AI-based smart energy meter fault diagnosis system

Abstract

This invention relates to the field of intelligent power judgment technology, and more particularly to an AI-based intelligent energy meter fault diagnosis system. The system includes an operational data acquisition module, a jump correlation judgment module, an operational deviation judgment module, a current waveform diagnosis module, and an anomaly alarm module. This invention synchronously acquires current data and cold storage operational data to determine whether current jumps are related to gating actions. Furthermore, by combining multiple cross-validations of baseline energy consumption and temperature changes, it effectively identifies abnormal refrigeration performance in cold storage and the type of intelligent energy meter fault, distinguishing between compressor malfunctions, insufficient airtightness in the cold storage, or faults in the meter itself. An ideal current waveform is constructed using AI algorithms and compared with the measured waveform to achieve intelligent fault type determination and classification. Simultaneously, the diagnostic results are linked with the operation and maintenance system, improving response efficiency. This system is particularly suitable for scenarios with high requirements for meter stability, such as cold chain warehousing, and has good practical value and promising prospects for widespread application.

Description

Technical Field

[0001] This invention relates to the field of intelligent power judgment technology, and in particular to an AI-based intelligent energy meter fault diagnosis system. Background Technology

[0002] As an important component of modern power systems, smart meters not only possess traditional electricity metering functions but also widely integrate data storage, communication management, power quality monitoring, and remote control functions, gradually evolving into intelligent electricity terminals. In various electricity consumption scenarios such as industry, commerce, cold chain warehousing, and residential buildings, smart meters have been widely deployed, supporting the efficient collection and management of distributed electricity consumption data.

[0003] Electricity meter fault diagnosis technology is a key component in ensuring data accuracy and terminal stability within the smart grid system. With the rapid increase in the number of smart meters, monitoring their operational status and identifying faults have become crucial aspects of operation and maintenance management. Fault diagnosis encompasses not only assessing the operational status of internal hardware components such as current sampling modules, voltage conversion modules, control chips, and communication modules, but also multiple technical areas including abnormal data identification, fault type classification, and remote fault location.

[0004] In industrial applications, especially in scenarios with high requirements for power reliability, such as cold chain warehousing, power rooms, and high-speed data centers, electricity meters need to adapt to multiple requirements, including high-frequency start-stop, high-current surges, precise measurement, and continuous monitoring. The frequent changes in current and voltage and the complex start-stop behavior of equipment in these scenarios provide a rich data foundation for fault diagnosis of smart meters and have also driven in-depth research into the practicality, response speed, and embedded implementation methods of diagnostic algorithms.

[0005] With the widespread deployment of smart meters in urban power grids, rural distribution networks, and industrial settings, intelligent fault diagnosis systems that combine artificial intelligence and edge computing technologies have become a research hotspot. The results of these efforts help improve equipment stability, optimize operation and maintenance strategies, and support refined energy management.

[0006] Chinese Patent Publication No. CN119904945A discloses a method and system for prepaid early warning of smart meters, including: real-time collection of data from each smart meter and transmission to a database; periodic transmission of database data to a pre-built TensorFlow; if data is missing, automatic completion of the missing data using TensorFlow based on historical and recent data, and return of the complete data to the database; periodically using TensorFlow to further predict electricity consumption n days later based on historical and recent data, and combining this prediction with the energy balance calculation to send energy consumption warnings and notifications to users. This invention, by combining the data completion and prediction capabilities of a large AI model, compensates for short-term communication failures and can provide personalized energy consumption reminders for different user groups. However, this method has poor data completion accuracy, making it difficult to adapt to complex or sudden changes in electricity consumption; the prediction model relies on a single trend analysis and does not consider individual user differences and environmental changes; the single prediction accuracy model results in weak collaborative optimization and improvement capabilities when facing diverse electricity demand, leading to certain limitations in the adaptability and effectiveness of this solution. Summary of the Invention

[0007] To address this, the present invention provides an AI-based smart energy meter fault diagnosis system to overcome the problems of low data completion accuracy and poor prediction accuracy in existing technologies, which result in weak real-time fault and prediction capabilities.

[0008] To achieve the above objectives, the present invention provides an AI-based smart energy meter fault diagnosis system, comprising:

[0009] Run the data acquisition module to collect current data and operating status data within each standard acquisition cycle, and align them according to a unified timestamp.

[0010] The jump correlation judgment module is connected to the running data acquisition module. It is used to acquire the jump signal in the current curve and determine whether it is synchronized with the opening and closing event of the cold storage door.

