Artificial intelligence monitoring system based on Internet of Things

Abstract

The invention discloses an artificial intelligence monitoring system based on the Internet of Things, and the system comprises a device access and protocol adaptation module which is responsible for enabling Internet of Things devices produced by different manufacturers to access the system, recognizing and analyzing a communication protocol of each Internet of Things device, converting the communication protocol into a unified communication format in the system, and transmitting the communication protocol to a server; the converted equipment data is transmitted to the data communication and transmission module; the data communication and transmission module is responsible for transmitting and exchanging equipment data and transmitting the equipment data to the equipment management and monitoring module; according to the invention, through the equipment access and protocol adaptation module, the Internet of Things equipment produced by different manufacturers is accessed to the system, and the communication protocol of the Internet of Things equipment is identified and analyzed and converted into a unified communication format in the system. The function greatly improves the compatibility of the equipment, so that the equipment from different manufacturers can be seamlessly connected and interoperate in the system, and the flexibility and expandability of the system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent security monitoring, and in particular to an artificial intelligence monitoring system based on the Internet of Things. Background Art

[0002] The Internet of Things refers to a network that connects various objects to the Internet through information sensing devices such as radio frequency identification, infrared sensors, global positioning systems, laser scanners, etc., to exchange and communicate information in order to achieve intelligent identification, positioning, tracking, monitoring and management. The rapid development of the Internet of Things technology has provided a solid technical foundation for the artificial intelligence monitoring system based on the Internet of Things. The core advantage of the Internet of Things technology lies in its excellent connectivity, which can build and connect a huge resource database, providing a rich data source for intelligent monitoring. At the same time, the Internet of Things technology provides strong support for accurate identification, intelligent positioning and tracking by improving the stability and security of communication. Artificial intelligence is a branch of computer science that simulates human intelligence and aims to create intelligent machines that can understand, learn and make decisions autonomously. Artificial intelligence technology has made significant progress in recent years, especially in the fields of machine learning and deep learning.

[0003] After searching, the invention patent with Chinese patent number CN117834256A discloses a security monitoring system for IoT devices based on big data, which uses sensors and probes deployed on IoT devices to monitor and collect the working status and upload traffic value of IoT devices in real time, and introduces data processing and analysis algorithms based on big data and artificial intelligence technology at the back end to perform time series collaborative analysis of these IoT device data, so as to perform real-time monitoring and early warning of the IoT devices. In this way, abnormal behaviors of IoT devices can be discovered and handled in a timely manner, preventing devices from being maliciously attacked or tampered with, and ensuring the normal operation of IoT devices, thereby improving the security and reliability of IoT devices and protecting the privacy and property safety of users.

[0004] However, in actual use of the above system, since IoT devices produced by different manufacturers usually use different communication standards, the devices are difficult to be compatible with each other, which limits the flexibility of the system. Suppose a smart home system contains smart bulbs, smart sockets, smart door locks, smart cameras and other devices from different manufacturers. Since these devices use different communication standards, it is difficult to achieve seamless connection and interoperability between them, which reduces the user experience and limits the flexibility of the system. Therefore, an artificial intelligence monitoring system based on the Internet of Things is proposed. Summary of the invention

[0005] The purpose of the present invention is to solve the problem that the existing IoT devices produced by different manufacturers usually use different communication standards, resulting in compatibility between devices and limiting the flexibility of the system, and to propose an artificial intelligence monitoring system based on the Internet of Things.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] An artificial intelligence monitoring system based on the Internet of Things, comprising:

[0008] Device access and protocol adaptation module: responsible for connecting IoT devices produced by different manufacturers to the system, identifying and parsing the communication protocol of each IoT device, converting it into a unified communication format within the system, and passing the converted device data to the data communication and transmission module;

[0009] Data communication and transmission module: responsible for the transmission and exchange of equipment data, and transmits equipment data to the equipment management and monitoring module;

