Intelligent monitoring system of Internet of Things sensing equipment based on AI (Artificial Intelligence)

Abstract

The invention, which relates to the equipment monitoring field, discloses an AI-based intelligent monitoring system for Internet-of-Things sensing equipment, comprising a sensing module for sensing equipment operation state information and environment information and storing the sensing information; the recording module is used for recording historical fault information and associated information of the equipment; the identification and regulation and control module is used for acquiring the associated information in the recording module, identifying the similarity between the associated information and the latest sensed equipment operation environment information in the sensing module, and regulating and controlling the operation frequency of the sensing module based on an identification result; according to the invention, intelligent sensing and differential storage of the equipment operation state and the environment information can be realized, similarity analysis can be carried out in combination with historical fault associated information and real-time environment data, the sensing frequency is dynamically regulated and controlled, and resource configuration is optimized while the monitoring precision is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of equipment monitoring technology, specifically to an AI-based intelligent monitoring system for IoT sensing devices. Background Technology

[0002] Intelligent equipment monitoring is an intelligent management system that uses sensors, the Internet of Things, big data, and AI technologies to collect, analyze, and issue early warnings for the real-time operation status of equipment. By deploying various monitoring modules, it can track parameters such as equipment temperature, vibration, and energy consumption in real time. Combined with preset thresholds and historical data modeling, it can automatically identify abnormal operating conditions, such as potential faults and efficiency fluctuations.

[0003] Patent application number 202411518113.0 discloses an intelligent monitoring system for substations based on IoT sensing, including sensors, interfaces, an equipment operation detection module, an intelligent analysis module, an alarm module, and a remote monitoring and control module. The sensors include temperature sensors, humidity sensors, current sensors, and voltage sensors, which collect data on the substation's temperature, humidity, current, and voltage to obtain operational data, including real-time and historical operational data. Point cloud video is acquired and constructed from the substation using laser light, resulting in point cloud video. The operational data and point cloud video are then sent to the equipment operation detection module and the intelligent analysis module via the interface, respectively. The interface connects the equipment operation detection module and the intelligent analysis module to the sensors. The equipment operation detection module uses a convolutional network method to detect the substation's operational data. When real-time operating data shows abnormalities, an operating alarm message is issued and transmitted to the alarm module. The intelligent analysis module uses an adaptive attention-based misaligned spatiotemporal feature extraction algorithm to detect intrusion behavior in point cloud video. If intrusion behavior is detected, an intrusion alarm message is issued and transmitted to the alarm module. The alarm module receives and saves the operating alarm message and the intrusion alarm message, and simultaneously issues an alarm. The remote monitoring and control module displays the substation's operating data, point cloud video, and all alarm information in real time, allowing users to remotely control the substation equipment. This application aims to solve the problem that "existing technologies can only perform simple monitoring of substations and cannot identify and detect intrusion behavior based on the monitored images. Existing computer vision technology is slow to identify dynamic geometric information in the monitoring when applied to substation monitoring, making it difficult to quickly issue danger alarms."

[0004] However, in the context of power operation and maintenance, power equipment is large in size, making it difficult to rely entirely on manual monitoring to maintain various types of power equipment.

[0005] To address this, an AI-based intelligent monitoring system for IoT sensing devices was proposed. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the existing technology, the present invention provides an AI-based intelligent monitoring system for IoT sensing devices, which can effectively solve the problems of the existing technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions;

[0008] This invention discloses an AI-based intelligent monitoring system for IoT sensing devices, comprising:

[0009] The system comprises the following modules: a sensing module for sensing equipment operating status and environmental information, and storing the sensed information; a recording module for recording historical fault information and related information; an identification and control module for acquiring related information from the recording module, identifying the similarity between the related information and the latest sensed equipment operating environment information in the sensing module, and controlling the operating frequency of the sensing module based on the identification results; an evaluation module for receiving information from the sensing module in real time and evaluating the health of the equipment operating status based on the sensed information; a judgment module for receiving the health risk assessment results of the equipment operating status from the evaluation module and determining whether the equipment operating status is healthy based on the evaluation results; an interaction module for acquiring the latest sensed equipment operating environment information applied during the operation of the identification and control module and the health risk assessment results of the equipment operating status received by the judgment module in real time, temporarily storing the equipment operating environment information and the health risk assessment results of the equipment operating status, and iterating on the temporarily stored equipment operating environment information and equipment operating status health risk assessment results at a specified period; and a protection module for receiving the judgment results of whether the equipment operating status is healthy from the judgment module and deciding whether the equipment should stop operating based on the judgment results.

