An Internet of Things device management system

Abstract

The present invention relates to the field of configuration management technology, specifically to an Internet of Things device management system, which includes a state recognition module, a fault diagnosis module, a network analysis module and a topology adjustment module. In the present invention, through real-time data collection and continuous analysis, the real-time response capability and adaptive adjustment capability of the system are enhanced, and the normal operation mode can be automatically updated and refined, so that device management is not limited to preset parameters, but can dynamically adapt to environmental changes. At the same time, automatic abnormal detection and rapid identification significantly improve the effectiveness and accuracy of fault diagnosis, continuous monitoring of network performance and automatic adjustment of network topology based on performance data enable the network to be optimized immediately when potential bottlenecks are found, avoiding the problem that performance bottlenecks affect system operation for a long time, not only improving the response speed of the network, but also improving the overall operating efficiency and stability, and effectively reducing maintenance costs.

Description

Technical Field

[0001] The present invention relates to the technical field of configuration management, and in particular to an Internet of Things device management system. Background Art

[0002] Configuration management is a key area in information technology and network management, involving monitoring, controlling and optimizing the configuration of various hardware and software resources in networks and systems, with the goal of ensuring that the configuration information of all systems and devices is up-to-date, consistent, and can quickly adapt to changing needs. It usually includes automated tools and processes for setting and maintaining system configuration standards to prevent unauthorized changes, while improving system stability and security, allowing technicians to monitor and adjust network behavior from a central location to ensure compatibility and optimal performance between devices and applications.

[0003] Among them, the IoT device management system is a system specially designed to monitor and manage various devices connected to the IoT network. Its main purpose is to optimize device performance and functions through remote monitoring, automated control and maintenance of device status. The IoT device management system enables users to configure devices from a central control point, monitor device health and safety, and perform troubleshooting and performance optimization, thereby improving device utilization and overall network efficiency, which is crucial to ensuring the stable operation of large-scale IoT deployments.

[0004] Existing IoT device management usually relies on static configuration management and manual updates, and lacks the ability to respond quickly to environmental changes. In a dynamic and large-scale IoT environment, static management methods cannot effectively cope with frequently changing network conditions and device status, resulting in management delays and untimely fault responses. At the same time, existing systems often fail to achieve real-time analysis and updates of device behavior patterns, making it impossible for the monitoring and control of device operation to accurately match the actual device status, which can easily lead to resource waste and potential security risks. In addition, the lack of an automated network performance adjustment mechanism means that the network cannot be adaptively adjusted in a timely manner when encountering performance bottlenecks, affecting the overall network efficiency and device performance. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose an Internet of Things device management system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: an Internet of Things device management system includes:

[0007] The state recognition module collects equipment operation data in real time through sensors, stores data streams, analyzes the stored data to determine the normal operation mode, analyzes the frequency and amplitude of each data point in the mode, compares the normal behavior to confirm the equipment behavior boundary, and generates equipment behavior characteristics;

[0008] The fault diagnosis module combines the real-time monitoring data with the equipment behavior characteristics, performs continuous comparison, detects abnormal data points, analyzes the difference between the data points and the normal data, identifies the abnormal pattern according to the significance of the difference, determines whether it matches the known fault type according to the characteristics and intensity of the abnormal pattern, and generates equipment fault warning information;

[0009] The network analysis module combines the device fault warning information to monitor network traffic and node activity status, analyze inter-node communication efficiency, detect inter-device data packet transmission speed and error rate, calculate total network traffic, evaluate network delay and data loss frequency, determine network operation efficiency, and generate network performance records;

[0010] Based on the network performance records, the topology adjustment module analyzes network traffic and node status, identifies bottleneck areas in data flows, replans network paths, adjusts the configuration of key nodes to match the current data requirements of the device, modifies connection weights and updates routing rules, responds to actual changes in network traffic, and obtains device configuration update records.

[0011] As a further solution of the present invention, the step of acquiring the device behavior characteristics is:

[0012] Initialize sensors, collect device operation data in real time, and perform data normalization.

[0013]

[0014] in, Represents the original data, and are the minimum and maximum values ​​in the data set, respectively;

[0015] Perform statistical analysis on the normalized equipment operation data, calculate the mean value, variance and frequency distribution of the parameters, and compare them with historical data to determine the normal operation mode.

[0016]

[0017] in, A comprehensive score representing the normal operating mode of the device. is the normalized value of the parameter, is the weight of the corresponding parameter, is the total number of parameters;

[0018] Apply Fourier transform to each data point to decompose the data time series into frequency domain and calculate the amplitude of each frequency.