[0010] Equipment management and monitoring module: responsible for centralized management and monitoring of access system equipment, real-time acquisition of equipment operating status and parameter information, and real-time monitoring of equipment health status and operating information, issuing alarms when abnormalities occur, and passing equipment operating status and parameters to the artificial intelligence analysis and decision-making module;

[0011] Artificial intelligence analysis and decision-making module: responsible for forming a strong classifier by gradually building and combining decision trees, inputting the equipment operation status and parameters into the strong classifier, predicting whether the equipment is in an abnormal state, and when the prediction result is a fault or the prediction probability is lower than the set threshold, the equipment is considered to be in an abnormal state and an alarm is triggered. Based on the prediction results of abnormal behavior, decision suggestions are generated, and the prediction results and decision suggestions are passed to the user interface and interaction module;

[0012] User interface and interaction module: responsible for presenting analysis results and decision recommendations to users in an intuitive manner, and supporting users to remotely control and configure the equipment.

[0013] The above technical solution further includes:

[0014] Preferably, the device access and protocol adaptation module includes a device discovery and access unit, a device authentication and security management unit, a protocol identification and parsing unit, and a data transmission unit. The device discovery and access unit is responsible for identifying and discovering new devices in the system, automatically detecting IoT devices by scanning the network or other means, and connecting them to the platform. The device authentication and security management unit is responsible for verifying the identity of the access device, ensuring that the device is legal and meets the security requirements of the system, and also ensuring that the communication between devices is encrypted and secure. The protocol identification and parsing unit is responsible for identifying the communication protocol of each IoT device, having an in-depth understanding of the communication standards of different manufacturers, and being able to accurately identify the protocol type used by the device, and parsing the device communication protocol, converting it into the system's internal communication format, including parsing the structure of the data packet, extracting useful information, and converting it into a data format that the system can understand and process. The data transmission unit is responsible for passing the parsed and converted device data to the data communication and transmission module, including ensuring the accuracy, integrity and real-time nature of the data so that subsequent modules can correctly process and analyze the data.

[0015] Preferably, the data communication and transmission module includes a data receiving and formatting unit, a data transmission control unit, a data cache and buffer unit, a data verification and error handling unit and a communication protocol management unit. The data receiving and formatting unit is responsible for receiving the device data provided by the device access and protocol adaptation module and then performing formatting processing to ensure the integrity and accuracy of the data, and performing preliminary verification on the received data. According to the specific requirements within the system, the data is further formatted to ensure that the data format meets the standards for internal processing and storage of the system. The data transmission control unit is responsible for controlling the data transmission process, including selecting a transmission path, determining a transmission rate, managing transmission resources, etc. The data cache and buffer unit is responsible for providing data cache and buffer functions to cope with data transmission delays and jitters, and ensuring the continuity and stability of data during transmission. The data verification and error handling unit is responsible for verifying the transmission data, and when data errors or transmission anomalies are found, taking error handling measures such as retransmission, discarding or recording error logs, etc. The communication protocol management unit is responsible for managing the internal communication protocol of the system, including protocol registration, parsing, conversion, etc.

[0016] Preferably, the device management and monitoring module includes a device status and health monitoring unit, a device performance evaluation unit, a data acquisition and storage unit, and a historical data query and playback unit. The device status and health monitoring unit is responsible for acquiring and monitoring the working status, operating parameters, and health status of the device, including temperature, power consumption, working mode, etc., and analyzing whether the device has a fault, an abnormality, or performance degradation. The device performance evaluation unit is responsible for analyzing the performance data of the device, providing evaluation indicators, operating efficiency, power consumption, etc., to help optimize the system configuration. The data acquisition and storage unit is responsible for collecting various types of operating data of the device and storing them. The historical data query and playback unit is responsible for querying and playing back the historical operating data of the device to help analyze the long-term performance of the device.

