IoT). The system includes: a multi-dimensional feature extraction and fusion module for extracting and fusing features from the input data stream and outputting a basic feature vector; a weight calculation and comprehensive scoring engine module for calculating the weights of each dimension based on the analytic hierarchy process (AHP), evaluating features, and generating a comprehensive importance score for the data stream; a priority mapping and resource allocation module for mapping the score to a priority level using a piecewise function, determining the display resource allocation ratio for each data stream; an update frequency control module for setting differentiated update frequencies according to priority levels and determining refresh timing using a timestamp comparison mechanism; a dynamic interface layout engine module for dynamically adjusting the size of visualization components based on resource allocation ratios to form a hierarchical interface layout; a synchronous update scheduler module for establishing a priority scheduling queue and triggering interface refresh according to the update frequency using a time polling mechanism; a load balancing and dynamic adjustment module for monitoring system load and dynamically adjusting the actual update frequency of each priority data stream to ensure smoothness; and an adaptive feedback and re-evaluation module for continuously monitoring the data stream and triggering recalculation and updates of weights, scores, priority levels, resource allocation, and frequencies when abnormal fluctuations are detected, achieving closed-loop system optimization.

[0016] Compared with the prior art, the beneficial effects of this application are as follows: This application establishes a dynamic weight allocation mechanism based on data importance, enabling differentiated display and updates of data streams with varying levels of importance. Furthermore, through dynamic prioritization and synchronous update mechanisms, this application ensures timely and accurate data stream updates, meeting the low-latency requirements of the Internet of Things (IoT); dynamic weight allocation and load balancing prevent resource waste, ensuring stable system operation under high loads; multi-dimensional feature extraction and feedback mechanisms enhance the system's adaptability to changes and reduce the impact of failures; and an adaptive interface layout presents information hierarchically and clearly, facilitating rapid user decision-making. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a real-time data stream visualization and synchronous update method for the Internet of Things (IoT) according to this application. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present application, the technical solutions in specific embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0019] This application provides a method for real-time data stream visualization and synchronous update in the Internet of Things, which aims to transform complex multi-source heterogeneous data into an intuitive graphical interface, providing timely and accurate information support for decision-makers.

[0020] The real-time data streams of the Internet of Things (IoT) originate from a variety of physical, chemical, and biological sensors, such as temperature sensors, humidity sensors, accelerometers, gyroscopes, light sensors, GPS, air quality sensors, acoustic sensors, and cameras. IoT real-time data streams also originate from actuators. Data types of IoT sensor data streams include: numerical values ​​(temperature parameters, humidity parameters, acceleration parameters, pressure parameters, etc.), strings (device status), Boolean values ​​(switch status), images, audio streams, geographic coordinates, timestamps, etc.

[0021] like Figure 1 As shown, this application discloses a method for real-time data stream visualization and synchronous update in the Internet of Things, comprising the following steps: S101 employs a multi-dimensional feature extraction algorithm to extract multi-dimensional features from the data stream and fuses features from different dimensions to obtain basic features. S102, Based on the basic characteristics, calculate the weight coefficients of each level using the analytic hierarchy process, evaluate the standardized feature vector, obtain the scores of each level, and calculate the comprehensive importance score by weighted summation of the weight coefficients and scores of each level. S103, A dynamic weight allocation table is established based on the comprehensive importance score. A piecewise function mapping mechanism is used to map each score interval to different priority levels. The display resource allocation ratio of each data stream is determined by the priority level. S104, set a differentiated update frequency based on the priority level identifier, and determine the refresh timing of each data stream through a timestamp comparison mechanism; S105, Based on the display resource allocation ratio of each data stream, the size of the visualization interface of each data stream is dynamically adjusted using an adaptive interface layout algorithm to obtain a hierarchical interface layout scheme. S106. Establish a synchronous update scheduling queue based on the update frequency and interface layout scheme of each data stream, use a priority queue data structure to manage data stream update tasks of different priority levels, and trigger the interface refresh operation of the corresponding data stream according to the preset update frequency through a time polling mechanism. S107, perform load balancing control according to the synchronous update scheduling queue, and dynamically adjust the actual update frequency of data streams of each priority level through the resource monitoring feedback mechanism; S108 employs a real-time monitoring and feedback mechanism to continuously track changes in the importance of each data stream. If abnormal data fluctuations are detected, the overall importance score is recalculated, and the dynamic weight allocation table and priority level are updated based on the overall importance score.

[0022] Furthermore, in S101, before performing multidimensional feature extraction on the data stream, the data stream is preprocessed and cleaned to ensure the quality and integrity of the data stream and reduce the impact of noise on feature extraction.

[0023] Specifically, data stream preprocessing and cleaning includes: The sliding window algorithm is used to calculate the mean and standard deviation of the data within the window.

[0024] Let the data stream sequence be: ; Define the window as: ; in, For window length, This refers to the current time.

[0025] When new data Upon arrival, the window updates to .

[0026] The sliding step size is Each time they arrive Each new data entry is swiped once.

[0027] Filter out abnormal fluctuations, such as removing outliers that exceed the mean by ±3 standard deviations.

[0028] Furthermore, data stream preprocessing and cleaning also includes: a) Use linear interpolation or LSTM prediction to fill in missing values, and use a lightweight model to fill in missing values ​​at the edges; b. Each data stream synchronizes its timestamp via the PTP protocol to align the sampling frequency.

