Circuit breaker based ai patrol method and system

Abstract

The application discloses an AI inspection method and system based on a circuit breaker, and relates to the technical field of AI inspection. The method comprises the following steps: receiving an intelligent inspection request, determining a plurality of to-be-inspected devices and a first inspection rule set matched by the plurality of to-be-inspected devices based on the intelligent inspection request; collecting real-time operation data of the plurality of to-be-inspected devices by using an intelligent circuit breaker, and obtaining a historical operation data set corresponding to the plurality of to-be-inspected devices that meet preset operation rules; inputting the historical operation data set into a dynamic baseline model to output a second inspection rule set; and generating a device inspection report according to the real-time operation data corresponding to the plurality of to-be-inspected devices based on the first inspection rule set and the second inspection rule set. Abnormities can be found and risks can be predicted from massive device data, and the reliability and user experience of a power system are improved.

Description

Technical Field

[0001] This application relates to the field of AI inspection technology, and more specifically, to an AI inspection method and system based on circuit breakers. Background Technology

[0002] In power systems, circuit breakers, as critical electrical protection devices, bear the important mission of quickly cutting off current when circuit faults occur, thereby protecting equipment and personnel safety. Traditional circuit breakers have undergone long-term development and are relatively mature in technology, mainly focusing on achieving basic circuit breaking protection functions. That is, when abnormal conditions such as overload or short circuit occur in the circuit, they can promptly disconnect the circuit to prevent the fault from escalating and ensure the stable operation of the power system.

[0003] With the rapid development of emerging technologies such as the Internet of Things, big data, and artificial intelligence, the power industry is undergoing profound changes, and intelligentization has become an important trend in circuit breaker development. Intelligent circuit breakers have emerged to meet this need. They not only fully retain the on / off protection functions of traditional circuit breakers but also leverage advanced sensor technology, communication technology, and intelligent algorithms to achieve real-time data acquisition and remote control functions. Furthermore, with the continuous advancement of smart grid construction and the widespread integration of distributed energy resources, the number of devices in the power system is increasing dramatically. Detecting anomalies and predicting risks from massive amounts of device data is a core issue in improving the reliability of the power system and the user experience.

[0004] Existing inspection methods often rely on simple threshold comparisons or empirical rules, which are insufficient to handle the complex and ever-changing operating conditions of power systems. For example, for current overload faults, traditional methods simply set a fixed current threshold, and a fault is identified when the current exceeds this threshold. However, in actual operation, the magnitude of the current is affected by various factors, such as load changes and ambient temperature. A fixed threshold may lead to misjudgments or missed judgments, resulting in inaccurate inspection results and thus affecting the user experience. Summary of the Invention

[0005] The main purpose of this application is to provide an AI inspection method and system based on circuit breakers to detect anomalies and predict risks from massive amounts of equipment data, thereby improving the reliability of the power system and the user experience.

[0006] To achieve the above objectives, according to one aspect of the embodiments of this application, an AI inspection method based on circuit breakers is proposed, comprising: receiving an intelligent inspection request, and determining multiple devices to be inspected and a first set of inspection rules matching the multiple devices to be inspected based on the intelligent inspection request; collecting real-time operating data of the multiple devices to be inspected using an intelligent circuit breaker, and obtaining a set of historical operating data corresponding to the multiple devices to be inspected that conform to preset operating rules; inputting the set of historical operating data into a dynamic baseline model and outputting a second set of inspection rules; and generating an equipment inspection report based on the real-time operating data corresponding to the multiple devices to be inspected according to the first set of inspection rules and the second set of inspection rules.

[0007] According to another aspect of the embodiments of this application, an AI inspection system based on circuit breakers is also provided, including:

[0008] The application and interaction module is used to receive intelligent inspection requests and equipment inspection reports. Intelligent circuit breakers are used to collect real-time operating data from multiple devices to be inspected. The intelligent inspection module is used to determine multiple devices to be inspected based on intelligent inspection requests, as well as a first set of inspection rules matching the multiple devices to be inspected; and to obtain a set of historical operating data corresponding to the multiple devices to be inspected that conform to preset operating rules; input the set of historical operating data into a dynamic baseline model and output a second set of inspection rules; and to generate an equipment inspection report based on the real-time operating data corresponding to the multiple devices to be inspected according to the first set of inspection rules and the second set of inspection rules.