[0017] Preferably, the artificial intelligence analysis and decision-making module includes a data preprocessing unit, a pattern recognition unit, an anomaly detection and early warning unit and a decision support unit. The data preprocessing unit is responsible for preprocessing the original data received from the equipment management and monitoring module to ensure data quality. The pattern recognition unit uses a gradient boosting tree to identify normal operating modes and abnormal behaviors of the equipment and distinguish between equipment failures and normal fluctuations. The anomaly detection and early warning unit is responsible for identifying abnormal conditions in equipment operation, such as failures, disconnections, etc., and issuing alarms in a timely manner. The decision support unit is responsible for generating decision recommendations based on data analysis results to guide users on how to adjust equipment configurations or respond to failures.

[0018] Preferably, the user interface and interaction module includes a device status display unit, a device control and configuration unit, an artificial intelligence analysis result display unit and a user personalization setting unit. The device status display unit is responsible for displaying the current status, parameters and health status of each device in a graphical interface, updating the operation status of the device in real time, and displaying the real-time data of the device in an intuitive manner. The device control and configuration unit is responsible for providing an interface for users to control the device, allowing users to remotely operate the device, such as turning on and off lights, adjusting socket power, adjusting door lock status, etc., and configuring device parameters, such as setting timer switches, adjusting modes, etc. The artificial intelligence analysis result display unit is responsible for displaying the analysis results and decision suggestions generated by the artificial intelligence analysis and decision module to the user, and helping the user understand the trend of device operation, possible risks and optimization suggestions through data visualization. The user personalization setting unit is responsible for providing personalized setting functions according to user preferences and habits, so that users can customize their own usage experience, such as improving efficiency through shortcuts, preset scenario modes, etc.

[0019] Preferably, the specific process of identifying the communication protocol of each IoT device, parsing the device communication protocol, and converting it into the system internal communication format is:

[0020] Protocol identification: Capture the data packets sent by IoT devices that attempt to access the system, identify the protocol type by analyzing the data packet feature fields or protocol data, including checking the port number, message format, specific protocol features, etc. of the data packet, and match the captured data packet features with the known communication protocols in the system. This is done by comparing the data packet features with the records in the protocol database. If the data packet features successfully match a known protocol, the communication protocol of the device is determined.

[0021] Protocol parsing: Based on the identified communication protocol, the protocol identification and parsing unit parses the structure of the data packet, which includes understanding the various fields of the data packet, the meaning of the fields, and the encoding method of the data. The data sent by the device, including the device status information, parameter information, event information, etc., are extracted from the parsed data packet, and the extracted device data is formatted according to the unified communication format within the system, including data re-encoding, field re-organization, data standardization, etc.

[0022] Preferably, the specific steps of using the gradient boosting tree to identify the normal operation mode and abnormal behavior of the device are:

[0023] Data preparation: Use the labeled equipment operation data set, including the normal operation data and fault data of the equipment. Suppose the training data set is {(X1,Y1),(X2,Y2),...,(X n ,Y n )}, where X i represents the feature vector, Y i Represent labels and preprocess data;

[0024] Model training: Initialize a decision tree and calculate its prediction error. Calculate the gradient of the loss function based on the prediction error, and gradually build multiple decision trees. Each decision tree corrects the prediction error of the previous decision tree.

[0025] Use boosting method to reduce the bias and variance of the model by gradually adjusting the sample weights and building a strong classifier. In each step, a new decision tree is added and its weight is adjusted to reduce the overall error in the combined strong classifier.

[0026] Model evaluation and optimization: Evaluate the performance of the model and adjust the model parameters such as the number of decision trees, learning rate, maximum depth, etc. according to the evaluation results to optimize the model performance;

[0027] Abnormal behavior mining and decision-making suggestions: The new equipment operating status and parameters are input into the trained gradient boosting tree model for prediction to determine whether the equipment is in an abnormal state. When the prediction result is a fault or the prediction probability is lower than the set threshold, the equipment is considered to be in an abnormal state and an alarm is triggered.

