Application terminal state judgment method and system based on principal component and density clustering

Abstract

The invention relates to the technical field of Internet of Things equipment management, in particular to an application terminal state judgment method and system based on principal components and density clustering. Standardized object models of various devices in an Internet of Things application terminal are constructed, and terminal data samples are collected based on the standardized object models; performing principal component analysis on the collected data samples, extracting key feature dimensions, and generating feature vectors after dimension reduction; performing clustering analysis on the feature vectors after dimension reduction by adopting a density clustering algorithm, and identifying a group distribution rule of the equipment state data to obtain a clustering result; and dynamically generating an equipment state recognition rule base according to a clustering result, and judging the real-time working state of the application terminal of the Internet of Things based on the equipment state recognition rule base. Key feature values are extracted through principal component analysis, a group distribution rule of Internet of Things application terminal state data is mined in combination with a density clustering algorithm, an application terminal state recognition rule base is dynamically generated, and a basis is provided for subsequent related intelligent applications.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things device management, and in particular to a method and system for determining the state of an application terminal based on principal components and density clustering. Background Art

[0002] IoT applications widely involve the coordinated operation of multiple sensing devices and systems. These applications typically consist of a diverse array of devices, typically found in smart buildings and smart vehicles. In smart buildings, various sensors, control devices, and security equipment work closely together to maintain efficient building operation and the safety and comfort of residents.

[0003] With the in-depth development of IoT applications and the surge in the number of devices, modern IoT devices are no longer single-function modules, but now integrate multi-module, multi-functional service capabilities. However, device data is often multi-sourced and heterogeneous, lacking a systematic description of device attributes and behaviors, which poses a huge challenge to determining the operating status of IoT terminals.

[0004] Existing technical solutions in the IoT environment require the collection, analysis, and modeling of metrics for various device models. A corresponding threshold range is set for each model to identify the device's operating status.

[0005] The existing solution has the following disadvantages: 1. Limited monitoring dimensions: This technical solution primarily focuses on monitoring individual characteristic parameters (such as energy consumption). Changes in these non-monitored characteristic parameters can cause discrepancies between device status determination and actual conditions. This is because the solution fails to fully cover all parameters that may affect device status.

[0006] 2. Difficulty in Data Integration: Data integration is extremely challenging due to differences in data formats and communication protocols among devices of different manufacturers and models, and the lack of a unified standard. Furthermore, the reusability of analysis and modeling work for a single device model is extremely low. Whenever new devices or features are needed, the system often requires extensive custom development, which is both costly and time-consuming.

[0007] 3. Nonlinear state determination challenges: Current technical solutions rely on threshold ranges to determine the device's operating state. However, this approach fails when the device's operating state is not linearly separable based on the monitored characteristic parameters. This means that in some cases, the solution may not accurately reflect the device's true state.

[0008] Therefore, existing IoT device status determination technologies suffer from problems such as a single monitoring dimension, difficulty integrating data, and misjudgment of nonlinear states. Traditional methods rely on manually set thresholds and are difficult to adapt to complex scenarios involving heterogeneous device data from multiple sources. Summary of the Invention

[0009] In view of this, an object of the present invention is to provide a method and system for determining the status of an application terminal based on principal components and density clustering, so as to solve the problems pointed out in the above background technology.

[0010] In order to achieve the above object, the technical solution adopted by the present invention is: The method for determining the state of an application terminal based on principal components and density clustering includes the following steps: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; S4. Dynamically generate a device status identification rule base according to the clustering results, and determine the real-time working status of the Internet of Things application terminal based on the device status identification rule base.

[0011] Furthermore, the construction of the standardized substance model in step S1 includes: Define the name, data type, unit, and value range of device attributes; Define the service name, delivery parameters and response parameters of the device command; Define the event name and output parameters of device events, and implement standardized integration of cross-vendor device data based on unified semantic descriptions.

[0012] Furthermore, the principal component analysis in step S2 includes: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

[0013] Furthermore, the parameter determination of the density clustering algorithm in step S3 includes: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

[0014] Furthermore, the dynamically generated device status identification rule base in step S4 includes: S41, calculating the silhouette coefficient of the clustering result, and generating a state determination rule according to the silhouette coefficient; S42: Storing the status determination rule in a database for real-time working status determination of the Internet of Things application terminal.

[0015] Furthermore, data preprocessing steps are also included: The collected data are then filled with missing values, processed for outliers and standardized.

[0016] Furthermore, the device status includes a normal state, an abnormal state, and a transition state, and the status determination rule is dynamically adjusted based on the density distribution of the cluster groups.