[0029] Furthermore, in S101, multidimensional feature extraction of the data stream includes: a. Extract business label features: Assign business labels to the data stream according to a predefined rule base. Output the label vector of the data stream, which is [business label 1, business label 2, ..., business label m]. If the data in the data stream belongs to one of the business labels, then set that business label to 1; otherwise, set it to 0.

[0030] b. Extract dynamic features: Calculate the rate of change of various types of data in the data stream within the window, record the frequency of data anomalies within a period, set the weighting coefficients for the rate of change and the weighting coefficients for the frequency of anomalies, and use a weighted summation algorithm to obtain the dynamic features of various types of data.

[0031] For example, suppose sensor one collects temperature and pressure data every second.

[0032] Using a sliding window algorithm with a window length of 10 seconds, the mean and standard deviation of temperature and pressure data within the window were calculated, yielding a mean temperature of 25.4°C, a standard deviation of 0.3, a mean pressure of 101.2 kPa, and a standard deviation of 0.5.

[0033] In a predefined rule base, data flow types include: safety critical, business monitoring, and environmental awareness. Temperature data is mapped to the "environmental awareness" label, and pressure data is mapped to the "safety critical" label. The output label vector is [1,0,1], indicating that temperature data and pressure data correspond to the safety critical and environmental awareness data flow types, respectively.

[0034] The normal temperature range was set at 20-30°C, and the pressure anomaly threshold was set at 100 kPa. The rate of change of temperature (ΔT = 0.1°C / s) and the rate of change of pressure (ΔP = 0.2 kPa / s) within a 10-second window were calculated. Considering historical anomaly frequencies, assuming three instances of temperature anomalies (exceeding 20-30°C) and two instances of pressure anomalies (exceeding 100 kPa) occurred in the past 24 hours, a weighted summation algorithm was used with weights of 0.6 (rate of change) and 0.4 (anomaly frequency). The calculated dynamic features of the temperature data were 0.6 × 0.1 + 0.4 × 3 = 1.26, and the dynamic features of the pressure data were 0.6 × 0.2 + 0.4 × 2 = 0.92. Finally, the label vectors and the dynamic features of various data types were fused to generate basic features in vector form. .

[0035] Furthermore, S101 also includes: predicting equipment failures based on basic features: comparing the basic features with the Euclidean distance of historical failure data to determine whether the equipment status is abnormal.

[0036] For two vectors and The Euclidean distance is: .

[0037] If the data change exceeds a preset threshold, the anomaly probability is calculated based on historical anomaly frequencies, and an anomaly detection result is generated by combining this with business tags. Based on the anomaly detection result, the K-means algorithm is used to cluster the basic features, generating a cluster center library for historical fault data. The Euclidean distance between the current basic feature and each cluster center representing a fault in the cluster center library is calculated, and then the minimum distance is compared with a threshold. If the minimum distance is less than the threshold, the equipment status is determined to be abnormal.

[0038] For example, suppose the historical fault data has two cluster centers representing faults. Fault vector 1: Fault vector 2: .

[0039] Calculate the distance to fault vector 1: Calculate the distance to fault vector 2: Minimum distance: The threshold T is set to 0.3. Since 0.253 < 0.3, the device is judged to be in an abnormal state.

[0040] Furthermore, in S102, before calculating the weight coefficients of each level using the analytic hierarchy process, the basic features are normalized to obtain standardized basic features, thereby obtaining standardized dynamic features.

[0041] Normalization is specifically achieved by performing a linear transformation on the basic features so that the result is mapped to the range [0,1], thus achieving equal scaling of the basic features; or, mapping the basic features to a distribution with a mean of 0 and a standard deviation of 1.

[0042] Furthermore, in S102, the Analytic Hierarchy Process (AHP) is used to determine the relative weights of each data stream type. These weights are subsequently used to calculate a comprehensive importance score, thereby supporting data stream prioritization and resource allocation decisions.

[0043] The weight coefficients for each dimension are calculated using the analytic hierarchy process (AHP), including: Construct a judgment matrix A, where the elements of A represent the relative importance ratio of each data stream type; By solving for the largest eigenvalue of matrix A The corresponding feature vector is then normalized to obtain the weight coefficients. Calculate the consistency ratio CR, CR = CI / RI, where CI is the consistency index. , where n is the matrix order; RI is the random consistency index, which is a predefined value.

[0044] For example, data stream types include: safety critical, business monitoring, and environmental awareness. In the safety critical data stream type, the data is the equipment failure rate, and the standardized dynamic characteristic of the equipment failure rate is assumed to be 0.8; in the business monitoring data stream type, the data is production efficiency, and the standardized dynamic characteristic of production efficiency is assumed to be 0.6; in the environmental awareness data stream type, the data is the temperature anomaly index, and the standardized dynamic characteristic of the temperature anomaly index is assumed to be 0.4.

[0045] Next, the Analytic Hierarchy Process (AHP) is used to construct a judgment matrix to compare the importance of each data stream type.

[0046] Assume the importance of security criticality compared to business monitoring is 3 (i.e., security criticality is 3 times more important than business monitoring); the importance of security criticality compared to environmental awareness is 6 (security criticality is 6 times more important than environmental awareness); and the importance of business monitoring compared to environmental awareness is 2 (business monitoring is 2 times more important than environmental awareness). Matrix A is: .