[0010] Furthermore, the sensing module operates continuously at a specified frequency while the device is in operation. When storing the sensing information, the sensing module distinguishes between storing the device operating status information and the environmental information. The sensing module is equipped with an environmental sensing module and a device sensing module at its lower level. The environmental sensing module is used to sense the device operating environment information, and the device sensing module is used to sense the device operating status information.

[0011] The environmental sensing module senses the following equipment operating environment information: temperature, humidity, dust concentration, and equipment vibration audio spectrum. The equipment operating status parameters sensed by the equipment sensing module include: rated parameter deviation. Rated parameters include: current, voltage, power, insulation resistance change rate, and harmonic content.

[0012] Furthermore, when the sensing module stores the sensing information, it acquires the equipment fault determination result in real time. When the determination result indicates that a fault has been identified, it simultaneously makes a decision to delete or retain the data stored in the sensing module based on a preset time interval and the corresponding time of the equipment fault determination result.

[0013] Using the time corresponding to the equipment fault determination result as the median of the preset time interval, the perceived information within the time period covered by the preset time interval is used as the retention target, and the remaining perceived information is used as the deletion target, and the deletion and retention processing of perceived information is carried out.

[0014] The fault information recorded in the recording module includes: fault level, equipment downtime caused by the fault, and the associated information of the equipment fault information is the equipment operating status information and environmental information corresponding to the equipment fault determination time.

[0015] Among them, the equipment fault level is defined by the system user. The higher the equipment fault level, the more serious the equipment fault problem, and vice versa.

[0016] Furthermore, the logical representation of adjusting the operating frequency of the sensing module based on the recognition result in the recognition and control module is as follows:

[0017]

[0018] In the formula: SIMM(a,i) represents the similarity between the latest sensed device operating environment information a and the associated information i in the sensing module; m represents the total number of information types; f a (j), f i (j) represents the corresponding value between the j-th information in a and the j-th information in i; ω j The configuration weight for the j-th type of information; SIMM NORR For reference similarity; [SIMM(a,i),SIMM(a,i+2),SIMM(a,i+3),...] represents the similarity between the latest sensed device operating environment information a in the sensing module and the associated information i, associated information i+1, associated information i+2,...; MAX[·] represents the maximum value within the parentheses;

[0019] In this context, the configuration weights of various information items are all positive numbers, and their sum is 1. m includes temperature, humidity, dust concentration, and the audio spectrum of equipment vibration. The calculation logic of SIMM(a,i+2), SIMM(a,i+3),... is the same as that of SIMM(a,i), and SIMM is continuously calculated based on the above formula. NORR , denoted as SIMM NORR 1. SIMM NORR 2. SIMM NORR 3. When the latest three results show a continuous upward trend, the operating frequency of the sensing module increases in a stepwise manner; conversely, the operating frequency of the sensing module decreases in a stepwise manner, and the increase in operating frequency does not exceed the preset upper limit of frequency, and the decrease does not exceed the initial operating frequency.

[0020] Furthermore, when the device vibration audio spectrum is applied to the SIMM(a,i) calculation, it is converted into a characterization value based on the following formula:

[0021]

[0022] In the formula: F is the characterization value of the equipment vibration audio spectrum; f peak f c K represents the dominant frequency, sideband energy ratio, and kurtosis; (f(min)) peak f(max) peak (f(min)) represents the dominant frequency standard range; c f(max) c () represents the standard range of sideband energy ratio; (K(min), K(max)) represents the standard range of kurtosis.

[0023] Furthermore, during the operation phase of the evaluation module, it receives no fewer than three sets of the latest stored equipment operating status information and environmental information from the sensing module, and performs an evaluation of the equipment operating status health based on the received information:

[0024] ;

[0025] Where: H is the health performance value of the equipment operating status; CV temp CV hum CV dust CV vib CV curr CV volt CV pow CV insu CV harm ω1 and ω2 are the coefficients of variation of the equipment operating environment temperature, humidity, dust concentration, equipment vibration audio spectrum, equipment operating current, voltage, power, insulation resistance change rate, and harmonic content deviation; ω1 and ω2 are weighting coefficients.

[0026] The coefficients of variation are calculated in the same way, using CV as the metric. temp For example, the larger the H value, the higher the risk of equipment operation; conversely, the smaller the H value, the healthier the equipment operation.

[0027]

[0028] Where: σ temp μ is the sample standard deviation. temp This is the sample mean.