[0009] Optionally, the aforementioned AI inspection system based on circuit breakers also includes: The communication module connects multiple smart circuit breakers and smart inspection modules. It is used to decapsulate the source data packets corresponding to the real-time operation data received from the smart circuit breakers, recapsulate them according to the protocol format of the smart inspection module to obtain the target data packets, and send the target data packets to the smart inspection module.

[0010] According to another aspect of the embodiments of this application, a computer-readable storage medium is provided, which stores computer instructions for causing a computer to perform the above-described AI inspection method based on circuit breakers.

[0011] According to another aspect of the embodiments of this application, an electronic device is also provided, the electronic device including: at least one processor, and a memory communicatively connected to the at least one processor; wherein, the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the at least one processor to perform the above-described AI inspection method based on circuit breakers.

[0012] The technical solutions provided by the embodiments of this application may include the following beneficial effects: In this application, the aforementioned AI inspection method and system based on circuit breakers can detect anomalies and predict risks from massive amounts of equipment data, thereby improving the reliability of the power system and the user experience. Attached Figure Description

[0013] The accompanying drawings, which form part of this application, are used to provide a further understanding of the application and to make other features, objects, and advantages of the application more apparent. The illustrative embodiments and descriptions of this application are used to explain the application and do not constitute an undue limitation of the application. In the drawings: Figure 1 A flowchart of an optional AI inspection method based on circuit breakers provided in this application; Figure 2 An architecture diagram of an optional circuit breaker-based AI inspection system provided for this application; Figure 3 A schematic diagram of an optional intelligent inspection module provided in this application. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in any order other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. Without conflict, the embodiments and features in the embodiments of this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0016] To improve the reliability of power systems and enhance user experience by identifying anomalies and predicting risks from massive amounts of equipment data, this application provides an AI-based inspection method and system for circuit breakers. As an optional implementation method, please refer to... Figure 1 The diagram illustrates a flowchart of an AI inspection method based on circuit breakers according to an embodiment of this application. The method includes at least one of the following steps (S102 to S108): S102, receive intelligent inspection request, and determine multiple devices to be inspected and a first set of inspection rules matching the multiple devices to be inspected based on the intelligent inspection request; S104 uses an intelligent circuit breaker to collect real-time operating data of multiple devices to be inspected and obtains a set of historical operating data corresponding to multiple devices to be inspected that conform to preset operating rules. S106: Input the historical operation data set into the dynamic baseline model and output the second inspection rule set; S108, Generate an equipment inspection report based on the real-time operating data of multiple devices to be inspected according to the first inspection rule set and the second inspection rule set.

[0017] In S102 above, the first inspection rule set includes fixed inspection standards determined according to the set of parameters to be inspected, wherein the set of parameters to be inspected includes the parameter types corresponding to the equipment to be inspected.

[0018] In S104 above, if the equipment to be inspected meets the preset operating rules, it can be indicated that the equipment is in normal operating condition. The corresponding historical operating data set includes historical operating data recorded during the normal operation of the equipment. Both the real-time operating data set and the historical operating data set can be matched with the parameter set to be inspected. That is to say, the data types indicated by the real-time operating data and the historical operating data set can correspond to the parameter types indicated by the parameters to be inspected.

[0019] In S106 above, the dynamic baseline model includes an intelligent model (AI model) that can automatically learn (such as through statistical methods or unsupervised learning) based on historical operational data sets and generate parameters within a normal fluctuation range. Specific functions include, but are not limited to: 1. Deviation detection: Compare real-time running data with the dynamic baseline model to identify situations that significantly deviate from the expected pattern (e.g., persistently high current in an empty office at midnight).

[0020] 2. Correlation Anomaly Analysis: Analyze whether the correlation between multiple parameters has been disrupted (e.g., high power but normal current may indicate measurement or metering anomalies).

[0021] 3. Pattern change identification: Sudden change in the operating mode of the equipment to be inspected (e.g., abnormal motor starting current waveform).

[0022] The second set of inspection rules includes dynamic inspection standards generated by the dynamic baseline model, which reflect the normal fluctuation range of parameters.

[0023] In S108 above, the equipment inspection report can analyze and compare the real-time operating data collected by the intelligent circuit breaker to make a comprehensive diagnosis, which may include the final judgment on whether the current operating status of the equipment is normal, whether there are any abnormalities or potential risks.