[0047] After normalization: Safety criticality weight coefficient W1 = 0.649, business monitoring weight coefficient W2 = 0.243, and environmental awareness weight coefficient W3 = 0.108. Adjust the weights according to the data flow type: If the data flow is safety critical, set W1=0.6, W2=0.3, W3=0.1; if it is business monitoring data flow, set W2=0.3, W1=0.5, W3=0.2; if it is environmental awareness data flow, set W3=0.1, W1=0.6, W2=0.3.

[0048] Assuming the current data flow is a safety-critical data flow type, we use W1=0.6, W2=0.3, and W3=0.1. Calculate the overall importance score R using the formula R=W1×S1+W2×S2+W3×S3. Substituting the values: R=0.6×0.8+0.3×0.6+0.1×0.4=0.48+0.18+0.04=0.7.

[0049] The analytic hierarchy process ensures that the sum of weights is 1, with a preset RI of 0.58. Verification shows that the AHP matrix consistency ratio CR = 0.02 < 0.1, meeting the consistency requirements. An overall importance score of R = 0.7 indicates high data importance, allowing for priority allocation of computational resources.

[0050] Specifically, in S103, based on the comprehensive importance score R, a piecewise function mapping mechanism is used to divide the score interval into [0.8, 1.0], [0.5, 0.8), and [0, 0.5). The score interval [0.8, 1.0] corresponds to a high priority level, the score interval [0.5, 0.8) corresponds to a medium priority level, and the score interval [0, 0.5) corresponds to a low priority level. The priority level identifier for each data stream is determined, and the allocation ratio of each data stream in the display resource pool is determined based on the priority level identifier.

[0051] In this application, a priority level identifier is assigned to each data stream, represented by an integer: 1 for high-level, 2 for medium-level, and 3 for low-level, and stored in the database. The data structure is {flow_id, priority, last_update_timestamp, update_frequency}, where update_frequency is automatically set to 100, 500, or 2000 milliseconds according to the priority.

[0052] A dynamic weight allocation table is constructed using priority level identifiers: if the data stream has a high priority level, a high weight coefficient is assigned; if the data stream has a medium priority level, a medium weight coefficient is assigned; if the data stream has a low priority level, a low weight coefficient is assigned, thus obtaining the dynamic weight allocation table.

[0053] The weight coefficients of each data stream are obtained from the dynamic weight allocation table. Combined with the total capacity of the resource pool, the allocation ratio of each data stream is calculated to obtain the allocation ratio value. The allocation ratio value is calculated by multiplying the weight coefficient by the resource pool capacity.

[0054] Based on the allocation ratio, display resources in the resource pool are allocated. A linear allocation algorithm is used to distribute resources proportionally to each data stream, resulting in a resource allocation result. The resource allocation result is then verified for the allocated resource amount for each data stream. If the deviation between the allocated resource amount and the allocation ratio exceeds a preset threshold, the weighting coefficients are adjusted, the allocation ratio is recalculated, and an updated resource allocation result is obtained. Using the updated resource allocation result, the resource allocation information is written to the resource pool's configuration table, and a mapping between data stream identifiers and resource amounts is used to generate the final resource allocation configuration.

[0055] For example, suppose data streams A, B, and C have scores of R_A=0.85, R_B=0.6, and R_C=0.3, respectively. According to the piecewise function, the score interval [0.8, 1.0] is mapped to high priority with an allocation ratio of 50%; the interval [0.5, 0.8) is medium priority with an allocation ratio of 30%; and the interval [0, 0.5) is low priority with an allocation ratio of 20%. First, the scores are classified using the piecewise function mapping mechanism: a function f(R) = {high if R≥0.8; medium if 0.5≤R<0.8; low if R<0.5} is defined. Applying this function to A, B, and C yields A as high, B as medium, and C as low. Next, the allocation ratio is determined based on the priority level. The total resource pool capacity is calculated to be 1000 units, with A receiving 500 units, B receiving 300 units, and C receiving 200 units. To ensure dynamic allocation, the system reads the data stream scores every minute, updates the priority levels, and uses a formula... Recalculate the allocation, where Allocate a quantity for the i-th data stream. Priority level ratio, (Total capacity).

[0056] Specifically, in S104, if the data stream has a high priority level, the update frequency is set to 100 milliseconds; if the data stream has a medium priority level, the update frequency is set to 500 milliseconds; and if the data stream has a low priority level, the update frequency is set to 2000 milliseconds.

[0057] Determining the refresh timing of each data stream through timestamp comparison mechanisms includes: Record the most recent update timestamp of each data stream; based on the preset refresh frequency, compare the current time with the sum of the most recent update timestamp and the refresh frequency. If the current time >= (last update time + frequency interval), it is determined to be a refresh point.

[0058] Furthermore, based on the refresh points, the state changes of the data stream are obtained to determine the update requirements. Using these requirements, a scheduling algorithm is employed to determine the actual refresh order of the data stream, resulting in an execution plan. Finally, the execution plan is verified using timestamps to assess the completion status of the refresh operations and obtain the update status.