[0028] Preferably, the specific steps of the model training are:

[0029] Information gain calculation: Use information gain calculation to divide each node of the decision tree. For feature A, the information gain formula is: Among them, H(D) is the entropy of data set D, H(D v ) is the entropy of the subsets divided according to different values ​​of feature A, |D v | and |D| are subsets of D v and the size of the full set D;

[0030] Train the initial decision tree: select features through information gain to divide data and recursively construct a decision tree;

[0031] Data input: New data from the equipment management and monitoring module will be input into the decision tree model for prediction;

[0032] Apply the initial decision tree for prediction: Let the new data feature be X new The decision tree model selects features during training, classifies along the tree branches, and finally outputs the classification result X pred , its value is normal or fault;

[0033] Calculate the forecast error: calculate the difference between the predicted value and the actual value;

[0034] Calculate the gradient: Derivative the loss function with respect to the model parameters to obtain the gradient;

[0035] Update model parameters: Use the gradient descent method to update model parameters according to the calculated gradients;

[0036] Construct new weak classifier: Construct a new weak classifier using the updated parameters.

[0037] The present invention has the following beneficial effects:

[0038] 1. In the present invention, IoT devices produced by different manufacturers are connected to the system through the device access and protocol adaptation module, and their communication protocols are identified and parsed and converted into a unified communication format within the system. This function greatly improves the compatibility of devices, allowing devices from different manufacturers to seamlessly connect and interoperate in the system, significantly improving the flexibility and scalability of the system.

[0039] 2. In the present invention, the artificial intelligence analysis and decision-making module uses artificial intelligence technology to conduct in-depth analysis and mining of the equipment's operating status and parameters, and can identify the equipment's operating mode and abnormal behavior, and provide early warnings and decision-making suggestions. This not only helps users to discover and solve equipment problems in a timely manner, but also provides strong data support for system optimization and upgrades. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a system architecture diagram of an artificial intelligence monitoring system based on the Internet of Things proposed by the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] like Figure 1 As shown, an artificial intelligence monitoring system based on the Internet of Things includes:

[0043] Device access and protocol adaptation module: responsible for connecting IoT devices produced by different manufacturers to the system, identifying and parsing the communication protocol of each IoT device, converting it into a unified communication format within the system, and passing the converted device data to the data communication and transmission module;

[0044] Data communication and transmission module: responsible for the transmission and exchange of equipment data, and transmits equipment data to the equipment management and monitoring module;

[0045] Equipment management and monitoring module: responsible for centralized management and monitoring of access system equipment, real-time acquisition of equipment operating status and parameter information, and real-time monitoring of equipment health status and operating information, issuing alarms when abnormalities occur, and passing equipment operating status and parameters to the artificial intelligence analysis and decision-making module;

[0046] Artificial intelligence analysis and decision-making module: responsible for forming a strong classifier by gradually building and combining decision trees, inputting the equipment operation status and parameters into the strong classifier, predicting whether the equipment is in an abnormal state, and when the prediction result is a fault or the prediction probability is lower than the set threshold, the equipment is considered to be in an abnormal state and an alarm is triggered. Based on the prediction results of abnormal behavior, decision suggestions are generated, and the prediction results and decision suggestions are passed to the user interface and interaction module;

[0047] User interface and interaction module: responsible for presenting analysis results and decision recommendations to users in an intuitive manner, and supporting users to remotely control and configure the equipment.

[0048] In an embodiment of the present invention, by using a parser with built-in multiple communication protocols or adopting dynamic protocol recognition technology, the module can convert the data of different devices into a unified communication format within the system. In this way, even if the devices use different communication standards, unified data processing and transmission can be achieved in the system. Through protocol adaptation, devices such as smart bulbs, smart sockets, smart door locks and smart cameras in the smart home system can still communicate through a unified interface even if they use different communication standards, thereby achieving interoperability between them. After the device access and protocol adaptation modules complete the data format unification, the data communication and transmission module can use standard communication protocols (such as TCP / IP, MQTT, etc.) to achieve data transmission and exchange between devices. This ensures that data between different devices can be seamlessly connected and interoperable. The device management and monitoring module can monitor the status of all devices in real time to ensure that the system can automatically identify and process different data formats of different devices. Regardless of the communication protocol used by the device, the module can integrate the data of all devices and perform health checks and monitoring in a unified manner.