[0029] Furthermore, the weighting coefficients ω1 and ω2 both take values ​​in the range (0, 1], and follow the following rules:

[0030] The three sets of latest stored device operating status information and environmental information received during the evaluation module operation phase are traversed, and the interval collection time of adjacent information is identified to obtain two interval collection times, denoted as a1 and a2.

[0031] When a1 > a2, then ω1 > ω2; when a1 < a2, then ω1 < ω2; when a1 = a2, then ω1 and ω2 are removed from the expression.

[0032] Furthermore, when the interaction module is running, it uses the temporarily stored device operating environment information and the device operating status health risk assessment results as an external knowledge base, and configures an AI question-answering engine. The AI ​​uses data from the external knowledge base as the answer content. The system user asks the AI ​​question-answering engine questions related to device operation through the interaction module. The AI ​​question-answering engine understands the questions raised by the system user, searches for relevant management data in the external knowledge base as the answer content, and provides feedback to the system user.

[0033] Furthermore, the determination module is equipped with a health determination threshold. The determination module compares the health determination threshold with the received evaluation results to determine whether the device is in a healthy operating state.

[0034] The protection module receives the judgment results from the judgment module in real time. If the judgment result is negative, the control device stops running; otherwise, the system operation is refreshed. The protection module monitors and evaluates the latest three health evaluation results of the device's operating status. If the three consecutive evaluation results show a continuous upward trend, the control device stops running; otherwise, the system operation is refreshed.

[0035] Furthermore, the sensing module is interconnected with the environmental sensing module and the device sensing module via a wireless network. The sensing module is interconnected with the recording module, the identification and control module via a wireless network. The identification and control module is interconnected with the evaluation module and the judgment module via a wireless network. The identification and control module and the judgment module are interconnected with the interaction module via a wireless network. The judgment module is interconnected with the protection module via a wireless network.

[0036] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects:

[0037] This invention provides an AI-based intelligent monitoring system for IoT sensing devices. During operation, the system can intelligently sense and differentiate the device's operating status and environmental information. It can combine historical fault correlation information with real-time environmental data for similarity analysis, dynamically adjust the sensing frequency, and optimize resource allocation while ensuring monitoring accuracy.

[0038] By calculating the coefficient of variation of multi-dimensional operating parameters and dynamically adjusting the weights, a health evaluation model that is more in line with the actual operating scenarios of the equipment is constructed, which can predict risks based on the trend of continuous monitoring data; an external knowledge base and AI question-answering engine are innovatively introduced to realize intelligent interaction of monitoring data and accumulation of management experience.

[0039] Meanwhile, relying on the dual shutdown control logic of data screening strategy for fault time intervals, health threshold comparison, and continuous evaluation trend analysis, a multi-level equipment operation protection system has been constructed. This not only improves the efficiency of fault tracing, but also significantly enhances the safety, stability, and intelligent level of equipment operation and maintenance management through tiered frequency adjustment and intelligent decision-making mechanisms. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0041] Figure 1 This is a schematic diagram of the structure of an AI-based intelligent monitoring system for IoT sensing devices. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0043] The present invention will be further described below with reference to embodiments.

[0044] Example:

[0045] This embodiment of the intelligent monitoring system for IoT sensing devices based on AI, such as Figure 1 As shown, it includes:

[0046] The sensing module is used to sense the operating status information of the equipment and environmental information, and to store the sensed information;

[0047] The sensing module operates continuously at a specified frequency while the device is in operation. When storing the sensing information, the sensing module distinguishes between storing the device operating status information and the environmental information. The sensing module has an environmental sensing module and a device sensing module at its lower level. The environmental sensing module is used to sense the device operating environment information, and the device sensing module is used to sense the device operating status information.

[0048] Among them, the environmental sensing module senses the equipment operating environment information including: temperature, humidity, dust concentration, and equipment vibration audio spectrum; the equipment operating status parameters sensed by the equipment sensing module include: rated parameter deviation; the rated parameters include: current, voltage, power, insulation resistance change rate, and harmonic content.

[0049] When the sensing module stores the sensed information, it acquires the equipment fault determination result in real time. When the determination result indicates that a fault has been identified, it simultaneously makes a decision to delete or retain the data stored in the sensing module based on a preset time interval and the corresponding time of the equipment fault determination result.

[0050] Using the time corresponding to the equipment fault determination result as the median of the preset time interval, the perceived information within the time period covered by the preset time interval is used as the retention target, and the remaining perceived information is used as the deletion target, and the deletion and retention processing of perceived information is carried out.

[0051] The fault information recorded in the recording module includes: fault level, downtime caused by the fault, and associated information of the equipment fault information, namely the equipment operating status information and environmental information corresponding to the time of the equipment fault determination.