[0059] For example, when data stream A (flow_id=A1, priority=1) is initialized, last_update_timestamp is set to the current timestamp 1704067200000 milliseconds (January 1, 2025 00:00:00), and update_frequency is 100 milliseconds.

[0060] Next, a scheduled task scans all data streams every 10 milliseconds to obtain the current timestamp, current_timestamp, for example, 1704067200100 milliseconds, and calculates the time difference delta = current_timestamp - last_update_timestamp.

[0061] For data flow A, delta = 100 milliseconds, which is equal to its update_frequency (100 milliseconds), triggering an update. The data refresh interface is called to update last_update_timestamp to 1704067200100 milliseconds and logs {flow_id=A1, update_time=1704067200100}.

[0062] If data stream B (priority=2, update_frequency=500 milliseconds, last_update_timestamp=1704067200000) has delta=100 milliseconds, which is less than 500 milliseconds, no update will be triggered.

[0063] Similarly, data stream C (priority=3, update_frequency=2000 milliseconds) is not updated because delta is less than 2000 milliseconds. The system ensures that high-priority data streams are updated frequently and low-priority data streams are kept resource-efficient by comparing delta and update_frequency (if delta >= update_frequency then trigger_update()).

[0064] If the data stream priority changes dynamically, for example, A1 changes from high to medium, the system detects the priority change, updates update_frequency to 500 milliseconds, and resets last_update_timestamp to the current timestamp.

[0065] Analysis shows that this mechanism ensures that high-priority data streams (such as real-time monitoring data) are updated frequently at 100 milliseconds to meet real-time requirements, while low-priority data streams (such as log archiving) are updated frequently at 2000 milliseconds to reduce system load. The time complexity of the timestamp comparison algorithm is O(n), where n is the number of data streams, making it suitable for large-scale data stream scenarios.

[0066] Furthermore, in S105, high-priority data streams occupy the main area of ​​the interface and are marked with a prominent color, while medium and low-priority data streams are assigned to secondary display areas.

[0067] In S105, based on the display resource allocation ratio of each data stream, an adaptive interface layout algorithm is used to dynamically adjust the size of the visualization interface of each data stream to obtain a hierarchical interface layout scheme, including: The available capacity of display resources is analyzed using a resource evaluation algorithm, and the screen resolution and pixel distribution of the display area are calculated to obtain the display resource allocation ratio. Based on the display resource allocation ratio, the interface is divided into main and secondary areas using a grid partitioning algorithm, and the size and position of each area are determined. If the number of high-priority data streams exceeds a preset threshold, they are allocated to the main display area, and a striking color is assigned to the high-priority data streams through a color mapping algorithm to generate a prominent visual effect. Based on the characteristics of low- and medium-priority data streams, they are allocated to secondary display areas, and by adjusting the transparency and grayscale values, a display effect distinct from the primary display area is generated. A dynamic adjustment algorithm is used to monitor changes in the data stream in real time. If the data stream priority changes, the allocation ratio is recalculated and the size of the main display area and the secondary display area is updated to obtain an adaptive interface layout scheme. By integrating the display content of the main display area and the secondary display area through a hierarchical layout algorithm, an interface layout scheme containing high-priority and medium-low-priority data streams is generated. If the display effect of the interface layout scheme does not reach the preset visual hierarchy threshold, the area boundaries and color contrast are adjusted through iterative optimization algorithms to obtain the final hierarchical interface layout scheme.

[0068] For example, based on the adaptive interface layout algorithm, the size of the visual components is first dynamically adjusted by the display resource allocation ratio. Assuming the screen resolution is 1920x1080 pixels, the total display resources are 100%, the high priority data stream is allocated 60% (1152x648 pixels), the medium priority is allocated 25% (480x270 pixels), and the low priority is allocated 15% (288x162 pixels).

[0069] The algorithm employs a weighted allocation model, with weights based on data flow priority: high priority has a weight of 0.6, medium priority has a weight of 0.25, and low priority has a weight of 0.15. The calculation formula is: component area = total area × weight. During dynamic adjustment, CSS media queries and JavaScript are used to monitor window size changes in real time to ensure a constant ratio.

[0070] When the window is scaled to 1440x810 pixels, the high-priority area is adjusted to 864x486 pixels, maintaining a 60% aspect ratio.

[0071] Next, the high-priority data stream occupies the main area of ​​the interface, centered, with coordinates ranging from (384,216,1536,864), and is identified using a striking color such as RGB(255,69,0). Through color contrast analysis (contrast calculation formula: (L1+0.05) / (L2+0.05), where L1 is the foreground brightness and L2 is the background brightness), the contrast ratio is ensured to be greater than 4.5:1 to meet readability requirements.

[0072] The secondary area is allocated to low-priority data streams, with medium-priority data located on the left (0,0,480,270) and low-priority data located at the bottom (0,810,288,162), using softer colors such as RGB(70,130,180) and RGB(144,238,144).

[0073] The layout uses a grid system (CSS Grid), defining 12x12 grid cells. High priority cells occupy 8x8 cells, while medium and low priority cells occupy 4x3 and 3x2 cells respectively.

[0074] The final result is a hierarchical interface layout. The algorithm uses priority weights, area allocation, and color contrast analysis to ensure clear visual hierarchy, efficient information transmission, and maintain proportion and readability during dynamic adjustments.