[0024] Through the above-described embodiments of this application, the AI ​​inspection method based on circuit breakers can detect anomalies and predict risks from massive amounts of equipment data, thereby improving the reliability of the power system and user experience. Through automated monitoring and analysis, it can improve the operation and maintenance efficiency and safety of electrical equipment. Based on the first inspection rule set and the second inspection rule set, it can adapt to the data monitoring needs in different scenarios.

[0025] As an optional implementation, the above-mentioned historical operational data set is input into the dynamic baseline model, and the output is a second inspection rule set, including: S1, The dynamic baseline model clusters the historical operational data set to obtain the clustering results; S2, determine the operating modes of multiple devices based on the clustering results; S3 determines the second inspection rule corresponding to each type of historical operating data under each equipment operating mode based on multiple equipment operating modes. The set of second inspection rules includes multiple second inspection rules.

[0026] In S1 above, clustering involves using existing algorithms (such as K-means, DBSCAN, etc.) to automatically group data based on various characteristics of historical operational datasets (such as collection timestamps, environmental information, device status (running / standby), load rate, runtime, etc.), thereby classifying data with similar characteristics into the same category. The clustering results include the data groups obtained after the clustering operation.

[0027] In S2 above, the equipment operation mode includes typical equipment working states that can be understood based on the clustering results, such as high temperature full load mode (characterized by high temperature + high power + long-term operation), low temperature standby mode (characterized by low temperature + low power + short-term operation), and variable temperature intermittent mode (characterized by temperature fluctuation + periodic power change), etc.

[0028] In S3 above, the types of historical operating data include different categories of operating parameters or status indicators collected from the equipment to be inspected. This can be understood as the classification of the operating data of the equipment to be inspected. Each category represents a specific monitoring dimension, such as: electrical parameters (voltage, current, power, power factor, etc.), status parameters (running time of the equipment, number of switching, load rate, etc.), and environmental parameters (temperature, humidity, etc. of the environment where the equipment is located).

[0029] The above-mentioned S1-S3 automatically identify a variety of typical equipment operation modes by clustering and analyzing historical operation data sets. Based on this, more refined second inspection rules are generated to match each mode, thereby realizing the transformation from fixed "one-size-fits-all" inspections to intelligent and precise inspections based on "scenario and mode". This significantly improves the ability to identify abnormal equipment states under complex operating conditions and the accuracy of early warnings.

[0030] As an optional implementation, the above-mentioned second inspection rule, which determines the type of historical operating data corresponding to each equipment operating mode under each equipment operating mode, includes: S3-1, determine the historical average running time for each type of equipment to be inspected based on the historical operating data set; S3-2, Obtain the preset fluctuation range corresponding to each type of historical operational data; S3-3 determines the second inspection rule corresponding to each type of historical operating data based on the historical average runtime, multiple equipment operating modes, and preset fluctuation range.

[0031] In S3-1 above, the historical average running time includes the average running time of each device to be inspected within the historical running data statistical period, which can indicate the normal working cycle of the device to be inspected.

[0032] In S3-2 above, the preset fluctuation range includes the pre-set range of normal numerical changes for each operating parameter under each operating mode of the equipment to be inspected (such as "high temperature full load mode"). For example, the preset fluctuation range of current is 90% to 110% of the rated current; the preset fluctuation range of voltage is -5% to +5% of the rated voltage.

[0033] In S3-3 above, the second inspection rule includes more specific and operational early warning triggering logic based on the preset fluctuation range. For example, in high temperature full load mode, the current rule is: the reference fluctuation range of the current is 90%-110% of the rated current (if it continuously exceeds 110% and the continuous running time is greater than 30 minutes, an early warning will be triggered; if it exceeds 120%, an early warning will be triggered immediately); the voltage rule is: the reference fluctuation range of the voltage is -5% to +5% of the rated voltage (if the fluctuation exceeds -10% to +10% and the continuous running time is greater than 15 minutes, an early warning will be triggered); while in low temperature standby mode, the reference fluctuation range of the current is 50%-100% of the rated current (if the current suddenly rises to more than 100%, an early warning will be triggered, which may indicate leakage or short circuit risk), and the voltage is -3% to +3% of the rated voltage (for 48V communication equipment, the baseline is 46.56V-49.44V; if the voltage is lower than 46V or higher than 50V, an early warning will be triggered to check the power supply or battery health status), etc.