[0052] Among them, the equipment failure level is defined by the system user. The higher the equipment failure level, the more serious the equipment failure problem, and vice versa.

[0053] The recording module is used to record historical fault information and related information of the equipment;

[0054] The identification and control module is used to acquire related information from the recording module, identify the similarity between the related information and the latest perceived device operating environment information in the sensing module, and control the operating frequency of the sensing module based on the identification results.

[0055] The logic for adjusting the operating frequency of the sensing module based on the recognition results in the recognition and control module is as follows:

[0056]

[0057] In the formula: SIMM(a,i) represents the similarity between the latest sensed device operating environment information a and the associated information i in the sensing module; m represents the total number of information types; f a (j), f i(j) represents the corresponding value between the j-th information in a and the j-th information in i; ω j The configuration weight for the j-th type of information; SIMM NORR For reference similarity; [SIMM(a,i),SIMM(a,i+2),SIMM(a,i+3),...] represents the similarity between the latest sensed device operating environment information a in the sensing module and the associated information i, associated information i+1, associated information i+2,...; MAX[·] represents the maximum value within the parentheses;

[0058] In this context, the configuration weights of various information items are all positive numbers, and their sum is 1. m includes temperature, humidity, dust concentration, and the audio spectrum of equipment vibration. The calculation logic of SIMM(a,i+2), SIMM(a,i+3),... is the same as that of SIMM(a,i), and SIMM is continuously calculated based on the above formula. NORR , denoted as SIMM NORR 1. SIMM NORR 2. SIMM NORR 3. When the latest three results show a continuous upward trend, the operating frequency of the sensing module increases stepwise; conversely, the operating frequency of the sensing module decreases stepwise, and the increase in operating frequency does not exceed the preset upper limit, and the decrease does not exceed the initial operating frequency.

[0059] The operating frequency of the sensing module is intelligently adjusted by the above logical formula to adapt the operating frequency of the sensing module to the operating status of the equipment, thereby collecting more effective data.

[0060] When the vibration audio spectrum of the device is applied to the calculation of SIMM(a,i), it is converted into a characterization value based on the following formula:

[0061]

[0062] In the formula: F is the characterization value of the equipment vibration audio spectrum; f peak f c K represents the dominant frequency, sideband energy ratio, and kurtosis; (f(min)) peak f(max) peak (f(min)) represents the dominant frequency standard range; c f(max) c () represents the standard range of sideband energy ratio; (K(min), K(max)) represents the standard range of kurtosis.

[0063] The above logical formula is used to convert the device vibration audio spectrum into a characterization value for the calculation of the above SIMM(a,i);

[0064] During the evaluation module's operation phase, it receives no fewer than three sets of the latest stored equipment operating status information and environmental information from the sensing module, and performs an evaluation of the equipment's operating status health based on the received information.

[0065] ;

[0066] Where: H is the health performance value of the equipment operating status; CV temp CV hum CV dust CV vib CV curr CV volt CV pow CV insu CV harm ω1 and ω2 are the coefficients of variation of the equipment operating environment temperature, humidity, dust concentration, equipment vibration audio spectrum, equipment operating current, voltage, power, insulation resistance change rate, and harmonic content deviation; ω1 and ω2 are weighting coefficients.

[0067] The coefficients of variation are calculated in the same way, using CV as the metric. temp For example, the larger the H value, the higher the risk of equipment operation; conversely, the smaller the H value, the healthier the equipment operation.

[0068]

[0069] Where: σ temp μ is the sample standard deviation. temp The mean of the sample;

[0070] The weighting coefficients ω1 and ω2 both take values ​​in the range (0, 1] and follow the following rules:

[0071] The three sets of latest stored device operating status information and environmental information received during the evaluation module operation phase are traversed, and the interval collection time of adjacent information is identified to obtain two interval collection times, denoted as a1 and a2.

[0072] When a1 > a2, then ω1 > ω2; when a1 < a2, then ω1 < ω2; when a1 = a2, then ω1 and ω2 are removed from the expression.

[0073] The health status of the equipment is represented in digital form through the above logical formula calculation, providing necessary data support for the further operation of the system's evaluation module in this embodiment;

[0074] The evaluation module is used to receive information from the sensing module in real time and evaluate the health status of the equipment based on the sensing information.

[0075] The determination module is used to receive the health insurance evaluation results of the equipment operating status from the evaluation module, and determine whether the equipment operating status is healthy based on the evaluation results.