[0019]

[0020] in, is the frequency amplitude value, is the frequency The complex result after Fourier transformation is, and are the real and imaginary parts of the result;

[0021] Compare and analyze deviations from normal patterns to identify device behavior characteristics.

[0022]

[0023] in, is the difference score, is the measured value of the current data point, is the average value of the normal mode, is the threshold for deviation from normal behavior.

[0024] As a further solution of the present invention, the step of detecting abnormal data points is:

[0025] Collect real-time monitoring data of the equipment. Let the data set be D, which represents the monitoring indicators of the equipment at each moment, including temperature, pressure and current;

[0026] Extracting a normal feature data set N of device behavior from the device behavior features, representing the behavior mode of the device in a non-fault state;

[0027] Compare each data point in the real-time dataset D with the normal feature dataset N and calculate the significance of the difference.

[0028]

[0029] in, A measure of the significance of the difference, is the i-th monitoring indicator value in the real-time data point, is the average value of the i-th indicator in the normal feature data set, is the weight, reflecting the importance of the indicator in fault diagnosis;

[0030] A threshold T is set to determine whether the difference significance E exceeds the threshold. If so, the data point is marked as an abnormal data point and recorded as an abnormal data point set.

[0031] As a further solution of the present invention, the step of obtaining the equipment fault warning information is:

[0032] Perform cluster analysis on the abnormal data point set, identify the difference abnormal pattern, calculate each cluster center,

[0033]

[0034] in, are data points in the same abnormal mode, is the weight based on the degree of abnormality, is the total number of abnormal data points;

[0035] Compare each abnormal pattern with the known fault type library, perform similarity evaluation, and determine the degree of closeness between each abnormal pattern and the fault type.

[0036]

[0037] in, Score the similarity. , , , They represent the feature values ​​in the fault library, the feature values ​​of the abnormal patterns, the feature matching weights and the adjustment parameters respectively;

[0038] Select the fault type with the highest similarity for each abnormal pattern. If the similarity S exceeds the set threshold, the matching fault type is confirmed, otherwise it is marked as an unknown type.

[0039] All matching results are combined to generate fault warning information and output the fault warning result set R.

[0040] As a further solution of the present invention, the steps for calculating the total network flow rate are:

[0041] Monitor and record the data packet sending and receiving of all nodes in the network, including the sending start and receiving end time of each data packet, as well as the total number of data packets sent and received;

[0042] Calculate the data packet transmission speed and error rate between each pair of nodes.

[0043]

[0044] in, is the data packet transmission speed, is the preset distance between nodes, indicating the physical distance of transmission. It's the time difference. is the correction factor to match the network congestion, It is the inter-node stability coefficient based on historical data analysis.

[0045]

[0046] in, is the error rate, is the number of packets successfully received, is the number of packets sent, is the data integrity weighting factor;

[0047] Use data on packet transmission speed and error rate to calculate the efficiency of inter-node communication.

[0048]

[0049] in, is the communication efficiency, is the adjustment factor for the effect of error;

[0050] The data of all node pairs are combined to calculate the total flow of the entire network.

[0051]

[0052] in, It is a normalization factor for the amount of data and provides a comprehensive evaluation of network performance.

[0053] As a further solution of the present invention, the steps of obtaining the network performance record are:

[0054] Collect communication delay data between nodes in the network, including the complete time from sending to receiving each data packet, and synchronously monitor and record the reception status of all data packets to identify lost data packets;

[0055] Calculate the average delay time and packet loss ratio for each node pair;

[0056] Analyze latency and data loss across the network,

[0057]

[0058] in, represents the comprehensive network performance index, Is a node To Node The average delay time, is the communication frequency weight of the corresponding node pair, Is a node To Node Number of packets lost, is the number of packets sent, It is a tuning parameter used to balance the effects of delay and packet loss rate in the formula;

[0059] Comprehensive performance index Conduct assessments to reveal the network's operational efficiency and data integrity, and obtain a network performance record.

[0060] As a further solution of the present invention, the network path replanning step is:

[0061] Based on the network performance records, extract the flow data and status information of each node in the network, and calculate the average flow and status indicators of each node;

[0062] Calculate the performance index of each node and mark the bottleneck node, calculate the traffic density for each path in the network,

[0063]

[0064]

[0065] in, is the performance index of node i, is the traffic data at a point in time, is the time point weight, is the time point reliability factor, is the traffic density of path j, is the bottleneck node flow, is the contribution of the node on the path, is the path sensitivity coefficient, , are the scaling factor and offset factor of the path density, respectively. , They represent the monitoring time point and the number of nodes on the path respectively;

[0066] Analyze the marked bottleneck nodes, select high-density traffic paths, and perform priority replanning;

[0067] According to the traffic density and node performance indicators, the network path is reconfigured to optimize the network data transmission efficiency and obtain the replanned path.