[0034] The above-mentioned S3-1 to S3-3 can upgrade the intelligent early warning from a simple judgment of "whether it is abnormal" to a precise and dynamic judgment of "in what mode, how much it deviates, and how long it lasts to be considered abnormal" by setting a quantified preset fluctuation range and a second inspection rule including a duration criterion for the operating parameters (such as current and voltage) in each device operating mode. This will greatly reduce false alarms and provide more timely and reliable early warnings for real potential faults.

[0035] As an optional implementation, the above-mentioned generation of equipment inspection reports based on the real-time operating data corresponding to multiple devices to be inspected according to the first inspection rule set and the second inspection rule set includes: S1, obtain the type of first inspection parameter corresponding to each first inspection rule in the first inspection rule set, and the type of second inspection parameter corresponding to each second inspection rule in the second inspection rule set; S2, based on multiple first inspection rule types and multiple second inspection rule types, perform rule matching between multiple first inspection rules and multiple second inspection rules to obtain a first matching rule set and a second matching rule set, wherein the first matching rule set includes multiple matching rule pairs; S3, obtain the target inspection rule type corresponding to each matching rule pair; S4 generates an equipment inspection report based on multiple matching rule pairs, multiple target inspection rule types, the first matching rule set, the second matching rule set, and the real-time operating data corresponding to multiple devices to be inspected.

[0036] In S1-S3 above, the first inspection parameter type includes the equipment parameter type targeted by the preset, fixed first inspection rule. For example, a fixed rule is "check if the current exceeds 100A", then "current" is the first inspection parameter type corresponding to this rule. The second inspection parameter type includes the equipment parameter type targeted by the dynamic second inspection rule generated by the dynamic baseline model. For example, the dynamic model generates a rule for "high temperature full load mode" that "check if the current continuously exceeds 110% of the rated value", then "current" is the second inspection parameter type corresponding to this rule. Matching rule pairs include: when the system finds that a first inspection parameter type and a second inspection parameter type point to the same equipment parameter, it will pair the first inspection rule and the second rule for that parameter to form a matching rule pair. For example, a fixed current rule is paired with a dynamic current rule.

[0037] The equipment inspection report in S4 above is a targeted and reliable final result obtained by the system after comprehensive analysis of real-time operating data.

[0038] The above-mentioned S1-S4 automatically match the parameter types of the first inspection rule with multiple second inspection rules, which can generate equipment inspection reports with unified rules, highlighting key points and making more accurate judgments, thereby improving the coordination and decision-making efficiency of the entire inspection system.

[0039] As an optional implementation, the above-mentioned generation of an equipment inspection report based on multiple matching rule pairs, multiple target inspection rule types, a first matching rule set, a second matching rule set, and the real-time operating data corresponding to multiple devices to be inspected includes: S4-1, send multiple matching rule pairs and multiple target inspection rule types to multiple monitoring devices corresponding to the target account, and determine the target inspection rule corresponding to each target inspection rule type based on the target account's selection trigger operation on the target monitoring device for multiple matching rule pairs; S4-2, Detect the real-time operating data corresponding to each of the devices to be inspected according to the multiple target inspection rules and the second matching rule set, and obtain the detection result; S4-3, if the detection result indicates that the real-time operating data does not match the target inspection rule and / or the second matching rule set, send early warning information to multiple monitoring devices; S4-4, Generate the equipment inspection report based on the warning information.

[0040] In S4-1 to S4-2 above, the selection trigger operation includes the system providing an interactive interface, allowing users (operators) to actively select from matching rule pairs, giving users ultimate control and flexibility. The target monitoring device includes one of multiple monitoring devices. The source of the target inspection rules includes rules actively selected by the user and second inspection rules output by the dynamic baseline model.

[0041] In S4-3 to S4-4 above, the warning information can be divided into high level (such as instantaneous high current), medium level (such as continuous slow rise in temperature) and low level (such as occasional small fluctuations in current) according to the degree of deviation of real-time operating data, the size of potential risks or the degree of urgency.

[0042] By introducing a user selection mechanism, S4-1 to S4-4 combine experience-driven custom rules with data-driven intelligent rules, achieving dual intelligent inspection through "human-machine collaboration," ultimately improving the flexibility, applicability, and intelligence level of the entire inspection system.