[0049] In one embodiment, the device access and protocol adaptation module includes a device discovery and access unit, a device authentication and security management unit, a protocol identification and parsing unit, and a data transmission unit. The device discovery and access unit is responsible for identifying and discovering new devices in the system, the device authentication and security management unit is responsible for verifying the identity of the access device, the protocol identification and parsing unit is responsible for identifying the communication protocol of each IoT device, and parsing the device communication protocol and converting it into the system internal communication format. The data transmission unit is responsible for passing the parsed and converted device data to the data communication and transmission module.

[0050] In the embodiment of the present invention, the process of device discovery and access is usually to discover and identify the device through a certain network protocol (such as SSDP, etc.). Assuming that device discovery is a broadcast process, the formula for device discovery can be expressed by the following signal propagation model: Among them, P received is the received signal strength, P transmitted is the strength of the transmitted signal, G is the system gain, d is the distance between devices, and α is the path loss exponent. Device authentication usually uses cryptographic algorithms to verify the identity of the device. The device authentication process is performed using a hash algorithm, expressed as: H auth =H(K||ID device ), where H auth is the authenticated hash value, H is the hash function, K is the shared key, and ID device A unique identifier for the device.

[0051] In one embodiment, the data communication and transmission module includes a data receiving and formatting unit, a data transmission control unit, a data cache and buffer unit, a data verification and error handling unit, and a communication protocol management unit. The data receiving and formatting unit is responsible for receiving the device data provided by the device access and protocol adaptation module and then performing formatting processing. The data transmission control unit is responsible for controlling the data transmission process. The data cache and buffer unit is responsible for providing data cache and buffer functions. The data verification and error handling unit is responsible for verifying the transmission data and taking error handling measures when data errors or transmission anomalies are found. The communication protocol management unit is responsible for managing the internal communication protocol of the system.

[0052] In the embodiment of the present invention, the data receiving and formatting unit receives the device data and performs parsing, conversion and formatting. Assume that the data transmitted by the device is X raw , and the target format is X formatted , then the process is expressed as the following formatted formula Among them, X raw The original data sent by the device, X formatted is the formatted data, and f is the format conversion function, which is responsible for converting the original data into a format that can be processed by the system. The parameters or rules required for format conversion, including data type, size, serialization method, etc.

[0053] In one embodiment, the equipment management and monitoring module includes an equipment status and health monitoring unit, an equipment performance evaluation unit, a data acquisition and storage unit, and a historical data query and playback unit. The equipment status and health monitoring unit is responsible for acquiring and monitoring the working status, operating parameters, and health status of the equipment. The equipment performance evaluation unit is responsible for analyzing the performance data of the equipment and providing evaluation indicators. The data acquisition and storage unit is responsible for collecting various types of equipment operation data and storing them. The historical data query and playback unit is responsible for querying and playing back the historical operation data of the equipment.

[0054] In the embodiment of the present invention, the device performance evaluation unit usually considers multiple performance parameters when evaluating the device performance, such as the output power, efficiency, response time, etc. of the device. Assume that the output power of the device is P out (t), the input power is P in (t), the efficiency of the equipment is expressed as: The performance evaluation of the equipment can also be combined with the working cycle and load of the equipment to form a comprehensive performance evaluation index: performance (t)=ω1η(t)+ω2U(t)+ω3R(t), where U(t) is the utilization rate of the equipment, R(t) is the response time of the equipment, and ω1, ω2, and ω3 are the weights of each performance parameter.

[0055] In one embodiment, the artificial intelligence analysis and decision-making module includes a data preprocessing unit, a pattern recognition unit, an anomaly detection and early warning unit, and a decision support unit. The data preprocessing unit is responsible for preprocessing the original data received from the equipment management and monitoring module. The pattern recognition unit uses a gradient boosting tree to identify the normal operation mode and abnormal behavior of the equipment, and distinguish between equipment failures and normal fluctuations. The anomaly detection and early warning unit is responsible for identifying abnormal conditions in equipment operation and issuing alarms in a timely manner. The decision support unit is responsible for generating decision recommendations based on data analysis results.