[0068] As a further solution of the present invention, the steps of obtaining the device configuration update record are:

[0069] According to the re-planned network path and node performance indicators, a new weight is set for each connection, and the new weight of the connection is calculated:

[0070]

[0071] in, represents the new connection weight, is the original connection weight, Indicates the change in node performance, is the maximum value of all node performance indicators. and is the weight adjustment coefficient, is the node's payload, is the maximum load monitored;

[0072] According to the new weights, updating routing rules in the network;

[0073] The new weights and routing rules are recorded to form a device configuration update record.

[0074] Compared with the prior art, the advantages and positive effects of the present invention are:

[0075] In the present invention, through real-time data collection and continuous analysis, the real-time response capability and adaptive adjustment capability of the system are enhanced, and the normal operation mode can be automatically updated and refined, so that equipment management is not limited to preset parameters, but can dynamically adapt to environmental changes. At the same time, automated anomaly detection and rapid identification significantly improve the effectiveness and accuracy of fault diagnosis. Continuous monitoring of network performance and automatic adjustment of network topology based on performance data enable the network to be optimized in real time when potential bottlenecks are found, avoiding the problem of performance bottlenecks affecting system operation for a long time. This not only improves the response speed of the network, but also improves the overall operating efficiency and stability, and effectively reduces maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 is a system flow chart of the present invention;

[0077] Figure 2 A flowchart for obtaining the behavioral characteristics of the device of the present invention;

[0078] Figure 3 is a flow chart for detecting abnormal data points of the present invention;

[0079] Figure 4 A flowchart for obtaining equipment fault warning information of the present invention;

[0080] Figure 5 The figure is a flow chart for calculating the total network flow rate of the present invention;

[0081] Figure 6 A flowchart for obtaining network performance records of the present invention;

[0082] Figure 7 A flowchart of network path replanning of the present invention;

[0083] Figure 8 A flowchart for obtaining update records for the device configuration of the present invention. DETAILED DESCRIPTION

[0084] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0085] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0086] Embodiment 1

[0087] See also Figure 1 , an Internet of Things device management system includes:

[0088] The state recognition module collects equipment operation data in real time through sensors, stores data streams, analyzes the stored data to determine the normal operation mode, analyzes the frequency and amplitude of each data point in the mode, compares the normal behavior to confirm the equipment behavior boundary, and generates equipment behavior characteristics;

[0089] The fault diagnosis module combines real-time monitoring data with continuous comparison of equipment behavior characteristics, detects abnormal data points, analyzes the differences between data points and normal data, identifies abnormal patterns based on the significance of the differences, determines whether the abnormal patterns match known fault types based on their characteristics and intensity, and generates equipment fault warning information;

[0090] The network analysis module combines equipment failure warning information to monitor network traffic and node activity status, analyze inter-node communication efficiency, detect data packet transmission speed and error rate between devices, calculate total network traffic, evaluate network delay and data loss frequency, determine network operation efficiency, and generate network performance records;

[0091] Based on the network performance records, the topology adjustment module analyzes the network traffic and node status, identifies the bottleneck areas in the data flow, and replans the network path. It adjusts the configuration of key nodes to match the current data requirements of the device, modifies the connection weights and updates the routing rules, responds to the actual changes in the network traffic, and obtains the device configuration update records.

[0092] Equipment behavior characteristics include frequency characteristics, amplitude characteristics and equipment behavior boundaries. Equipment fault warning information includes abnormal data identification results, fault pattern identification results and fault type matching results. Network performance records include delay evaluation results, packet loss rate and error frequency. Equipment configuration update records include node reconfiguration results, path optimization results and weight adjustment records.

[0093] See also Figure 2 , the steps to obtain device behavior characteristics are:

[0094] Initialize sensors, collect device operation data in real time, and perform data normalization.

[0095]

[0096] in, Represents the original data, and are the minimum and maximum values ​​in the data set, respectively;

[0097] Perform statistical analysis on the normalized equipment operation data, calculate the mean value, variance and frequency distribution of the parameters, and compare them with historical data to determine the normal operation mode.