[0017] The application terminal state determination system based on principal component and density clustering includes a processor and a memory, wherein the memory stores a program or instruction. When the program or instruction is executed by the processor, the following steps are implemented: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; S4. Dynamically generate a device status identification rule base according to the clustering results, and determine the real-time working status of the Internet of Things application terminal based on the device status identification rule base.

[0018] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

[0019] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

[0020] The beneficial effects of the present invention are: The present invention provides a method and system for determining the status of an application terminal based on principal component analysis and density clustering. The method constructs standardized object models for various types of devices in an IoT application terminal, including device attributes, commands, and events. Terminal data samples are collected based on the standardized object models. Principal component analysis is performed on the collected data samples to extract key feature dimensions and generate reduced-dimensionality feature vectors. A density clustering algorithm is then used to cluster the reduced-dimensional feature vectors to identify the group distribution patterns of the device status data and obtain clustering results. A device status identification rule base is dynamically generated based on the clustering results, and the real-time operating status of the IoT application terminal is determined based on the device status identification rule base. At the data layer, multi-dimensional indicators such as runtime attributes and event parameters of different IoT devices are collected. Using machine learning, principal component analysis (PCA) is used to extract key feature values ​​through training with a large amount of sample data. The density clustering algorithm is then used to explore the group distribution patterns of IoT application terminal status data. A dynamic application terminal status identification rule base is generated to determine the operating status of the application terminal and calculate its usage rate, providing a foundation for subsequent related intelligent applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 FIG2 is a flowchart of the steps of the method for determining the state of an application terminal based on principal components and density clustering of the present invention; Figure 2 FIG2 is a structural block diagram of the application terminal state determination system based on principal component and density clustering of the present invention; Description of Figure Numbers: 1-Processor; 2-Memory. DETAILED DESCRIPTION

[0022] The present invention will be further described below with reference to the accompanying drawings and specific embodiments: like Figure 1 As shown, the application terminal status determination method based on principal component and density clustering provided by the present invention includes the following steps: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; S4. Dynamically generate a device status identification rule base according to the clustering results, and determine the real-time working status of the Internet of Things application terminal based on the device status identification rule base.

[0023] From the above description, it can be seen that the present invention has the following beneficial effects: The present invention provides a method for determining the status of an application terminal based on principal component analysis and density clustering. This method constructs standardized object models for various types of devices in an IoT application terminal, including device attributes, commands, and events. Terminal data samples are collected based on the standardized object models. Principal component analysis is performed on the collected data samples to extract key feature dimensions and generate reduced-dimensionality feature vectors. A density clustering algorithm is then used to cluster the reduced-dimensional feature vectors to identify the group distribution patterns of the device status data and obtain clustering results. A device status identification rule base is dynamically generated based on the clustering results, and the real-time operating status of the IoT application terminal is determined based on the device status identification rule base. At the data layer, multi-dimensional indicators such as runtime attributes and event parameters of different IoT devices are collected. Using machine learning, principal component analysis (PCA) is used to extract key feature values ​​through training with a large amount of sample data. A density clustering algorithm is then used to explore the group distribution patterns of IoT application terminal status data. This dynamically generates an application terminal status identification rule base for determining the operating status of the application terminal and calculating its usage rate, providing a foundation for subsequent related intelligent applications.

[0024] Furthermore, the construction of the standardized substance model in step S1 includes: Define the name, data type, unit, and value range of device attributes; Define the service name, delivery parameters and response parameters of the device command; Define the event name and output parameters of device events, and implement standardized integration of cross-vendor device data based on unified semantic descriptions.

[0025] From the above description, it can be seen that the construction of a standardized object model is achieved through the above steps.

[0026] Furthermore, the principal component analysis in step S2 includes: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

[0027] From the above description, it can be seen that through the above steps, principal component analysis is performed on the collected data samples to extract key feature dimensions.

[0028] Furthermore, the parameter determination of the density clustering algorithm in step S3 includes: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

[0029] From the above description, it can be seen that the parameters of the density clustering algorithm are determined through the above steps.

[0030] Furthermore, the dynamically generated device status identification rule base in step S4 includes: S41, calculating the silhouette coefficient of the clustering result, and generating a state determination rule according to the silhouette coefficient; S42: Storing the status determination rule in a database for real-time working status determination of the Internet of Things application terminal.

[0031] From the above description, it can be seen that through the above steps, a device status identification rule base can be dynamically generated.

[0032] Furthermore, data preprocessing steps are also included: The collected data are then filled with missing values, processed for outliers and standardized.