[0056] In the embodiment of the present invention, the anomaly detection and early warning unit is responsible for real-time monitoring of the operating status of the device, identifying any possible abnormal situation and issuing an alarm. Assuming that the health index H(t) of the device is calculated based on multiple parameters (such as temperature, vibration, power, etc.), the purpose of anomaly detection is to determine whether H(t exceeds the set threshold θ, i.e.

[0057] In one embodiment, the user interface and interaction module includes a device status display unit, a device control and configuration unit, an artificial intelligence analysis result display unit, and a user personalization setting unit. The device status display unit is responsible for displaying the current status, parameters, and health status of each device in a graphical interface. The device control and configuration unit is responsible for providing an interface for users to control the device, allowing users to remotely operate the device and configure device parameters. The artificial intelligence analysis result display unit is responsible for displaying the analysis results and decision suggestions generated by the artificial intelligence analysis and decision module to the user. The user personalization setting unit is responsible for providing personalized setting functions based on user preferences and habits.

[0058] In one embodiment, the specific process of identifying the communication protocol of each IoT device, parsing the device communication protocol, and converting it into the system internal communication format is as follows:

[0059] Protocol identification: Capture the data packets sent by IoT devices that attempt to access the system, identify the protocol type by analyzing the data packet feature fields or protocol data, and match the captured data packet features with the known communication protocols in the system. If the data packet features successfully match a known protocol, the communication protocol of the device is determined;

[0060] Protocol parsing: Based on the identified communication protocol, the protocol identification and parsing unit parses the structure of the data packet, extracts the data sent by the device from the parsed data packet, and formats the extracted device data according to the unified communication format within the system.

[0061] In one embodiment, the specific steps of using the gradient boosting tree to identify the normal operation mode and abnormal behavior of the device are:

[0062] Data preparation: Use the labeled equipment operation data set, including the normal operation data and fault data of the equipment. Suppose the training data set is {(X1,Y1),(X2,Y2),...,(X n ,Y n )}, where X i represents the feature vector, Y i Represent labels and preprocess data;

[0063] Model training: Initialize a decision tree and calculate its prediction error. Calculate the gradient of the loss function based on the prediction error, and gradually build multiple decision trees. Each decision tree corrects the prediction error of the previous decision tree.

[0064] Use boosting method to reduce the bias and variance of the model by gradually adjusting the sample weights and building a strong classifier. In each step, a new decision tree is added and its weight is adjusted to reduce the overall error in the combined strong classifier.

[0065] Model evaluation and optimization: Evaluate the performance of the model and adjust the model parameters such as the number of decision trees, learning rate, maximum depth, etc. according to the evaluation results to optimize the model performance;

[0066] Abnormal behavior mining and decision-making suggestions: The new equipment operating status and parameters are input into the trained gradient boosting tree model for prediction to determine whether the equipment is in an abnormal state. When the prediction result is a fault or the prediction probability is lower than the set threshold, the equipment is considered to be in an abnormal state and an alarm is triggered.

[0067] In one embodiment, the specific steps of the model training are:

[0068] Information gain calculation: Use information gain calculation to divide each node of the decision tree. For feature A, the information gain formula is: Among them, H(D) is the entropy of data set D, H(D v ) is the entropy of the subsets divided according to different values ​​of feature A, |D v | and |D| are subsets of D v and the size of the full set D;

[0069] Train the initial decision tree: select features through information gain to divide data and recursively construct a decision tree;

[0070] Data input: New data from the equipment management and monitoring module will be input into the decision tree model for prediction;

[0071] Apply the initial decision tree for prediction: Let the new data feature be X new The decision tree model selects features during training, classifies along the tree branches, and finally outputs the classification result Xpred , its value is normal or fault;

[0072] Calculate the forecast error: calculate the difference between the predicted value and the actual value;

[0073] Calculate the gradient: Derivative the loss function with respect to the model parameters to obtain the gradient;

[0074] Update model parameters: Use the gradient descent method to update model parameters according to the calculated gradients;

[0075] Construct new weak classifier: Construct a new weak classifier using the updated parameters.