[0043] As an optional implementation, after generating the equipment inspection report based on the early warning information, the following steps are also included: S1, extract inspection anomaly data from equipment inspection reports. The inspection anomaly data includes the abnormal equipment identification, the cause of the equipment anomaly, and the warning information level. S2, send the inspection anomaly data to multiple monitoring devices, and determine the first optimization information based on the target account's response information to the multiple inspection anomaly data on the target monitoring devices; S3, obtain the equipment manufacturing information corresponding to the abnormal equipment identifier and the equipment historical operation information corresponding to the first time period; S4. Input the equipment's factory information and historical operating information into the optimization analysis model, and output the second optimization information. The optimization analysis model is a model that has been trained using training samples in advance. S5, optimize the dynamic baseline model based on the first optimization information and the second optimization information.

[0044] In S1-S2 above, the response information includes the judgment result generated by the system after preliminary analysis of the detected equipment anomaly; the first optimization information includes information determined based on the response information for optimizing the dynamic baseline model.

[0045] In S3-S5 above, the first time period is used to indicate the total time from the factory to the most recent completion of the operation of the device corresponding to the abnormal device identifier, and can indicate information such as "device aging"; the second optimization information includes information obtained by the optimization analysis model based on the device factory information and the device's historical operation information for optimizing the dynamic baseline model; through model optimization, the dynamic baseline model can be continuously trained and iteratively improved.

[0046] Through the above S1-S5, the system can achieve root cause inference and feedback. That is, the system will attempt to perform preliminary classification and root cause inference on the detected anomalies (such as suspected "hidden electricity use", "equipment aging", "loose wiring"), and push them to the user in the form of "suspected anomaly" or "low confidence alarm", and collect user feedback (confirmation, ignoring, false alarm) to continuously optimize the model, and build a complete closed loop of "detection-inference-feedback-optimization", which greatly improves the accuracy of anomaly diagnosis and the system's self-learning ability.

[0047] As an optional implementation, the above-mentioned acquisition of a set of historical operating data corresponding to multiple devices to be inspected that conform to preset operating rules includes: S1, filter the reference historical operation data set corresponding to multiple devices to be inspected in the second time period from the preset storage location, which conforms to the preset operation rules. The second time period is a subset of the first time period, and the end timestamp of the second time period is the same as the end timestamp of the first time period. S2, obtain the application scenario corresponding to each device to be inspected, and the types of historical operation data included in the historical operation data set; S3, based on the application scenario and multiple types of historical operational data, derives multiple derived features; S4 will fuse historical operational data sets and derived features to obtain historical operational data sets.

[0048] In S1 above, the second time period can indicate a specific, relatively short time range for analyzing the recent operating status of the equipment. For example, assuming the first time period is from 12:00 on January 2, 2020 to 20:00 on January 2, 2026, and the second time period is from 12:00 on January 2, 2025 to 20:00 on January 2, 2026, the historical operating information of the equipment corresponding to the first time period can be understood as the operating information from the time the equipment was manufactured to the time the equipment was last used. The second time period is the operating information of the equipment for the past year. Through the first time period and the second time period, the current life cycle data (first time period) and recent data (second time period) of the equipment can be distinguished to examine the long-term aging pattern and short-term operating characteristics respectively.

[0049] In S2-S3 above, the application scenarios include the specific application environment of the equipment to be inspected (such as commercial buildings, factories, and residences); feature derivation includes constructing new parameters (derived features) with stronger physical meaning or judgment value based on the original monitoring data through calculation or combination, such as three-phase imbalance, load rate, daily / weekly comparative volatility, etc. Derived features include the types of derived features and the derived feature data corresponding to each type of derived feature.

[0050] In S4 above, data fusion includes integrating and analyzing the reference historical operating data set and derived features to obtain the historical operating data set, such as analyzing the correlation status changes and data fluctuations of load characteristics, seasonal and workday patterns and user operation records (history of opening and closing operations).

[0051] The above-mentioned S1-S4, through in-depth feature derivation based on application scenarios and multiple types of historical operating data, and by integrating the reference historical operating data set and derived features, constructs an extremely rich analytical dimension, which can more comprehensively and accurately depict the real operating status and trends of the equipment to be inspected, and significantly improves the depth and reliability of status assessment and anomaly diagnosis.

[0052] For example, the implementation process of the above-mentioned AI inspection method based on circuit breakers may include the following three stages: 1. First stage (basic functions): Implement the full functionality of the rule-based inspection mode, including flexible rule configuration, reliable scheduling execution, and multi-channel alarms (alarm information can be sent to multiple devices). This stage can immediately meet the customer's specific monitoring needs.