[0033] From the above description, it can be seen that through the above steps, the collected data is preprocessed to improve the accuracy of subsequent data processing.

[0034] Furthermore, the device status includes a normal state, an abnormal state, and a transition state, and the status determination rule is dynamically adjusted based on the density distribution of the cluster groups.

[0035] See Figure 2 The present invention also provides an application terminal state determination system based on principal component and density clustering, comprising a processor 1 and a memory 2, wherein the memory 2 stores a program or instruction, and when the program or instruction is executed by the processor 1, the following steps are implemented: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; S4. Dynamically generate a device status identification rule base according to the clustering results, and determine the real-time working status of the Internet of Things application terminal based on the device status identification rule base.

[0036] From the above description, it can be seen that the present invention has the following beneficial effects: The present invention provides a system for determining the status of an application terminal based on principal component analysis and density clustering. This system constructs standardized object models for various types of devices in an IoT application terminal, including device attributes, commands, and events. Terminal data samples are collected based on these standardized object models. Principal component analysis is performed on the collected data samples to extract key feature dimensions and generate reduced-dimensionality feature vectors. A density clustering algorithm is then used to cluster the reduced-dimensional feature vectors to identify the group distribution patterns of the device status data and obtain clustering results. A device status identification rule base is dynamically generated based on the clustering results, and the real-time operating status of the IoT application terminal is determined based on this device status identification rule base. At the data layer, multi-dimensional indicators such as runtime attributes and event parameters of different IoT devices are collected. Using machine learning, principal component analysis (PCA) is used to extract key feature values ​​through training with a large amount of sample data. A density clustering algorithm is then used to explore the group distribution patterns of IoT application terminal status data. This dynamically generates an application terminal status identification rule base for determining the operating status of the application terminal and calculating its usage rate, providing a foundation for subsequent related intelligent applications.

[0037] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the construction of the standardized object model in step S1 includes: Define the name, data type, unit, and value range of device attributes; Define the service name, delivery parameters and response parameters of the device command; Define the event name and output parameters of device events, and implement standardized integration of cross-vendor device data based on unified semantic descriptions.

[0038] From the above description, it can be seen that the construction of a standardized object model is achieved through the above steps.

[0039] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the principal component analysis in step S2 includes: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

[0040] From the above description, it can be seen that through the above steps, principal component analysis is performed on the collected data samples to extract key feature dimensions.

[0041] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the parameter determination of the density clustering algorithm in step S3 includes: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

[0042] From the above description, it can be seen that the parameters of the density clustering algorithm are determined through the above steps.

[0043] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the dynamically generated device state identification rule base in step S4 includes: S41, calculating the silhouette coefficient of the clustering result, and generating a state determination rule according to the silhouette coefficient; S42: Storing the status determination rule in a database for real-time working status determination of the Internet of Things application terminal.

[0044] From the above description, it can be seen that through the above steps, a device status identification rule base can be dynamically generated.

[0045] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: a data preprocessing step is also included: The collected data are then filled with missing values, processed for outliers and standardized.

[0046] From the above description, it can be seen that through the above steps, the collected data is preprocessed to improve the accuracy of subsequent data processing.

[0047] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the device state includes a normal state, an abnormal state, and a transition state, and the state determination rule is dynamically adjusted based on the density distribution of the cluster groups.

[0048] Several preferred embodiments or application examples are listed below to help those skilled in the art better understand the technical content of the present invention and the technical contribution made by the present invention relative to the prior art: Preferred embodiment one: like Figure 1As shown, the application terminal status determination method based on principal component and density clustering provided by the present invention includes the following steps: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. The construction of the standardization model in step S1 includes: Define the name, data type, unit, and value range of device attributes; Define the service name, delivery parameters and response parameters of the device command; Define the event name and output parameters of device events, and implement standardized integration of cross-vendor device data based on unified semantic descriptions.

[0049] In this embodiment, when analyzing the functions and characteristics of various devices within an IoT application terminal, it is important to focus on the device's real-time status (such as temperature, humidity, and state), the controls executed, and the information it actively reports. Based on this information, the basic components of the object model can be determined, including properties, commands, and events. Properties: These describe the static characteristics of a device and its real-time status. Commands: These represent the control operations a device can perform. Events: These describe the status or information actively reported by a device.

[0050] Device attribute definitions include the attribute name, data type (e.g., integer, floating point, string), unit, and value range. Device control definitions include the service name, delivery and response parameters, as well as the parameter name and data type. Device event definitions include the event name and output parameters. For example, the smokeAlarm event for a smoke detector includes output parameters such as smoke concentration and duration.