[0076] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An artificial intelligence monitoring system based on the Internet of Things, characterized in that: include: Device access and protocol adaptation module: responsible for connecting IoT devices produced by different manufacturers to the system, identifying and parsing the communication protocol of each IoT device, converting it into a unified communication format within the system, and passing the converted device data to the data communication and transmission module; Data communication and transmission module: responsible for the transmission and exchange of equipment data, and transmits equipment data to the equipment management and monitoring module; Equipment management and monitoring module: responsible for centralized management and monitoring of access system equipment, real-time acquisition of equipment operating status and parameter information, and real-time monitoring of equipment health status and operating information, issuing alarms when abnormalities occur, and passing equipment operating status and parameters to the artificial intelligence analysis and decision-making module; Artificial intelligence analysis and decision-making module: responsible for forming a strong classifier by gradually building and combining decision trees, inputting the equipment operation status and parameters into the strong classifier, predicting whether the equipment is in an abnormal state, and when the prediction result is a fault or the prediction probability is lower than the set threshold, the equipment is considered to be in an abnormal state and an alarm is triggered. Based on the prediction results of abnormal behavior, decision suggestions are generated, and the prediction results and decision suggestions are passed to the user interface and interaction module; User interface and interaction module: responsible for presenting analysis results and decision recommendations to users in an intuitive manner, and supporting users to remotely control and configure the equipment.

2. According to claim 1, an artificial intelligence monitoring system based on the Internet of Things is characterized in that: The device access and protocol adaptation module includes a device discovery and access unit, a device authentication and security management unit, a protocol identification and parsing unit, and a data transmission unit. The device discovery and access unit is responsible for identifying and discovering new devices in the system, the device authentication and security management unit is responsible for verifying the identity of the access device, the protocol identification and parsing unit is responsible for identifying the communication protocol of each IoT device, parsing the device communication protocol, and converting it into the system internal communication format. The data transmission unit is responsible for passing the parsed and converted device data to the data communication and transmission module.

3. The artificial intelligence monitoring system based on the Internet of Things according to claim 1 is characterized in that: The data communication and transmission module includes a data receiving and formatting unit, a data transmission control unit, a data cache and buffer unit, a data verification and error handling unit and a communication protocol management unit. The data receiving and formatting unit is responsible for receiving the device data provided by the device access and protocol adaptation module and then performing formatting processing. The data transmission control unit is responsible for controlling the data transmission process. The data cache and buffer unit is responsible for providing data cache and buffer functions. The data verification and error handling unit is responsible for verifying the transmission data and taking error handling measures when data errors or transmission anomalies are found. The communication protocol management unit is responsible for managing the internal communication protocol of the system.

4. The artificial intelligence monitoring system based on the Internet of Things according to claim 1 is characterized in that: The equipment management and monitoring module includes an equipment status and health monitoring unit, an equipment performance evaluation unit, a data acquisition and storage unit, and a historical data query and playback unit. The equipment status and health monitoring unit is responsible for acquiring and monitoring the working status, operating parameters, and health status of the equipment. The equipment performance evaluation unit is responsible for analyzing the performance data of the equipment and providing evaluation indicators. The data acquisition and storage unit is responsible for collecting various types of equipment operation data and storing them. The historical data query and playback unit is responsible for querying and playing back the historical operation data of the equipment.

5. The artificial intelligence monitoring system based on the Internet of Things according to claim 1 is characterized in that: The artificial intelligence analysis and decision-making module includes a data preprocessing unit, a pattern recognition unit, an anomaly detection and early warning unit, and a decision support unit. The data preprocessing unit is responsible for preprocessing the original data received from the equipment management and monitoring module. The pattern recognition unit uses a gradient boosting tree to identify the normal operation mode and abnormal behavior of the equipment, and distinguish between equipment failures and normal fluctuations. The anomaly detection and early warning unit is responsible for identifying abnormal conditions in equipment operation and issuing alarms in a timely manner. The decision support unit is responsible for generating decision recommendations based on data analysis results.