[0075] Furthermore, in S106, a synchronous update scheduling queue is established based on the update frequency of each data stream and the interface layout scheme, including: allocating priorities according to the priority level of the data stream to obtain an ordered task sequence; extracting update tasks from the task sequence, obtaining the current timestamp through a time polling mechanism, and determining the task trigger time.

[0076] In S106, a priority queue data structure is used to manage data stream update tasks with different priority levels, including: if the current timestamp matches the preset update frequency, the highest priority task is extracted from the priority queue to trigger the data stream update operation.

[0077] In S106, the interface refresh operation of the corresponding data stream is triggered according to the preset update frequency through the time polling mechanism, including: obtaining the corresponding data stream content according to the triggered data stream update operation and generating the interface refresh instruction; updating the data stream display content through the interface refresh instruction to obtain the updated interface state.

[0078] Furthermore, S106 also includes: extracting refresh operation feedback from the updated interface state, determining whether the task level in the priority queue needs to be adjusted; if the task level is adjusted, updating the task sequence in the priority queue and generating a new task scheduling plan.

[0079] In this application, data stream priorities are defined. For example, high-priority data streams (such as real-time stock prices) are updated every 1 second, medium-priority data streams (such as news summaries) every 5 seconds, and low-priority data streams (such as historical statistics) every 30 seconds. The priority queue adopts a min-heap structure, with heap nodes storing task information, including data stream ID, priority, update frequency, and next trigger time. During initialization, all data stream tasks are inserted into the heap according to priority.

[0080] For example, task A (priority 1, frequency 1 second) is inserted at time t=0, and task B (priority 2, frequency 5 seconds) is also inserted at t=0. The time polling mechanism is driven by a timer with a precision of 0.1 seconds to check the trigger time of the top task. If the current time t is greater than or equal to the trigger time of the top task, the task execution interface will pop up to refresh.

[0081] The refresh operation calls a predefined API, such as calling updateStockPrice() to update the stock price interface, with the stock code and timestamp as parameters. After the task is executed, the next trigger time is calculated (e.g., if task A is reinserted into the heap after t=1 seconds, the trigger time is t+1 seconds).

[0082] To avoid frequent heap operations, a time wheel algorithm is used for optimization. The time wheel is divided into 100 slots (covering 10 seconds, with a slot interval of 0.1 seconds). Tasks are assigned to corresponding slots according to their trigger time. For example, task A, with a trigger time of t=1.2 seconds, is assigned to slot 12. During polling, only the task in the current slot is checked, reducing the time complexity to O(1). If high-priority tasks (such as stock prices) are triggered frequently, it may cause delays in low-priority tasks. Analysis shows that the average waiting time for tasks in the heap is 0.5 seconds, and the maximum delay does not exceed 1 second (because high-priority tasks account for 20%). If task B fails to complete in time due to network latency, the system records its failure count. After more than 3 failures, its priority is reduced to 3 to ensure system stability. Finally, the interface refresh frequency is consistent with the preset, for example, the stock price interface updates every second, the news summary updates every 5 seconds, and the historical statistics update every 30 seconds, meeting the real-time requirements.

[0083] Furthermore, in S107, load balancing control is performed based on the synchronous update scheduling queue, including: Obtain the system's CPU utilization and memory usage, calculate the difference between these values ​​and preset thresholds, and obtain the resource usage status. If the resource usage status shows that the CPU utilization exceeds the preset threshold, the update operation of low-priority data streams is paused, the paused data stream identifier is recorded, and an update pause list is generated; specifically, if the system CPU utilization exceeds 80%, the update of low-priority data streams is paused. Based on the updated pause list, the scheduling queue of low-priority data streams is adjusted, system CPU resources are reallocated, and an optimized scheduling plan is obtained. Compare the difference between memory utilization and a preset threshold to determine whether the cache size of the medium-priority data stream needs to be reduced, and generate a cache adjustment instruction; specifically, if the memory utilization exceeds 75%, reduce the cache size of the medium-priority data stream. Based on the cache adjustment instructions, reduce the cache size of medium-priority data streams, release memory resources, and obtain the updated cache allocation status. A dynamic adjustment mechanism is adopted, which combines resource usage status and cache allocation status to calculate the optimal update frequency of data streams at all levels and generate frequency adjustment parameters. By adjusting the frequency parameters, the execution frequency of each level of data flow in the scheduling queue is updated, resulting in the final load balancing control scheme.

[0084] For example, firstly, a resource monitoring module is deployed, using the Linux system's proc / stat and proc / meminfo interfaces to collect CPU utilization and memory usage data every second. CPU utilization is calculated by dividing the sum of user-mode, system-mode, and I / O wait times by the total time, assuming the current value is 85%. When CPU utilization exceeds 80%, a load balancing algorithm is triggered, pausing low-priority data stream updates. Specifically, this is achieved by setting the flag of low-priority tasks in the scheduling queue to "pause," ensuring that high-priority data streams (such as real-time video streams with a bandwidth requirement of 10Mbps) have priority access to CPU resources. After the pause, the update frequency of low-priority data streams (such as log synchronization with a bandwidth requirement of 1Mbps) is reduced to 0 times / second, releasing approximately 5% of CPU resources.