[0098]

[0099] in, A comprehensive score representing the normal operating mode of the device. is the normalized value of the parameter, is the weight of the corresponding parameter, is the total number of parameters;

[0100] Apply Fourier transform to each data point to decompose the data time series into frequency domain and calculate the amplitude of each frequency.

[0101]

[0102] in, is the frequency amplitude value, is the frequency The complex result after Fourier transformation is, and are the real and imaginary parts of the result;

[0103] Compare and analyze deviations from normal patterns to identify device behavior characteristics.

[0104]

[0105] in, is the difference score, is the measured value of the current data point, is the average value of the normal mode, is the threshold for deviation from normal behavior.

[0106] Collection and preliminary processing of equipment operation data:

[0107] Data collection: Assuming the device has sensors , the sensor records data once a second, including voltage , Current and speed wait.

[0108] Data normalization: In order to process the data, it needs to be normalized to ensure that the data is compared on the same scale. (such as voltage), the normalized calculation process is as follows:

[0109] Calculation formula:

[0110]

[0111] in, and It is the minimum and maximum value of the parameter calculated within a certain time window (for example, every day), which can be calculated by simply traversing the data points.

[0112] Determination of normal operating mode:

[0113] Data feature statistics: Based on the normalized data, calculate the statistical characteristics of each parameter, such as the mean value and standard deviation .

[0114] Mode determination: Determine the normal operating mode of the device by comparing the statistical characteristics of real-time data with historical data. The comparison process involves multiple parameters, each of which There is a weight .

[0115] Calculation formula:

[0116]

[0117] in, is the normalized value of each parameter, The weight is given based on the importance and impact analysis of historical data.

[0118] Detailed analysis of frequency and amplitude:

[0119] Fourier transform: For each data point Apply Fourier transform to convert it from time domain to frequency domain, and then get .

[0120] Amplitude calculation: Calculate the amplitude of each frequency component.

[0121] Calculation formula:

[0122]

[0123] in, is the frequency The result after Fourier transformation is and Im are the real and imaginary parts of the result.

[0124] Generation of device behavior characteristics:

[0125] Behavior boundary judgment: Use the previous analysis results to determine whether the device is within the normal operation mode.

[0126] Signature Generation: All identified normal and abnormal features are recorded.

[0127] Calculation formula:

[0128]

[0129] if , then the behavior is considered abnormal. is the current measured value, is the average value in normal operation mode, It is a set threshold, which is set based on experience or statistical analysis to determine when behavioral deviations are significant enough to be considered abnormal.

[0130] See also Figure 3 , the detection steps of abnormal data points are:

[0131] Collect real-time monitoring data of the equipment. Let the data set be D, which represents the monitoring indicators of the equipment at each moment, including temperature, pressure and current;

[0132] Extracting a normal feature data set N of device behavior from the device behavior features, which represents the behavior mode of the device in a non-fault state;

[0133] Compare each data point in the real-time dataset D with the normal feature dataset N and calculate the significance of the difference.

[0134]

[0135] in, A measure of the significance of the difference, is the i-th monitoring indicator value in the real-time data point, is the average value of the i-th indicator in the normal feature data set, is the weight, reflecting the importance of the indicator in fault diagnosis;

[0136] Set the threshold T to determine whether the difference significance E exceeds the threshold. If it exceeds the threshold, mark the data point as an abnormal data point and record it as a set of abnormal data points.

[0137] Parameter definition and calculation process:

[0138] Monitoring index value :

[0139] Extracted directly from real-time monitoring data, assuming that at time point t, the data monitored by the device is ,These values ​​reflect the current operating status of the equipment, such as temperature, pressure, current, etc.;

[0140] Normal eigenvalue :

[0141] Calculated from the historical monitoring data of the equipment. For each monitoring indicator, calculate the average value of its historical data when there is no fault, that is:

[0142]

[0143] here represents the value of the i-th monitoring indicator at the j-th observation, and m is the total number of historical data points;

[0144] Weight :

[0145] Weight The indicator is set according to its importance in fault diagnosis. It can be assigned by expert experience or adjusted dynamically based on historical fault data analysis of the correlation between each indicator and the fault.

[0146] Calculate the significance of the difference :

[0147] Difference calculation: For each monitoring indicator, calculate the difference between its current value and the historical normal value.