[0076] The judgment module is equipped with a health judgment threshold. The judgment module compares the health judgment threshold with the received evaluation results to determine whether the device is in a healthy operating state.

[0077] The protection module receives the judgment results from the judgment module in real time. If the judgment result is negative, the control device stops running. Otherwise, the system operation is refreshed. The protection module monitors and evaluates the latest three health evaluation results of the device's operating status. If the three consecutive evaluation results show a continuous upward trend, the control device stops running. Otherwise, the system operation is refreshed.

[0078] The interaction module is used to acquire in real time the latest perceived equipment operating environment information and the equipment operating status health risk assessment results received by the judgment module during the operation phase of the identification and control module. It temporarily stores the equipment operating environment information and the equipment operating status health risk assessment results, and iterates the temporarily stored equipment operating environment information and equipment operating status health risk assessment results based on a specified period.

[0079] When the interaction module is running, it uses the temporarily stored equipment operating environment information and the equipment operating status health risk assessment results as an external knowledge base, and configures an AI question answering engine. The AI ​​uses the data in the external knowledge base as the answer content. The system user asks the AI ​​question answering engine questions related to equipment operation through the interaction module. The AI ​​question answering engine understands the questions asked by the system user, searches for the corresponding management data in the external knowledge base as the answer content, and provides feedback to the system user.

[0080] It is important to note that:

[0081] When the interactive module starts running, it integrates the temporarily stored device operating environment information and the device operating status health evaluation results into an external knowledge base.

[0082] To achieve efficient question-and-answer interaction, the system is equipped with an AI question-and-answer engine based on Natural Language Processing (NLP) technology. This engine employs a Transformer architecture and combines it with pre-trained models such as BERT. Through fine-tuning of corpora from the device operation and maintenance domain, it can accurately understand device operation-related questions raised by users on the system side. For example, whether a user asks "How to solve the problem of high CPU usage on the current device?" or "What is the reason for the frequent device restarts recently?", the AI ​​question-and-answer engine can parse the semantics of the question and extract key information.

[0083] Upon receiving a user's question, the AI ​​question-answering engine first performs intent recognition and entity extraction to determine the core request. Then, the engine performs a multi-dimensional search in an external knowledge base, using techniques such as inverted indexes and semantic matching to quickly locate relevant data. If a directly matching answer exists, the AI ​​question-answering engine will structure and optimize its language to provide a concise and easy-to-understand response to the user. If no exact match is found, the engine will perform logical reasoning based on related data in the knowledge base to generate a comprehensive answer. Simultaneously, to ensure the accuracy and timeliness of the responses, the system also manually reviews and dynamically updates the content output by the AI ​​question-answering engine, forming a closed-loop management system.

[0084] The protection module is used to receive the judgment result of whether the device is operating healthily from the judgment module, and decide whether the device should stop operating based on the judgment result;

[0085] The sensing module is connected to the environmental sensing module and the device sensing module via a wireless network. The sensing module is connected to the recording module, the identification and control module via a wireless network. The identification and control module is connected to the evaluation module and the judgment module via a wireless network. The identification and control module and the judgment module are connected to the interaction module via a wireless network. The judgment module is connected to the protection module via a wireless network.

[0086] In this embodiment, the sensing module senses the device's operating status and environmental information, stores the sensed information, the environmental sensing module synchronously senses the device's operating environment information, the device sensing module senses the device's operating status information in real time, the recording module records the device's historical fault information and related information, the identification and control module further obtains related information from the recording module, identifies the similarity between the related information and the latest sensed device operating environment information in the sensing module, and adjusts the operating frequency of the sensing module based on the identification results, then the evaluation module receives the information sensed by the sensing module in real time, evaluates the health of the device's operating status based on the sensed information, and receives the device operating status health risk evaluation results from the evaluation module through the judgment module, and determines whether the device's operating status is healthy based on the evaluation results, the interaction module obtains the latest sensed device operating environment information applied by the identification and control module in real time and the device operating status health risk evaluation results received by the judgment module, temporarily stores the device operating environment information and device operating status health risk evaluation results, and iterates the temporarily stored device operating environment information and device operating status health risk evaluation results according to a specified period, and finally, the protection module receives the determination result of whether the device's operating status is healthy from the judgment module, and decides whether the device should stop operating based on the determination result.

[0087] Through the operation of the above system, online intelligent monitoring services are effectively provided for power equipment in power operation and maintenance scenarios. Real-time monitoring is based on multi-source data to ensure the long-term stable operation of power equipment.