[0085] Next, the memory usage is checked. Assuming the current value is 78%, which exceeds the 75% threshold, the system reduces the cache size of medium-priority data streams (such as batch data processing, which uses 50MB of cache). The algorithm reduces the cache size from 50MB to 30MB, reducing memory usage by 40%. This is achieved by adjusting the maxmemory parameter of the Redis cache.

[0086] The resource monitoring and feedback mechanism analyzes resource usage trends every 5 seconds and uses a sliding window algorithm to calculate the average CPU and memory usage over the past 30 seconds. If the CPU usage drops below 70%, the low-priority data stream updates are resumed, and the frequency is adjusted to once per second. If the memory usage drops below 60%, the medium-priority data stream cache is increased to 40MB.

[0087] Analysis showed that a 10% reduction in CPU utilization can improve system response time by approximately 20ms, and freeing up 20MB of memory can reduce page swapping frequency by approximately 30%.

[0088] Through the above dynamic adjustments, the system maintains stable operation under high load, and the latency of high-priority services (such as video streams) is controlled within 50ms, ensuring that resource allocation matches business needs.

[0089] Furthermore, in S108, a real-time monitoring and feedback mechanism is adopted to continuously track the changes in the importance of each data stream, including: acquiring real-time data from each data stream through a sensor network, analyzing the characteristics of the data streams using stream processing technology, and obtaining a comprehensive importance score for each data stream.

[0090] In S108, if abnormal data fluctuations are detected, the comprehensive importance score is recalculated. This includes: if the dynamic characteristics of the data stream exceed a preset threshold, it is judged as an abnormal fluctuation, triggering the recalculation of the comprehensive importance score to obtain an updated comprehensive importance score.

[0091] In S108, the dynamic weight allocation table and priority level are updated according to the changes in the comprehensive importance score, including: adjusting the weight allocation table using a weighted average algorithm based on the updated comprehensive importance score, determining new weight coefficients, and updating the dynamic weight allocation table; updating the priority level identifier using the adjusted weight coefficients, and generating a new priority sequence.

[0092] Furthermore, S108 also includes: adopting the updated priority sequence to reallocate system resources and obtain an optimized resource configuration scheme; adjusting the data stream processing order through the optimized resource configuration scheme to maintain the system's adaptive capability; and continuously monitoring system performance based on the changing trend of the data stream processing order to obtain real-time performance feedback data.

[0093] For example, the system continuously tracks the changes in the importance of each data stream through a real-time monitoring and feedback mechanism. Assume there are three data streams A, B, and C, each generating 1000 records per second, including metrics such as traffic, latency, and error rate. The monitoring module collects data every 5 seconds and uses a sliding window algorithm (window size 30 seconds) to calculate the statistical characteristics of each data stream. For example, the mean traffic of stream A is μ = 500 kb / s, and the standard deviation is σ = 50 kb / s.

[0094] Anomaly detection uses the Z-score method. If a data point deviates from the mean by more than 3σ (e.g., if the A flow suddenly increases to 650kb / s, Z=(650-500) / 50=3), an anomaly alarm is triggered.

[0095] When recalculating the importance score, a weighted scoring formula is used: S = 0.4 × traffic share + 0.3 × latency sensitivity + 0.3 × error rate, where traffic share = traffic of this flow / total traffic, and latency sensitivity and error rate are obtained by normalizing historical data (for example, the latency sensitivity of A is 0.8 and the error rate is 0.05).

[0096] Assuming the calculated scores for A, B, and C are 0.85, 0.65, and 0.5 respectively, the system updates the weight allocation table based on the scores and uses the softmax function to map the scores to weights: W_i=exp(S_i) / Σexp(S_j), resulting in weights of 0.45, 0.33, and 0.22 for A, B, and C respectively.

[0097] Priority levels are assigned based on weight thresholds (>0.4 for high, 0.3-0.4 for medium, <0.3 for low), with A for high priority, B for medium, and C for low.

[0098] After the weights and priorities are updated, the system adjusts the resource allocation. For example, high-priority data stream A is allocated 60% of the bandwidth (600Mbps), B is allocated 30% (300Mbps), and C is allocated 10% (100Mbps).

[0099] Closed-loop adjustments are executed once per minute through a feedback loop to ensure that the system dynamically optimizes resource allocation based on changes in data flow and maintains its adaptive capabilities.

[0100] This application also provides a real-time data stream visualization and synchronous update system for the Internet of Things, which includes: a multi-dimensional feature extraction and fusion module, a weight calculation and comprehensive scoring engine module, a priority level mapping and resource allocation module, an update frequency control module, a dynamic interface layout engine module, a synchronous update scheduler module, a load balancing and dynamic adjustment module, and an adaptive feedback and re-evaluation module.

[0101] The system comprises the following modules: a multi-dimensional feature extraction and fusion module, which extracts and fuses features from the input data stream to output a basic feature vector; a weight calculation and comprehensive scoring engine module, which calculates the weights of each dimension based on the analytic hierarchy process (AHP), evaluates features, and generates a comprehensive importance score for each data stream; a priority mapping and resource allocation module, which maps scores to priority levels using a piecewise function to determine the display resource allocation ratio for each data stream; an update frequency control module, which sets differentiated update frequencies based on priority levels and uses a timestamp comparison mechanism to determine refresh timing; a dynamic interface layout engine module, which dynamically adjusts the size of visualization components based on resource allocation ratios to create a hierarchical interface layout; a synchronous update scheduler module, which establishes a priority scheduling queue and triggers interface refreshes according to the update frequency using a time-polling mechanism; a load balancing and dynamic adjustment module, which monitors system load and dynamically adjusts the actual update frequency of each priority data stream to ensure smoothness; and an adaptive feedback and re-evaluation module, which continuously monitors the data stream and, when abnormal fluctuations are detected, triggers the recalculation and update of weights, scores, priority levels, resource allocation, and frequencies to achieve closed-loop system optimization.