[0148]

[0149] Weighted squared differences: Calculate the square of each difference and multiply it by the corresponding weight,

[0150]

[0151] Sum: sum all weighted squared differences,

[0152]

[0153] Square root: Take the square root of the sum to get the significance of the difference ,

[0154]

[0155] The calculation process will directly extract monitoring indicators from the real-time monitoring data of the equipment, use historical data to calculate normal characteristic values, and combine the weights set by experts or obtained by data analysis to calculate the degree of abnormality of each data point, which can effectively identify abnormal data points.

[0156] See also Figure 4 , the steps to obtain equipment fault warning information are:

[0157] Perform cluster analysis on the set of abnormal data points, identify the difference abnormal patterns, calculate each cluster center,

[0158]

[0159] in, are data points in the same abnormal mode, is the weight based on the degree of abnormality, is the total number of abnormal data points;

[0160] Compare each abnormal pattern with the known fault type library, perform similarity evaluation, and determine the degree of closeness between each abnormal pattern and the fault type.

[0161]

[0162] in, Score the similarity. , , , They represent the feature values ​​in the fault library, the feature values ​​of the abnormal patterns, the feature matching weights and the adjustment parameters respectively;

[0163] Select the fault type with the highest similarity for each abnormal pattern. If the similarity S exceeds the set threshold, the matching fault type is confirmed, otherwise it is marked as an unknown type.

[0164] All matching results are combined to generate fault warning information and output the fault warning result set R.

[0165] Cluster center calculation of abnormal patterns:

[0166] Collect a collection of abnormal data points , these data points are the outliers identified in the previous step by comparing with the normal data;

[0167] For each data point Assign a weight , the weight is determined based on the degree of anomaly, which can be evaluated by the absolute value or square of the difference between the data point and the normal data. Assuming the degree of anomaly,

[0168]

[0169] is the corresponding value of normal data, and ;

[0170] Then use the weighted average formula to calculate the cluster center , the values ​​of all data points Multiply by its corresponding weight , and then divided by the sum of the weights.

[0171] Similarity evaluation of fault type matching:

[0172] Defining a fault type library , where each fault type is represented by a feature set express;

[0173] For each abnormal pattern cluster center pass express;

[0174] Assign a weight to each feature ,The weight indicates the importance of the feature in fault identification.,For example, some features may be more diagnostically valuable than other features;

[0175] Use an exponential decay function Calculate the impact of each feature difference, where is the attenuation parameter, which controls the influence of feature differences on the total similarity;

[0176] Calculate features The similarity contribution of

[0177]

[0178] And sum up the similarity contributions of all features to get the total similarity .

[0179] See also Figure 5 , the calculation steps of the total network traffic rate are:

[0180] Monitor and record the data packet sending and receiving of all nodes in the network, including the sending start and receiving end time of each data packet, as well as the total number of data packets sent and received;

[0181] Calculate the data packet transmission speed and error rate between each pair of nodes.

[0182]

[0183] in, is the data packet transmission speed, is the preset distance between nodes, indicating the physical distance of transmission. It's the time difference. is the correction factor to match the network congestion, It is the inter-node stability coefficient based on historical data analysis.

[0184]

[0185] in, is the error rate, is the number of packets successfully received, is the number of packets sent, is the data integrity weighting factor;

[0186] Use data on packet transmission speed and error rate to calculate the efficiency of inter-node communication.

[0187]

[0188] in, is the communication efficiency, is the adjustment factor for the effect of error;

[0189] The data of all node pairs are combined to calculate the total flow of the entire network.

[0190]

[0191] in, It is a normalization factor for the amount of data and provides a comprehensive evaluation of network performance.

[0192] Calculate the packet transmission speed:

[0193] Assumptions:

[0194] node To Node Distance Unit distance (such as kilometers)

[0195] Transfer time Unit of time (such as seconds)

[0196] Correction Factor To adapt to network congestion

[0197] Inter-node stability coefficient

[0198] calculate When calculating the network congestion and the stability between nodes, we need to consider the network congestion and the stability between nodes.