[0051] Based on a standardized object model, the IoT platform collects characteristic data, including status attributes, event parameters, delivery parameters, and response parameters, for various types of devices connected to application terminals. This allows the IoT platform to more effectively manage and utilize data from various devices.

[0052] S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; Furthermore, the principal component analysis in step S2 includes: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

[0053] In order to effectively process these characteristic data of application terminals, the principal component analysis (PCA) algorithm is used to reduce the dimension and simplify the data. The significance of this process is: Eliminate irrelevant noise: Remove noise data that has no significant impact on the sample and reduce data complexity.

[0054] Simplify data usage: Make the dataset easier to manage and display, and highlight the main characteristics of the sample. Reduce computing costs: Reduce the computational burden of subsequent algorithms and improve processing efficiency.

[0055] The following content explains how to implement principal component analysis and achieve dimensionality reduction on the device dataset.

[0056] Convert the data samples collected by the IoT application terminal (including: status attribute values ​​of services under each device, reporting event parameters, sending and response parameters, etc.) into feature vectors , where each eigenvalue corresponds to a spatial element in the vector Assuming there are data samples from M IoT application terminals, the feature data set can be represented as an M*N matrix.

[0057] Principal component analysis (PCA) centers the features of a dataset, selecting the direction with the largest variance as the first principal component direction and the orthogonal direction with the next largest variance as the second principal component direction. By rotating and scaling the coordinate axes, the difference between each dimension's data and the new coordinate axes is calculated. The data is sorted from largest to smallest based on the degree of variance, and the top N dimensions with the largest variance are selected as the final result.

[0058] S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; Furthermore, the parameter determination of the density clustering algorithm in step S3 includes: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

[0059] In this embodiment, the implementation process of applying the terminal status data group characteristics through density cluster analysis is as follows: After principal component analysis, the data samples of the application terminal are used to extract the dimensions with effective characteristics accounting for 95% of the entire data set. The filtered samples are then subjected to data preprocessing, including filling missing values, outlier processing, and standardization.

[0060] Then, the KNN algorithm is used to obtain the algorithm parameters of density clustering (DBSCAN), including: neighborhood radius (Epsilon), that is, the maximum distance between two points considered as neighbors; and the minimum number of points required to form a dense area (MinPts).

[0061] Perform clustering using the selected density clustering algorithm (DBSCAN). For each unvisited data point, if it has not yet been assigned to any cluster, check whether there are at least MinPts points around it within a distance of Epsilon. If the conditions are met, create a new cluster and add the point and all its reachable points to this cluster. Repeat the above process until all data sample points have been visited.

[0062] Analyze the cluster structure obtained by the density clustering algorithm (DBSCAN) and identify the characteristics of different groups: Core points: are inside the cluster and have enough points around them.

[0063] Boundary point: Although it belongs to a cluster, its surroundings do not meet the conditions for becoming a core point.

[0064] Noise points: isolated points that are neither core points nor boundary points of any cluster.

[0065] S4. Dynamically generate a device state identification rule base based on the clustering results, and determine the real-time working state of the IoT application terminal based on the device state identification rule base. The device state includes normal state, abnormal state, and transition state, and the state determination rules are dynamically adjusted based on the density distribution of the cluster groups.

[0066] Furthermore, the dynamically generated device status identification rule base in step S4 includes: S41, calculating the silhouette coefficient of the clustering result, and generating a state determination rule according to the silhouette coefficient; S42: Storing the status determination rule in a database for real-time working status determination of the Internet of Things application terminal.

[0067] Using the results of density clustering, we can delineate the silhouette coefficients of the application terminal's operating status and store these trained silhouette coefficients in a database. When receiving characteristic data from each device on an IoT terminal, the application service uses physical model metrics such as status attribute values, reported event parameters, and dispatch and response parameters to derive the principal component eigenvectors of the data. It then calculates the clustering of the vectors and determines the operating status corresponding to the coefficient silhouette, thereby determining the current operating status.

[0068] Furthermore, data preprocessing steps are also included: The collected data are then filled with missing values, processed for outliers and standardized.

[0069] Preferred embodiment 2: See Figure 2 The present invention also provides an application terminal state determination system based on principal component and density clustering, comprising a processor 1 and a memory 2, wherein the memory 2 stores a program or instruction, and when the program or instruction is executed by the processor 1, the following steps are implemented: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; S4. Dynamically generate a device status identification rule base according to the clustering results, and determine the real-time working status of the Internet of Things application terminal based on the device status identification rule base.