[0053] 2. Second Phase (Dynamic Baseline Model Pilot): Select typical projects or equipment types and deploy the baseline modeling and basic deviation detection functions of the dynamic baseline model inspection. Collect data and verify the effectiveness of the algorithm.

[0054] 3. Third Phase (Full Integration): Improve the dynamic baseline model to achieve correlation analysis and pattern recognition. Allow users to convert abnormal patterns discovered by the dynamic baseline model into fixed rules with a single click, forming a closed-loop operation and maintenance system of "dynamic baseline model discovery - manual confirmation - rule solidification". Display rule alerts and dynamic baseline model warnings uniformly in the management interface.

[0055] According to another aspect of the present invention, an AI inspection system based on circuit breakers is also provided, such as... Figure 2 As shown, it includes: The application and interaction module is used to receive intelligent inspection requests and equipment inspection reports. Intelligent circuit breakers (which can be a single intelligent circuit breaker or a cluster of intelligent circuit breakers) are used to collect real-time operating data from multiple devices to be inspected. The intelligent inspection module (also known as the cloud platform) is used to determine multiple devices to be inspected based on intelligent inspection requests, as well as a first set of inspection rules matching the multiple devices to be inspected; and to obtain a set of historical operating data corresponding to the multiple devices to be inspected that conform to preset operating rules; input the set of historical operating data into the dynamic baseline model and output a second set of inspection rules; and to generate an equipment inspection report based on the real-time operating data corresponding to the multiple devices to be inspected according to the first set of inspection rules and the second set of inspection rules.

[0056] Specifically, the aforementioned application and interaction modules can also provide web portals or mobile applications for users such as operations and maintenance personnel and project administrators, enabling rule configuration, task management, inspection report viewing, and alarm handling. Users can define inspection tasks and configure rules through a user-friendly graphical interface. Key elements include: 1. Equipment to be inspected: Circuit breakers can be selected individually, in groups, or logically divided by project, area, etc.

[0057] 2. Inspection parameters: Supports all measurable parameters of circuit breakers, such as A / B / C three-phase current, line voltage, total active power, switch status, terminal temperature, etc.

[0058] 3. Conditional Logic: Supports a rich set of relational and logical operators to construct complex judgment conditions.

[0059] 4. Numerical judgment: greater than, less than, equal to, within range, etc. (e.g., "air conditioning circuit current <1A").

[0060] 5. Status judgment: equal to or not equal to (e.g., "Main equipment switch status == closed").

[0061] 6. Time setting: Supports triggering inspections at absolute time points (e.g., "14:00 every day"), periodic time (e.g., "09:00 every Monday"), and relative time (e.g., "30 minutes after the equipment is turned on").

[0062] 7. Composite Rules: Supports combining multiple simple conditions using "AND / OR / NOT" to form composite business logic (e.g., "time between 09:00-17:00 and switch status is open and current > 0.5A" is considered abnormal standby).

[0063] Intelligent circuit breakers can use fieldbuses (such as Manton CAN bus) to collect real-time electrical parameters (current, voltage, power, power factor, etc.) and status information (closing / opening, alarm) and send them to a communication module (which can connect multiple intelligent circuit breakers).

[0064] As the core of the system, the intelligent inspection module can provide data access, storage, computing and analysis services. All functional modules of this circuit breaker-based AI inspection system are deployed here, including the rule engine, dynamic baseline model, alarm center and so on.

[0065] 1. Task scheduling: The rule engine triggers inspection tasks on time according to the configured time strategy, and extracts the latest data of the target device from the real-time data pool or time series database.

[0066] 2. Condition Determination: The engine substitutes the data into the rule conditions for calculation and logical determination.

[0067] 3. Alarm Notification: Once a rule is triggered (conditions are met), the system immediately generates an alarm event. Alarms support multi-channel push notifications: platform messages, SMS, email, DingTalk / WeChat Work robots, etc., and different notification strategies can be set according to the rule level.

[0068] The aforementioned AI inspection system based on circuit breakers enables dual intelligent inspection of real-time operational data: firstly, user-customizable rule-based inspection, which transforms user's operational experience and specific management requirements into automatically executable digital rules, meeting customers' precise monitoring needs for specific equipment, specific times, and specific parameters; secondly, intelligent inspection (AI inspection) based on a dynamic baseline model (the normal fluctuation range derived from analyzing historical normal data using machine learning algorithms), which can autonomously learn from historical data and project characteristics to achieve automatic identification and early warning of abnormal patterns. This circuit breaker-based AI inspection system fully leverages the ability of intelligent circuit breakers to collect real-time data such as current, voltage, active power, and switch status, constructing a complete inspection system from the equipment to be inspected to the application and interaction modules.