[0199]

[0200] Calculate the error rate:

[0201] Assumptions:

[0202] node Successfully sent Packets

[0203] Successfully received Packets

[0204] Data integrity weighting factor

[0205] The error rate needs to reflect the loss of data packets. Substituting the value into the value,

[0206]

[0207] Calculate the communication efficiency between nodes:

[0208] Assumptions:

[0209] Error Impact Adjustment Factor

[0210] Communication efficiency takes into account the influence of speed and error rate. Substituting the values ​​into the formula, and have to,

[0211]

[0212] Calculate the total network traffic:

[0213] Assumptions:

[0214] Nodes

[0215] Normalization factor

[0216] The total traffic is the sum of the number of packets sent and received by all nodes. Assuming that each node sends and receives 1000 packets on average, we can get the following formula:

[0217]

[0218] See also Figure 6 , the steps to obtain network performance records are:

[0219] Collect communication delay data between nodes in the network, including the complete time from sending to receiving each data packet, and synchronously monitor and record the reception status of all data packets to identify lost data packets;

[0220] Calculate the average delay time and packet loss ratio for each node pair;

[0221] Analyze latency and data loss across the network,

[0222]

[0223] in, represents the comprehensive network performance index, Is a node To Node The average delay time, is the communication frequency weight of the node pair, Is a node To Node Number of packets lost, is the number of packets sent, It is a tuning parameter used to balance the effects of delay and packet loss rate in the formula;

[0224] Comprehensive performance index Conduct assessments to reveal the network's operational efficiency and data integrity, and obtain a network performance record.

[0225] Collect data:

[0226] For each pair of nodes Collection Delay and packet loss.

[0227] Assume there are three nodes with the following data:

[0228] ms, ms, ms

[0229] , ,

[0230] , ,

[0231] Calculating Packet Loss Rate , Count the number of lost packets and then calculate the packet loss rate for each pair of nodes,

[0232] , loss rate = 10 / 100 = 0.10

[0233] , loss rate = 5 / 150 = 0.033

[0234] , loss rate = 5 / 120 = 0.042

[0235] Calculate the communication frequency weight :

[0236] The weight is proportional to the amount of data packets sent, set:

[0237]

[0238]

[0239]

[0240] Comprehensive calculation G:

[0241]

[0242] See also Figure 7 , the network path replanning steps are:

[0243] Based on the network performance records, the traffic data and status information of each node in the network are extracted, and the average traffic and status indicators of each node are calculated;

[0244] Calculate the performance index of each node and mark the bottleneck node, calculate the traffic density for each path in the network,

[0245]

[0246]

[0247] in, is the performance index of node i, is the traffic data at a point in time, is the time point weight, is the time point reliability factor, is the traffic density of path j, is the bottleneck node flow, is the contribution of the node on the path, is the path sensitivity coefficient, , are the scaling factor and offset factor of the path density, respectively. , They represent the monitoring time point and the number of nodes on the path respectively;

[0248] Analyze the marked bottleneck nodes, select high-density traffic paths, and perform priority replanning;

[0249] According to the traffic density and node performance indicators, the network path is reconfigured to optimize the network data transmission efficiency and obtain the replanned path.

[0250] : Comprehensively reflects the data processing capacity of the node within a specific time;

[0251] : reflects the data transmission volume of the node at that moment;

[0252] : Adjust according to the peak traffic period to reflect the importance of different time periods;

[0253] : Considering the reliability of data transmission, the weights are adjusted to reflect the data quality;

[0254] : Taking into account the system baseline noise or background flow;

[0255] : reflects the concentration of traffic distribution on the path;

[0256] : Display the traffic size of the node;

[0257] : reflects the influence of the node on a specific path;

[0258] : reflects the sensitivity of the node to the path performance;

[0259] , : are the scaling factor and offset factor of the path density, respectively, used to correct and balance the path evaluation results;

[0260] , : Respectively represent the monitoring time point and the number of nodes on the path to ensure the comprehensiveness and accuracy of the analysis.

[0261] Calculation performance indicators :

[0262] The data of a node at three different time points were monitored, and the following data were obtained:

[0263] Mbps (traffic data at the first point in time)

[0264] Mbps (traffic data at the second time point)

[0265] Mbps (traffic data at the third time point)

[0266] The corresponding time point weight is:

[0267] (off-peak hours)

[0268] (peak hours)

[0269] (Transition period)

[0270] The reliability factor of the time point is obtained based on the data quality assessment:

[0271] (High data quality)

[0272] (low data quality, possibly due to network congestion)

[0273] (Data quality is very high)

[0274] First, calculate the weighted flow at each time point:

[0275]

[0276]

[0277]

[0278] Then calculate the average:

[0279]

[0280] Setting constant offset is 10, and the final performance index is:

[0281]

[0282] Calculating traffic density :

[0283] Assume that a path contains two nodes and the relevant data is as follows:

[0284]

[0285] The contribution of a node on a path is:

[0286]

[0287] Path sensitivity coefficient, based on the node's sensitivity to path performance evaluation:

[0288] (Highly sensitive)

[0289] (More sensitive)

[0290] Calculate the weighted flow that a node contributes to a path:

[0291]

[0292]

[0293] Calculate the average traffic density:

[0294]

[0295] Assumption adjustment factor and , the final path traffic density is:

[0296]

[0297] See also Figure 8 , the steps to obtain the device configuration update record are:

[0298] According to the re-planned network path and node performance indicators, a new weight is set for each connection, and the new weight of the connection is calculated:

[0299]

[0300] in, represents the new connection weight, is the original connection weight, Indicates the change in node performance, is the maximum value of all node performance indicators. and is the weight adjustment coefficient, is the node's payload, is the maximum load monitored;

[0301] Based on the new weights, the routing rules in the network are updated to ensure that data flows can preferentially pass through connections with high performance and high weights, and path optimization is achieved by adjusting the routing table.

[0302] The new weights and routing rules are recorded to form a device configuration update record.

[0303] The parameters are described as follows:

[0304] : New connection weight, weight adjusted according to node performance;

[0305] : Original connection weight, the weight set in the previous network configuration;

[0306] : The change in node performance, indicating the increase or decrease in node performance since the last evaluation;

[0307] : The maximum value of all node performance indicators, used to standardize the impact of performance changes;

[0308] and : Weight adjustment coefficient, Adjust the weight impact based on performance changes, Adjust the weight impact based on payload;

[0309] : The node's effective load, which reflects the amount of data currently processed by the node;

[0310] : Maximum load monitored, used to normalize the impact of load.

[0311] Parameter calculation and acquisition method:

[0312] Get :

[0313] Original connection weights It is directly read from the database of the network management system, assuming that the original weight is 5;

[0314] calculate and :

[0315] Assume that a node in the network has improved its performance by 10 units since the last check, and the best node performance is 200 units, then,

[0316]

[0317]

[0318] calculate and :

[0319] Assume that the node currently processes 100GB of data, and a node in the network processes 150GB at peak. Then,

[0320]

[0321]

[0322] choose and :

[0323] The coefficients are preset based on the historical performance data and management strategies of the network.

[0324]

[0325]

[0326] Substitute the values ​​into the formula to calculate:

[0327]

[0328] New connection weights can be obtained How is the original weight It is adjusted according to changes in node performance and load, ensuring that network performance is dynamically adjusted to adapt to changing data demands and node status.

[0329] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An Internet of Things device management system, characterized in that: The system comprises: The state recognition module collects equipment operation data in real time through sensors, stores data streams, analyzes the stored data to determine the normal operation mode, analyzes the frequency and amplitude of each data point in the mode, compares the normal behavior to confirm the equipment behavior boundary, and generates equipment behavior characteristics; The fault diagnosis module combines the real-time monitoring data with the equipment behavior characteristics, performs continuous comparison, detects abnormal data points, analyzes the difference between the data points and the normal data, identifies the abnormal pattern according to the significance of the difference, determines whether it matches the known fault type according to the characteristics and intensity of the abnormal pattern, and generates equipment fault warning information; The network analysis module combines the device fault warning information to monitor network traffic and node activity status, analyze inter-node communication efficiency, detect inter-device data packet transmission speed and error rate, calculate total network traffic, evaluate network delay and data loss frequency, determine network operation efficiency, and generate network performance records; The topology adjustment module analyzes network traffic and node status based on the network performance record, identifies bottleneck areas in the data flow, and replans the network path, adjusts the configuration of key nodes to match the current data requirements of the device, modifies the connection weight and updates the routing rules, responds to the actual changes in network traffic, and obtains device configuration update records; The steps of replanning the network path are: Based on the network performance records, extract the flow data and status information of each node in the network, and calculate the average flow and status indicators of each node; Calculate the performance index of each node and mark the bottleneck node, calculate the traffic density for each path in the network, in, is the performance index of node i, is the traffic data at a point in time, is the time point weight, is the time point reliability factor, is the traffic density of path j, is the bottleneck node flow, is the contribution of the node on the path, is the path sensitivity coefficient, , are the scaling factor and offset factor of the path density, respectively. , They represent the monitoring time point and the number of nodes on the path respectively; Analyze the marked bottleneck nodes, select high-density traffic paths, and perform priority replanning; Reconfigure the network path according to traffic density and node performance indicators, optimize network data transmission efficiency, and obtain a replanned path; The steps for obtaining the device configuration update record are: According to the re-planned network path and node performance indicators, a new weight is set for each connection, and the new weight of the connection is calculated: in, represents the new connection weight, is the original connection weight, Indicates the change in node performance, is the maximum value of all node performance indicators. and is the weight adjustment coefficient, is the node's payload, is the maximum load monitored; According to the new weights, updating routing rules in the network; The new weights and routing rules are recorded to form a device configuration update record.