6. The artificial intelligence monitoring system based on the Internet of Things according to claim 1 is characterized in that: The user interface and interaction module includes a device status display unit, a device control and configuration unit, an artificial intelligence analysis result display unit and a user personalization setting unit. The device status display unit is responsible for displaying the current status, parameters and health status of each device in a graphical interface. The device control and configuration unit is responsible for providing an interface for users to control the device, allowing users to remotely operate the device and configure device parameters. The artificial intelligence analysis result display unit is responsible for displaying the analysis results and decision suggestions generated by the artificial intelligence analysis and decision module to the user. The user personalization setting unit is responsible for providing personalized setting functions according to user preferences and habits.

7. The artificial intelligence monitoring system based on the Internet of Things according to claim 2 is characterized in that: The specific process of identifying the communication protocol of each IoT device, parsing the device communication protocol, and converting it into the system internal communication format is as follows: Protocol identification: Capture the data packets sent by IoT devices that attempt to access the system, identify the protocol type by analyzing the data packet feature fields or protocol data, and match the captured data packet features with the known communication protocols in the system. If the data packet features successfully match a known protocol, the communication protocol of the device is determined; Protocol parsing: Based on the identified communication protocol, the protocol identification and parsing unit parses the structure of the data packet, extracts the data sent by the device from the parsed data packet, and formats the extracted device data according to the unified communication format within the system.

8. The artificial intelligence monitoring system based on the Internet of Things according to claim 5 is characterized in that: The specific steps of using the gradient boosting tree to identify the normal operation mode and abnormal behavior of the device are: Data preparation: Use the labeled equipment operation data set, including the normal operation data and fault data of the equipment. Suppose the training data set is {(X1,Y1),(X2,Y2),...,(X n ,Y n )}, where X i represents the feature vector, Y i Represent labels and preprocess data; Model training: Initialize a decision tree and calculate its prediction error. Calculate the gradient of the loss function based on the prediction error, and gradually build multiple decision trees. Each decision tree corrects the prediction error of the previous decision tree. Use boosting method to reduce the bias and variance of the model by gradually adjusting the sample weights and building a strong classifier. In each step, a new decision tree is added and its weight is adjusted to reduce the overall error in the combined strong classifier. Model evaluation and optimization: Evaluate the performance of the model and adjust the model parameters such as the number of decision trees, learning rate, maximum depth, etc. according to the evaluation results to optimize the model performance; Abnormal behavior mining and decision-making suggestions: The new equipment operating status and parameters are input into the trained gradient boosting tree model for prediction to determine whether the equipment is in an abnormal state. When the prediction result is a fault or the prediction probability is lower than the set threshold, the equipment is considered to be in an abnormal state and an alarm is triggered.

9. The artificial intelligence monitoring system based on the Internet of Things according to claim 8, characterized in that: The specific steps of the model training are: Information gain calculation: Use information gain calculation to divide each node of the decision tree. For feature A, the information gain formula is: Among them, H(D) is the entropy of data set D, H(D v ) is the entropy of the subsets divided according to different values ​​of feature A, |D v | and |D| are subsets of D v and the size of the full set D; Train the initial decision tree: select features through information gain to divide data and recursively construct a decision tree; Data input: New data from the equipment management and monitoring module will be input into the decision tree model for prediction; Apply the initial decision tree for prediction: Let the new data feature be X new The decision tree model selects features during training, classifies along the tree branches, and finally outputs the classification result X pred , its value is normal or fault; Calculate the forecast error: calculate the difference between the predicted value and the actual value; Calculate the gradient: Derivative the loss function with respect to the model parameters to obtain the gradient; Update model parameters: Use the gradient descent method to update model parameters according to the calculated gradients; Construct new weak classifier: Construct a new weak classifier using the updated parameters.

Images