[0102] It should be noted that the terms "first," "second," and similar terms used in this application specification and claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, "a" or "one," and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. "A plurality" or "several" indicates at least two. Unless otherwise stated, terms such as "front," "back," "left," "right," "lower," and / or "upper" are for illustrative purposes only and are not limited to a location or spatial orientation. Terms such as "comprising" or "including" indicate that the elements or objects preceding "comprising" encompass the elements or objects listed following "comprising" or "including" and their equivalents, and do not exclude other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect.

[0103] The singular forms “a,” “the,” and “the” used in this application specification and appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0104] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A method for real-time data stream visualization and synchronous update in the Internet of Things, characterized in that, The method includes: S101 employs a multi-dimensional feature extraction algorithm to extract multi-dimensional features from the data stream and fuses features from different dimensions to obtain basic features; S102, Based on the basic features, the weight coefficients of each dimension are calculated using the analytic hierarchy process (AHP), the standardized feature vector is evaluated, scores for each dimension are obtained, and the overall importance score is calculated. S103, based on the comprehensive importance score, a piecewise function mapping mechanism is used to map each score interval to different priority levels, and the display resource allocation ratio of each data stream is determined by the priority level; S104, set differentiated update frequencies according to priority levels, and determine the refresh timing of each data stream through a timestamp comparison mechanism; S105, dynamically adjust the size of the visualization interface of each data stream according to the display resource allocation ratio of each data stream to obtain a hierarchical interface layout scheme. S106. Establish a synchronous update scheduling queue based on the update frequency and interface layout scheme of each data stream, manage data stream update tasks of different priority levels, and trigger the interface refresh operation of the corresponding data stream according to the preset update frequency through a time polling mechanism. S107, Perform load balancing control based on the synchronous update scheduling queue, and dynamically adjust the actual update frequency of data streams at each priority level; S108 continuously tracks changes in the importance of each data stream. If abnormal data fluctuations are detected, the overall importance score is recalculated, and the weight coefficients and priority levels are updated based on the overall importance score.

2. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 1, characterized in that, In S101, multidimensional feature extraction of the data stream includes: Based on the predefined rule base, business labels are assigned to the data stream, and the label vector of the data stream is output. The label vector is [business label 1, business label 2, ..., business label m]. If the data in the data stream belongs to one of the business labels, the business label of that type is set to 1, otherwise it is set to 0. Calculate the rate of change of various types of data in the data stream, record the frequency of data anomalies within a period, set the weighting coefficients for the rate of change and the weighting coefficients for the frequency of anomalies, and use a weighted summation algorithm to obtain the dynamic characteristics of various types of data.

3. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 2, characterized in that, S101 also includes: If the data change exceeds a preset threshold, the anomaly probability is calculated based on the historical anomaly frequency, and anomaly detection results are generated by combining business tags. Based on the anomaly detection results, the K-means algorithm is used to cluster the basic features to generate a cluster center library of historical fault data. The Euclidean distance between the current basic feature and each cluster center representing a fault in the cluster center library is calculated, and then the minimum distance is compared with the threshold. If the minimum distance is less than the threshold, the equipment status is determined to be abnormal.

4. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 1, characterized in that, In S102, the weight coefficients for each dimension are calculated using the analytic hierarchy process (AHP), including: Construct a judgment matrix, where each element represents the relative importance of each data stream type; obtain the weight coefficients by solving for the largest eigenvalue and the corresponding eigenvector of the judgment matrix, and then normalizing the eigenvector; calculate the consistency ratio to determine whether the consistency requirements are met.

5. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 1, characterized in that, In S104, the timer for each data stream is determined by a timestamp comparison mechanism, including: Record the most recent update timestamp of each data stream; based on the preset refresh frequency, compare the current time with the sum of the most recent update timestamp and the refresh frequency. If the current time >= (last update time + frequency interval), it is determined to be a refresh point. S104 also includes: obtaining the state change of the data stream based on the refresh point to obtain the update requirement; using the update requirement, employing a scheduling algorithm to determine the actual refresh order of the data stream to obtain the execution plan; and verifying the execution plan using timestamps to determine the completion status of the refresh operation and obtain the update status.

6. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 1, characterized in that, In S105, the size of the visualization interface for each data stream is dynamically adjusted according to the display resource allocation ratio of each data stream to obtain a hierarchical interface layout scheme, including: A resource assessment algorithm is used to analyze the available capacity of display resources, calculate the screen resolution and pixel distribution of the display area, and obtain the display resource allocation ratio. Based on the display resource allocation ratio, a grid partitioning algorithm is applied to divide the interface into primary and secondary areas, determining the size and position of each area. If the number of high-priority data streams exceeds a preset threshold, they are allocated to the primary display area, and a color mapping algorithm is used to assign a striking color to the high-priority data streams, generating a prominent visual effect. Based on the characteristics of medium and low-priority data streams, they are allocated to secondary display areas, and the display effect is generated to be distinct from the primary display area by adjusting the transparency and grayscale values. A dynamic adjustment algorithm is used to monitor data stream changes in real time. If the data stream priority changes, the allocation ratio is recalculated, and the sizes of the primary and secondary display areas are updated to obtain an adaptive interface layout scheme. A hierarchical layout algorithm is used to integrate the display content of the primary and secondary display areas, generating an interface layout scheme that includes high-priority and medium-low-priority data streams. If the display effect of the interface layout scheme does not reach the preset visual hierarchy threshold, an iterative optimization algorithm is used to adjust the area boundaries and color contrast to obtain the final hierarchical interface layout scheme.

7. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 1, characterized in that, In S106, a synchronous update scheduling queue is established based on the update frequency of each data stream and the interface layout scheme, including: Priorities are assigned based on data flow priority levels to obtain an ordered task sequence; update tasks are extracted from the task sequence, and the current timestamp is obtained through a time polling mechanism to determine the task trigger time. A priority queue data structure is used to manage data stream update tasks with different priority levels, including: if the current timestamp matches the preset update frequency, the highest priority task is extracted from the priority queue and the data stream update operation is triggered; The interface refresh operation is triggered by the time polling mechanism according to the preset update frequency, including: obtaining the corresponding data stream content based on the triggered data stream update operation and generating the interface refresh command; updating the data stream display content through the interface refresh command and obtaining the updated interface state. S106 also includes: extracting refresh operation feedback from the updated interface state to determine whether the task level in the priority queue needs to be adjusted; if the task level is adjusted, updating the task sequence in the priority queue and generating a new task scheduling plan.

8. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 1, characterized in that, In S107, load balancing control is performed based on the synchronous update scheduling queue, including: The system acquires system CPU utilization and memory usage, calculates the difference between these values ​​and preset thresholds to obtain the resource usage status. If the resource usage status shows that CPU utilization exceeds the preset threshold, the update operation of low-priority data streams is paused, the paused data stream identifier is recorded, and an update pause list is generated. Based on the update pause list, the scheduling queue of low-priority data streams is adjusted, and system CPU resources are reallocated to obtain an optimized scheduling plan. The difference between memory usage and the preset threshold is compared to determine whether the cache size of medium-priority data streams needs to be reduced, and a cache adjustment instruction is generated. Based on the cache adjustment instruction, the cache size of medium-priority data streams is reduced, memory resources are released, and the updated cache allocation status is obtained. A dynamic adjustment mechanism is adopted, combining resource usage status and cache allocation status, to calculate the optimal update frequency of each level of data stream and generate frequency adjustment parameters. Through the frequency adjustment parameters, the execution frequency of each level of data stream in the scheduling queue is updated to obtain the final load balancing control scheme.

9. The method for real-time data stream visualization and synchronous update in the Internet of Things according to claim 1, characterized in that, In S108, the importance changes of each data stream are continuously tracked, including: Real-time data is acquired from various data streams through sensor networks, and the characteristics of the data streams are analyzed using stream processing techniques to obtain a comprehensive importance score for each data stream. If abnormal data fluctuations are detected, the overall importance score will be recalculated. This includes: if the dynamic characteristics of the data stream exceed a preset threshold, it will be judged as an abnormal fluctuation, triggering a recalculation of the overall importance score to obtain an updated overall importance score. The weight coefficients and priority levels are updated based on the changes in the overall importance score, including: adjusting the weight allocation table using a weighted average algorithm based on the updated overall importance score to determine the new weight coefficients; and updating the priority level identifiers using the adjusted weight coefficients to generate a new priority sequence. S108 also includes: using the updated priority sequence to reallocate system resources and obtain an optimized resource configuration scheme; adjusting the data stream processing order through the optimized resource configuration scheme to maintain the system's adaptive capability; and continuously monitoring system performance based on the changing trend of the data stream processing order to obtain real-time performance feedback data.

10. A real-time data stream visualization and synchronous update system for the Internet of Things, characterized in that, The system includes: The multidimensional feature extraction and fusion module is used to extract and fuse features from the input data stream and output a basic feature vector. The weight calculation and comprehensive scoring engine module is used to calculate the weights of each dimension based on the analytic hierarchy process, evaluate features, and generate a comprehensive importance score for the data stream. The priority mapping and resource allocation module is used to map scores to priority levels through a segmentation function and determine the display resource allocation ratio for each data stream. The update frequency control module is used to set differentiated update frequencies according to priority levels and to determine the refresh timing using a timestamp comparison mechanism. The dynamic interface layout engine module is used to dynamically adjust the size of visual components according to the resource allocation ratio to form a hierarchical interface layout. The synchronous update scheduler module is used to establish a priority scheduling queue and trigger interface refresh according to the update frequency through a time polling mechanism. The load balancing and dynamic adjustment module is used to monitor system load and dynamically adjust the actual update frequency of data streams of different priorities to ensure smoothness. The adaptive feedback and re-evaluation module is used to continuously monitor the data flow. When abnormal fluctuations are detected, it triggers the recalculation and update of weights, scores, priorities, resource allocation, and frequencies to achieve closed-loop optimization of the system.

Images