[0070] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the construction of the standardized object model in step S1 includes: Define the name, data type, unit, and value range of device attributes; Define the service name, delivery parameters and response parameters of the device command; Define the event name and output parameters of device events, and implement standardized integration of cross-vendor device data based on unified semantic descriptions.

[0071] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the principal component analysis in step S2 includes: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

[0072] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the parameter determination of the density clustering algorithm in step S3 includes: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

[0073] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the dynamically generated device state identification rule base in step S4 includes: S41, calculating the silhouette coefficient of the clustering result, and generating a state determination rule according to the silhouette coefficient; S42: Storing the status determination rule in a database for real-time working status determination of the Internet of Things application terminal.

[0074] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: a data preprocessing step is also included: The collected data are then filled with missing values, processed for outliers and standardized.

[0075] Furthermore, when the program or instruction is executed by the processor, the following steps are specifically implemented: the device state includes a normal state, an abnormal state, and a transition state, and the state determination rule is dynamically adjusted based on the density distribution of the cluster groups.

[0076] The present invention has been described with reference to the above embodiments and accompanying drawings. However, the above embodiments are merely exemplary embodiments of the present invention. It should be noted that the disclosed embodiments do not limit the scope of the present invention. On the contrary, modifications and equivalents falling within the spirit and scope of the claims are intended to be within the scope of the present invention.

Claims

1. The application terminal status determination method based on principal component and density clustering is characterized by: The following steps are involved: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; S4. Dynamically generate a device status identification rule base according to the clustering results, and determine the real-time working status of the Internet of Things application terminal based on the device status identification rule base.

2. The method for determining the state of an application terminal based on principal component and density clustering according to claim 1, characterized in that: The construction of the standardization model in step S1 includes: Define the name, data type, unit, and value range of device attributes; Define the service name, delivery parameters and response parameters of the device command; Define the event name and output parameters of device events, and implement standardized integration of cross-vendor device data based on unified semantic descriptions.

3. The method for determining the state of an application terminal based on principal component and density clustering according to claim 1, characterized in that: The principal component analysis in step S2 includes: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

4. The method for determining the state of an application terminal based on principal component and density clustering according to claim 1, characterized in that: The parameter determination of the density clustering algorithm in step S3 includes: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

5. The method for determining the state of an application terminal based on principal component and density clustering according to claim 1, characterized in that: The dynamically generated device status identification rule base in step S4 includes: S41, calculating the silhouette coefficient of the clustering result, and generating a state determination rule according to the silhouette coefficient; S42: Storing the status determination rule in a database for real-time working status determination of the Internet of Things application terminal.

6. The method for determining the state of an application terminal based on principal component and density clustering according to claim 1, characterized in that: It also includes data preprocessing steps: The collected data are then filled with missing values, processed for outliers and standardized.

7. The method for determining the state of an application terminal based on principal component and density clustering according to claim 1, characterized in that: The device status includes a normal state, an abnormal state and a transition state, and the status determination rule is dynamically adjusted based on the density distribution of the cluster groups.

8. The application terminal status determination system based on principal component and density clustering is characterized by: The system comprises a processor and a memory, wherein the memory stores a program or instruction, and when the program or instruction is executed by the processor, the following steps are implemented: S1. Build standardized object models for various types of devices in IoT application terminals, including device attributes, commands, and events, and collect terminal data samples based on the standardized object models. S2. Perform principal component analysis on the collected data samples, extract key feature dimensions, and generate feature vectors after dimensionality reduction; S3. Use density clustering algorithm to perform cluster analysis on the feature vector after dimensionality reduction, identify the group distribution pattern of device status data, and obtain clustering results; S4. Dynamically generate a device status identification rule base according to the clustering results, and determine the real-time working status of the Internet of Things application terminal based on the device status identification rule base.

9. The application terminal status determination system based on principal component and density clustering according to claim 8, characterized in that: When the program or instruction is executed by the processor, the following steps are specifically implemented: S21. Centralize the data samples; S22. Calculate the covariance matrix and determine the principal component direction based on the data sample after the centralization process; S23. Select the top N principal components whose cumulative variance contribution rates reach a preset threshold as key feature dimensions.

10. The application terminal status determination system based on principal component and density clustering according to claim 8, characterized in that: When the program or instruction is executed by the processor, the following steps are specifically implemented: S31. Obtaining density clustering algorithm parameters using a KNN algorithm, where the algorithm parameters include a neighborhood radius and a minimum number of points required to form a dense area; S32. Clustering the data samples according to the algorithm parameters to distinguish core points, boundary points and noise points.

Images