[0088] The following is an example application of the system described in the above embodiments:

[0089] Scenario: Monitoring of the main transformer in a substation of a power grid company. The goal is to monitor the equipment's operating status in real time, predict internal anomalies, and provide intelligent operation and maintenance support.

[0090] I. Sensing Module Acquires Multidimensional Information from Devices

[0091] Environmental perception section:

[0092] Monitor the equipment's operating environment using deployed sensors:

[0093] Real-time sensing of the temperature and humidity of the space where the device is located to determine whether it is in a normal environmental range;

[0094] Detecting the concentration of specific gases in the environment; abnormal concentrations may indicate a risk of equipment leakage.

[0095] The vibration signals of the equipment during operation are collected, the vibration frequency and energy distribution characteristics are analyzed, and the operating status of the internal mechanical components is determined.

[0096] Equipment status sensing section:

[0097] Access the device control system to obtain electrical and insulation related parameters:

[0098] Monitor the current and voltage values ​​of the equipment to determine the load condition and voltage stability;

[0099] Track insulation-related parameters, such as the rate of change of winding resistance, the degree of dielectric loss, and the gas composition in the oil, to assess whether the insulation performance has deteriorated or become damp.

[0100] Data storage mechanism:

[0101] The collected environmental information and device status information are categorized and stored. By default, data is collected continuously at certain time intervals. If an abnormal situation is detected, the collection frequency is automatically increased.

[0102] II. Recording module stores historical anomaly data.

[0103] The system records all faults and abnormal events since the equipment was put into operation, including:

[0104] The time and duration of each anomaly;

[0105] Anomaly level (defined by operations and maintenance personnel based on the degree of impact);

[0106] Information related to the occurrence of anomalies, such as environmental parameters and equipment status parameters at the time, provides historical references for subsequent analysis.

[0107] III. Identification and Control Module Dynamically Adjusts Monitoring Frequency

[0108] Similarity analysis:

[0109] The system compares the currently collected environmental and equipment parameters with historical abnormal event correlation information in real time and calculates the similarity between the two. If the similarity exceeds the preset standard and the calculation results show an upward trend for multiple consecutive times, it indicates that the equipment may be approaching an abnormal state.

[0110] Frequency control logic:

[0111] When an upward trend in similarity is detected, the system automatically increases the sampling frequency of the sensing module to track device status changes more intensively; if the similarity drops to the normal range, the default sampling frequency is restored, ensuring monitoring effectiveness while avoiding resource waste.

[0112] IV. Evaluation Module: Comprehensive Assessment of Equipment Health Status

[0113] The system selects the latest sets of collected data and calculates a comprehensive health status value of the equipment by analyzing the fluctuations and interrelationships of various parameters. A higher value indicates a greater risk to equipment operation, while a lower value indicates a more stable condition. During the evaluation process, the influence weight of each parameter is dynamically adjusted based on the data collection time interval to ensure the accuracy of the results.

[0114] V. Risk Management of Judgment and Safeguard Module

[0115] Status determination: The system presets health determination standards, compares the comprehensive evaluation of health performance values ​​with the standards, and if the values ​​exceed the threshold, the device is determined to be in an abnormal state, triggering an early warning mechanism.

[0116] Risk control measures:

[0117] If the equipment status is abnormal, the system will first send an alarm message to the operation and maintenance personnel, indicating the type of abnormality and possible causes.

[0118] If the abnormal trend continues, the system will automatically adjust the equipment's operating load to reduce internal losses; if the risk further intensifies, it will execute shutdown protection to prevent the fault from escalating and causing a serious accident.

[0119] VI. The interaction module provides intelligent question-and-answer support.

[0120] External knowledge base: The system integrates real-time collected environmental parameters, equipment health assessment results, and historical anomaly correlation data into a knowledge base for the intelligent question-answering engine to use.

[0121] Example of AI question-and-answer application:

[0122] A user asked: "Could you provide me with the latest operating status parameters / environmental status parameters for my device?"

[0123] Engine response:

[0124] Retrieve the corresponding data from the knowledge base and provide feedback to the user who asked the question.

[0125] VII. The Core Value of the System in Power Scenarios

[0126] Fault prediction capability: Through multi-parameter correlation analysis and historical data comparison, potential internal anomalies (such as winding overheating and insulation dampness) can be detected in advance, transforming passive maintenance into proactive maintenance and reducing the probability of sudden failures.

[0127] Improved operational efficiency: The intelligent question-and-answer engine responds to user inquiries in real time and provides targeted handling suggestions based on historical experience, reducing manual troubleshooting time and improving fault response speed.