[0069] As an optional implementation, the above-mentioned circuit breaker-based AI inspection system also includes: The communication module connects multiple smart circuit breakers and smart inspection modules. It is used to decapsulate the source data packets corresponding to the real-time operation data received from the smart circuit breakers, recapsulate them according to the protocol format of the smart inspection module to obtain the target data packets, and send the target data packets to the smart inspection module.

[0070] As an optional implementation method, such as Figure 3 As shown, the above-mentioned intelligent inspection module includes: The decapsulation module is used to decapsulate the target data packet to obtain real-time running data; Storage modules are used to store the device's operational data and derived characteristics; A dynamic baseline model is used to cluster historical operating data sets to obtain clustering results; multiple equipment operating modes are determined based on the clustering results; and a second inspection rule is determined for each type of historical operating data under each equipment operating mode, wherein the set of second inspection rules includes multiple second inspection rules. The rules engine connects the dynamic baseline model and the application and interaction module, and is used to generate equipment inspection reports based on the real-time operating data of multiple devices to be inspected, according to the first inspection rule set and the second inspection rule set.

[0071] The aforementioned intelligent inspection module can discover unknown, potential, or abnormal patterns that are difficult to describe by fixed rules, realizing the transformation from "humans looking for anomalies" to "anomalies looking for people".

[0072] This circuit breaker-based AI inspection system deeply integrates IoT data acquisition and intelligent analysis technologies, achieving the following results: 1. Improve operation and maintenance efficiency: Automate and intelligentize manual periodic meter reading and troubleshooting work, freeing up manpower.

[0073] 2. Dynamic baseline model fusion: Strengthen the deep integration of dynamic baseline models with the system to lay the foundation for future enhanced service charges.

[0074] 3. Optimize energy management: Identify abnormal power consumption and standby energy consumption to provide data support for energy-saving renovations.

[0075] 4. Enable data-driven decision-making: The accumulated knowledge base of equipment operation and anomalies can provide valuable input for product improvement and after-sales service upgrades.

[0076] Driven by both rules and AI (dynamic baseline model), this system enables smart circuit breakers to move from "perception" to "cognition," allowing them to detect anomalies and predict risks from massive amounts of equipment data. This improves the reliability of the power system and the user experience, providing a solid foundation for building a smart power distribution and intelligent energy management system.

[0077] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0078] Obviously, those skilled in the art should understand that the various units or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps into a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0079] The above are merely preferred embodiments of this application and are not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. An AI-based inspection method for circuit breakers, characterized in that, include: Receive intelligent inspection requests, and determine multiple devices to be inspected and a first set of inspection rules matching the multiple devices to be inspected based on the intelligent inspection requests; A smart circuit breaker is used to collect real-time operating data of multiple devices to be inspected, and a set of historical operating data corresponding to multiple devices to be inspected that conform to preset operating rules is obtained. The dynamic baseline model clusters the historical operational data set to obtain clustering results; Based on the clustering results, various equipment operation modes are determined; The historical average runtime of each type of equipment to be inspected is determined based on the historical operational data set. Obtain the preset fluctuation range corresponding to each type of historical operational data; The second inspection rule corresponding to each type of historical operating data is determined based on the historical average runtime, various device operating modes, and the preset fluctuation range. The set of second inspection rules output by the dynamic baseline model includes multiple second inspection rules. Obtain the first inspection rule type corresponding to each first inspection rule in the first inspection rule set, and the second inspection rule type corresponding to each second inspection rule in the second inspection rule set; Based on multiple first inspection rule types and multiple second inspection rule types, multiple first inspection rules are matched with multiple second inspection rules to obtain a first matching rule set and a second matching rule set, wherein the first matching rule set includes multiple matching rule pairs; Obtain the target inspection rule type corresponding to each of the matching rule pairs; Multiple matching rule pairs and multiple target inspection rule types are sent to multiple monitoring devices corresponding to the target account, and the target inspection rule corresponding to each target inspection rule type is determined based on the target account's selection trigger operation on the target monitoring device for multiple matching rule pairs. The real-time operating data corresponding to each of the devices to be inspected is detected according to multiple target inspection rules and the second matching rule set to obtain the detection result; If the detection result indicates that the real-time operating data does not match the target inspection rule and / or the second matching rule set, an early warning message is sent to multiple monitoring devices. Equipment inspection reports are generated based on the aforementioned early warning information.