2. The IoT device management system according to claim 1, characterized in that: The steps for acquiring the device behavior characteristics are as follows: Initialize sensors, collect device operation data in real time, and perform data normalization. in, Represents the original data, and are the minimum and maximum values ​​in the data set, respectively; Perform statistical analysis on the normalized equipment operation data, calculate the mean value, variance and frequency distribution of the parameters, and compare them with historical data to determine the normal operation mode. in, A comprehensive score representing the normal operating mode of the device. is the normalized value of the parameter, is the weight of the corresponding parameter, is the total number of parameters; Apply Fourier transform to each data point to decompose the data time series into frequency domain and calculate the amplitude of each frequency. in, is the frequency amplitude value, is the frequency The complex result after Fourier transformation is, and are the real and imaginary parts of the result; Compare and analyze deviations from normal patterns to identify device behavior characteristics. in, is the difference score, is the measured value of the current data point, is the average value of the normal mode, is the threshold for deviation from normal behavior.

3. The IoT device management system according to claim 1, characterized in that: The steps for detecting abnormal data points are: Collect real-time monitoring data of the equipment. Let the data set be D, which represents the monitoring indicators of the equipment at each moment, including temperature, pressure and current; Extracting a normal feature data set N of device behavior from the device behavior features, representing the behavior mode of the device in a non-fault state; Compare each data point in the real-time dataset D with the normal feature dataset N and calculate the significance of the difference. in, A measure of the significance of the difference, is the i-th monitoring indicator value in the real-time data point, is the average value of the i-th indicator in the normal feature data set, is the weight, reflecting the importance of the indicator in fault diagnosis; A threshold T is set to determine whether the difference significance E exceeds the threshold. If so, the data point is marked as an abnormal data point and recorded as an abnormal data point set.

4. The Internet of Things device management system according to claim 1, characterized in that: The steps for obtaining the equipment fault warning information are as follows: Perform cluster analysis on the abnormal data point set, identify the difference abnormal pattern, calculate each cluster center, in, are data points in the same abnormal mode, is the weight based on the degree of abnormality, is the total number of abnormal data points; Compare each abnormal pattern with the known fault type library, perform similarity evaluation, and determine the degree of closeness between each abnormal pattern and the fault type. in, Score the similarity. , , , They represent the feature values ​​in the fault library, the feature values ​​of the abnormal patterns, the feature matching weights and the adjustment parameters respectively; Select the fault type with the highest similarity for each abnormal pattern. If the similarity S exceeds the set threshold, the matching fault type is confirmed, otherwise it is marked as an unknown type. All matching results are combined to generate fault warning information and output the fault warning result set R.

5. The IoT device management system according to claim 1, characterized in that: The calculation steps of the total network flow rate are as follows: Monitor and record the data packet sending and receiving of all nodes in the network, including the sending start and receiving end time of each data packet, as well as the total number of data packets sent and received; Calculate the data packet transmission speed and error rate between each pair of nodes. in, is the data packet transmission speed, is the preset distance between nodes, indicating the physical distance of transmission. It's the time difference. is the correction factor to match the network congestion, It is the inter-node stability coefficient based on historical data analysis. in, is the error rate, is the number of packets successfully received, is the number of packets sent, is the data integrity weighting factor; Use data on packet transmission speed and error rate to calculate the efficiency of inter-node communication. in, is the communication efficiency, is the adjustment factor for the effect of error; The data of all node pairs are combined to calculate the total flow of the entire network. in, It is a normalization factor for the amount of data and provides a comprehensive evaluation of network performance.

6. The Internet of Things device management system according to claim 1, characterized in that: The steps of obtaining the network performance record are: Collect communication delay data between nodes in the network, including the complete time from sending to receiving each data packet, and synchronously monitor and record the reception status of all data packets to identify lost data packets; Calculate the average delay time and packet loss ratio for each node pair; Analyze latency and data loss across the network, in, represents the comprehensive network performance index, Is a node To Node The average delay time, is the communication frequency weight of the corresponding node pair, Is a node To Node Number of packets lost, is the number of packets sent, It is a tuning parameter used to balance the effects of delay and packet loss rate in the formula; Comprehensive performance index Conduct assessments to reveal the network's operational efficiency and data integrity, and obtain a network performance record.

Images