[0128] Equipment safety assurance: Through dynamic load regulation and automatic shutdown mechanisms, risks are effectively controlled to avoid power grid outages caused by equipment failures, thus ensuring the stability and reliability of power supply.

[0129] This example uses a power transformer as an example to fully demonstrate the entire process of the system from data acquisition, analysis and early warning to intelligent decision-making, highlighting its practical application value in the status management of power equipment and meeting the needs of power grid operation and maintenance for intelligent and lean equipment management.

[0130] In summary, the system in the above embodiments can intelligently perceive and differentiate the equipment operating status and environmental information during operation. It can perform similarity analysis by combining historical fault correlation information and real-time environmental data, dynamically adjust the sensing frequency, optimize resource allocation while ensuring monitoring accuracy, and construct a health evaluation model that is more in line with the actual operating scenario of the equipment through the calculation of the coefficient of variation of multi-dimensional operating parameters and the dynamic adjustment mechanism of weights. It can predict risks based on the trend of continuous monitoring data. The system innovatively introduces an external knowledge base and an AI question-answering engine to realize intelligent interaction of monitoring data and accumulation of management experience. At the same time, relying on the data screening strategy of fault time interval, the dual shutdown control logic of health threshold comparison and continuous evaluation trend analysis, a multi-level equipment operation protection system is constructed. This not only improves the efficiency of fault tracing, but also significantly enhances the safety, stability and intelligent level of operation and maintenance management of equipment through tiered frequency adjustment and intelligent decision-making mechanisms.

[0131] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An AI-based intelligent monitoring system for IoT sensing devices, characterized in that, include: The sensing module is used to sense the operating status information of the equipment and environmental information, and to store the sensed information; The recording module is used to record historical fault information and related information of the equipment; The identification and control module is used to acquire related information from the recording module, identify the similarity between the related information and the latest perceived device operating environment information in the sensing module, and control the operating frequency of the sensing module based on the identification results. The evaluation module is used to receive information from the sensing module in real time and evaluate the health status of the equipment based on the sensing information. The determination module is used to receive the health insurance evaluation results of the equipment operating status from the evaluation module, and determine whether the equipment operating status is healthy based on the evaluation results. The interaction module is used to acquire in real time the latest perceived equipment operating environment information and the equipment operating status health risk assessment results received by the judgment module during the operation phase of the identification and control module. It temporarily stores the equipment operating environment information and the equipment operating status health risk assessment results, and iterates the temporarily stored equipment operating environment information and equipment operating status health risk assessment results based on a specified period. The protection module receives the judgment result from the judgment module regarding whether the device's operating status is healthy, and decides whether the device should stop operating based on the judgment result.

2. The AI-based intelligent monitoring system for IoT sensing devices according to claim 1, characterized in that, The sensing module operates continuously at a specified frequency while the device is running. When storing the sensing information, the sensing module distinguishes between storing the device operating status information and the environmental information. The sensing module is equipped with an environmental sensing module and a device sensing module. The environmental sensing module is used to sense the device operating environment information, and the device sensing module is used to sense the device operating status information. The environmental sensing module senses the following equipment operating environment information: temperature, humidity, dust concentration, and equipment vibration audio spectrum. The equipment operating status parameters sensed by the equipment sensing module include: rated parameter deviation. Rated parameters include: current, voltage, power, insulation resistance change rate, and harmonic content.

3. The AI-based intelligent monitoring system for IoT sensing devices according to claim 1, characterized in that, When the sensing module stores the sensed information, it acquires the equipment fault determination result in real time. When the determination result indicates that a fault has been identified, it simultaneously makes a decision to delete or retain the data stored in the sensing module based on a preset time interval and the corresponding time of the equipment fault determination result. Using the time corresponding to the equipment fault determination result as the median of the preset time interval, the perceived information within the time period covered by the preset time interval is used as the retention target, and the remaining perceived information is used as the deletion target, and the deletion and retention processing of perceived information is carried out. The fault information recorded in the recording module includes: fault level, equipment downtime caused by the fault, and the associated information of the equipment fault information is the equipment operating status information and environmental information corresponding to the equipment fault determination time. Among them, the equipment fault level is defined by the system user. The higher the equipment fault level, the more serious the equipment fault problem, and vice versa.