2. The method according to claim 1, characterized in that, After generating the equipment inspection report based on the early warning information, it also includes: Extract inspection anomaly data from the equipment inspection report. The inspection anomaly data includes the abnormal equipment identifier, the cause of the equipment anomaly, and the warning information level. The inspection anomaly data is sent to multiple monitoring devices, and the first optimization information is determined based on the response information of the target account to the multiple inspection anomaly data on the target monitoring devices. Obtain the device manufacturing information corresponding to the abnormal device identifier and the device historical operation information corresponding to the first time period; The equipment's manufacturing information and historical operating information are input into the optimization analysis model, and the second optimization information is output. The optimization analysis model is a model that has been trained in advance using training samples. The dynamic baseline model is optimized based on the first optimization information and the second optimization information.

3. The method according to claim 2, characterized in that, Obtain a set of historical operating data corresponding to multiple devices to be inspected that conform to preset operating rules, including: The reference historical operation data set corresponding to multiple devices to be inspected within a second time period is selected from a preset storage location and conforms to the preset operation rules. The second time period is a subset of the first time period, and the end timestamp of the second time period is the same as the end timestamp of the first time period. Obtain the application scenario corresponding to each of the devices to be inspected, as well as the types of historical operating data included in the reference historical operating data set; Based on the application scenario and multiple types of historical operational data, feature derivation is performed to obtain multiple derived features; The historical operation data set is obtained by fusing the reference historical operation data set and the derived features.

4. An AI inspection system based on circuit breakers, characterized in that, include: The application and interaction module is used to receive intelligent inspection requests and equipment inspection reports. Intelligent circuit breakers are used to collect real-time operating data from multiple devices to be inspected. The intelligent inspection module is used to determine multiple devices to be inspected and a first set of inspection rules matching the multiple devices to be inspected based on the intelligent inspection request. And acquire the historical operating data sets corresponding to multiple devices to be inspected that conform to the preset operating rules; The intelligent inspection module includes a dynamic baseline model for clustering the historical operating data set to obtain clustering results; determining multiple equipment operating modes based on the clustering results; determining the historical average operating time corresponding to each type of equipment to be inspected based on the historical operating data set; obtaining a preset fluctuation range corresponding to each type of historical operating data; and determining a second inspection rule corresponding to each type of historical operating data based on the historical average operating time, multiple equipment operating modes, and the preset fluctuation range. The second inspection rule set output by the dynamic baseline model includes multiple second inspection rules. The intelligent inspection module further includes a rule engine, connected to the dynamic baseline model and the application and interaction module, used to obtain the first inspection rule type corresponding to each first inspection rule in the first inspection rule set, and the second inspection rule type corresponding to each second inspection rule in the second inspection rule set; based on the multiple first inspection rule types and the multiple second inspection rule types, to perform rule matching between the multiple first inspection rules and the multiple second inspection rules, to obtain a first matching rule set and a second matching rule set, wherein the first matching rule set includes multiple matching rule pairs; to obtain the target inspection rule type corresponding to each matching rule pair; and to match the multiple matching rule pairs... The matching rule pairs and multiple target inspection rule types are sent to multiple monitoring devices corresponding to the target account. Based on the target account's selection trigger operation on the target monitoring device for multiple matching rule pairs, the target inspection rule corresponding to each target inspection rule type is determined. The real-time operating data corresponding to each device to be inspected is detected according to the multiple target inspection rules and the second matching rule set to obtain the detection result. If the detection result indicates that the real-time operating data does not match the target inspection rule and / or the second matching rule set, an early warning information is sent to the multiple monitoring devices. The device inspection report is generated based on the early warning information.

5. The system according to claim 4, characterized in that, The system also includes: The communication module connects multiple smart circuit breakers and the smart inspection module. It is used to decapsulate the source data packets corresponding to the real-time operation data received from the smart circuit breakers, recapsulate them according to the protocol format of the smart inspection module to obtain target data packets, and send the target data packets to the smart inspection module.

6. The system according to claim 5, characterized in that, The intelligent inspection module includes: The decapsulation module is used to decapsulate the target data packet to obtain the real-time running data; Storage modules are used to store the operating data and derived characteristics of the storage device.

Images