4. The AI-based intelligent monitoring system for IoT sensing devices according to claim 1, characterized in that, The logical representation of adjusting the operating frequency of the sensing module based on the recognition result in the recognition and control module is as follows: In the formula: SIMM(a,i) represents the similarity between the latest sensed device operating environment information a and the associated information i in the sensing module; m represents the total number of information types; f a (j), f i (j) represents the corresponding value between the j-th information in a and the j-th information in i; ω j The configuration weight for the j-th type of information; SIMM NORR For reference similarity; [SIMM(a,i),SIMM(a,i+2),SIMM(a,i+3),...] represents the similarity between the latest sensed device operating environment information a in the sensing module and the associated information i, associated information i+1, associated information i+2,...; MAX[·] represents the maximum value within the parentheses; In this context, the configuration weights of various information items are all positive numbers, and their sum is 1. m includes temperature, humidity, dust concentration, and the audio spectrum of equipment vibration. The calculation logic of SIMM(a,i+2), SIMM(a,i+3),... is the same as that of SIMM(a,i), and SIMM is continuously calculated based on the above formula. NORR , denoted as SIMM NORR 1. SIMM NORR 2. SIMM NORR 3. When the latest three results show a continuous upward trend, the operating frequency of the sensing module increases in a stepwise manner; conversely, the operating frequency of the sensing module decreases in a stepwise manner, and the increase in operating frequency does not exceed the preset upper limit of frequency, and the decrease does not exceed the initial operating frequency.

5. The AI-based intelligent monitoring system for IoT sensing devices according to claim 4, characterized in that, When the vibration audio spectrum of the device is applied to the calculation of SIMM(a,i), it is converted into a characterization value based on the following formula: In the formula: F is the characterization value of the equipment vibration audio spectrum; f peak f c K represents the dominant frequency, sideband energy ratio, and kurtosis; (f(min)) peak f(max) peak (f(min)) represents the dominant frequency standard range; c f(max) c () represents the standard range of sideband energy ratio; (K(min), K(max)) represents the standard range of kurtosis.

6. The AI-based intelligent monitoring system for IoT sensing devices according to claim 1, characterized in that, During the operation phase of the evaluation module, it receives no less than three sets of the latest stored equipment operating status information and environmental information from the sensing module, and performs a health evaluation of the equipment operating status based on the received information. ； In the formula: This refers to the health performance value of the equipment's operating status; The coefficient of variation of the equipment operating environment temperature, humidity, dust concentration, equipment vibration audio spectrum, equipment operating current, voltage, power, insulation resistance change rate, and harmonic content deviation; These are the weighting coefficients; The calculation methods for each coefficient of variation are the same, using... For example; The larger the value, the higher the risk of equipment operation; conversely, the smaller the value, the healthier the equipment operation. ； In the formula: The standard deviation is the sample standard deviation. This is the sample mean.

7. The AI-based intelligent monitoring system for IoT sensing devices according to claim 6, characterized in that, The weighting coefficients ω1 and ω2 both take values ​​in the range (0, 1] and follow the following rules: The three sets of latest stored device operating status information and environmental information received during the evaluation module operation phase are traversed, and the interval collection time of adjacent information is identified to obtain two interval collection times, denoted as a1 and a2. When a1 > a2, then ω1 > ω2; when a1 < a2, then ω1 < ω2; when a1 = a2, then ω1 and ω2 are removed from the expression.

8. The AI-based intelligent monitoring system for IoT sensing devices according to claim 1, characterized in that, When the interaction module is running, it uses the temporarily stored equipment operating environment information and equipment operating status health risk assessment results as an external knowledge base, and configures an AI question-answering engine. The AI ​​uses data from the external knowledge base as the answer content. The system user asks the AI ​​question-answering engine questions related to equipment operation through the interaction module. The AI ​​question-answering engine understands the questions raised by the system user, searches for relevant management data in the external knowledge base as the answer content, and provides feedback to the system user.

9. The AI-based intelligent monitoring system for IoT sensing devices according to claim 1, characterized in that, The determination module is equipped with a health determination threshold. The determination module compares the health determination threshold with the received evaluation results to determine whether the device is in a healthy operating state. The protection module receives the judgment results from the judgment module in real time. If the judgment result is negative, the control device stops running; otherwise, the system operation is refreshed. The protection module monitors and evaluates the latest three health evaluation results of the device's operating status. If the three consecutive evaluation results show a continuous upward trend, the control device stops running; otherwise, the system operation is refreshed.

10. The AI-based intelligent monitoring system for IoT sensing devices according to claim 1, characterized in that, The sensing module is interconnected with the environmental sensing module and the device sensing module via a wireless network. The sensing module is interconnected with the recording module, the identification and control module via a wireless network. The identification and control module is interconnected with the evaluation module and the judgment module via a wireless network. The identification and control module and the judgment module are interconnected with the interaction module via a wireless network. The judgment module is interconnected with the protection module via a wireless network.

Images