[0011] The deviation judgment module is connected to the jump correlation judgment module. It is used to determine whether the number of cold storage opening and closing events in the current standard collection cycle exceeds a set threshold when the jump signal is synchronized with the cold storage door opening and closing event. When the set threshold is exceeded, the offset category comparison result is obtained through baseline energy consumption judgment. When the offset category comparison result is the second offset category comparison result, the expected temperature judgment is performed. The offset category comparison result includes the first offset category comparison result and the second offset category comparison result.

[0012] A current waveform diagnostic module, which is connected to the operation deviation judgment module, is used to perform current waveform diagnosis when the first offset category comparison result is obtained.

[0013] An abnormal alarm module is connected to the jump correlation judgment module, the operation deviation judgment module and the current waveform diagnosis module, and is used to receive the abnormal judgment results output by each module and send corresponding alarm information.

[0014] Furthermore, the operation data acquisition module includes an operation current data extraction unit and an operation sensing data acquisition unit, wherein,

[0015] The current data extraction unit is used to obtain current data within each standard acquisition cycle, including each current value read from the smart meter, the timestamp of each current value, the cumulative power within the standard acquisition cycle, and the current curve of the current value within the standard acquisition cycle.

[0016] The operation perception data acquisition unit is used to collect operation perception data of the cold storage, including cold storage operation time, number of cold storage opening and closing events, timestamps of cold storage opening and closing events, internal temperature of the cold storage, and temperature difference between the inside and outside of the cold storage.

[0017] Furthermore, the transition association judgment module includes a transition detection unit and a transition gating synchronization judgment unit, wherein,

[0018] The jump detection unit is used to determine whether there is a jump signal in the current curve read by the smart meter, and to obtain the jump signal parameters when there is a jump signal.

[0019] The transition signal parameters are the timestamp of each transition signal and the peak value of each transition signal;

[0020] The jump gate synchronization judgment unit calls the timestamp of the alignment processing of the cold storage door opening and closing event and the current jump, maps the two timestamps onto the same time axis, matches the cold storage door opening and closing event and the current jump event in chronological order, and judges whether the current jump is synchronized with the cold storage door opening and closing event based on the matching result of the timestamp of the cold storage opening and closing event and the timestamp of the jump signal.

[0021] Furthermore, the current jump detection unit includes a data detection subunit, a first current jump subunit, and a second current jump subunit, wherein,

[0022] The data detection subunit is used to acquire the actual current data within each standard acquisition cycle and to determine whether the jump signal read by the smart meter is an actual jump signal.

[0023] The first current switching subunit is used to obtain the judgment result of the current acquisition fault of the smart meter when the switching signal read by the smart meter within the standard acquisition period is the actual switching signal;

[0024] The second current switching subunit is used to obtain the fault result of the smart meter display when the switching signal is not the actual switching signal, obtain the actual switching signal and the switching signal, obtain the timestamp of the two signal alignment processing, map the two timestamps onto the same time axis and obtain the time interval of the two signals, so as to dynamically adjust the time deviation of the smart meter.

[0025] Furthermore, the operational deviation judgment module includes a gating action comparison unit, a baseline energy consumption comprehensive judgment unit, and a expected temperature judgment unit, wherein,

[0026] The gate control action comparison unit is used to compare the number of cold storage opening and closing events in the current collection period with the standard opening and closing number threshold and obtain the comparison result. When the comparison result is the first opening and closing comparison result, it performs baseline energy consumption comprehensive judgment, and when the comparison result is the second opening and closing comparison result, it performs current waveform diagnosis.

[0027] The baseline energy consumption comprehensive judgment unit is used to perform baseline energy consumption comprehensive judgment. The judgment process is to analyze the current waveform baseline, obtain the measured offset based on the analysis results, and perform current waveform diagnosis or expected temperature judgment based on the measured offset and offset category.

[0028] The expected temperature judgment unit is used to obtain the temperature values ​​and temperature curves of the temperature values ​​within the standard acquisition period in the cold storage. It calculates the real-time temperature change rate of the temperature curve at different time periods and determines whether the direction of change is negative. It also determines the relationship between the absolute value of the slope of the average temperature change rate of the curve within the standard acquisition period and the preset cooling rate threshold. If the slope is negative and its absolute value is lower than the set threshold, there is an abnormal refrigeration performance fault.

[0029] Furthermore, the baseline energy consumption comprehensive judgment unit includes a baseline offset judgment subunit and an energy consumption verification subunit, wherein,

[0030] The baseline offset determination subunit is used to analyze the changes in the current waveform baseline, identify whether there is an offset, and if so, record the offset as the measured offset.

[0031] The energy consumption verification subunit is used to perform energy consumption verification to determine the offset category and obtain the offset category comparison result. When the first offset category comparison result is obtained, the current waveform diagnosis is performed, and when the second offset category comparison result is obtained, the expected temperature is determined.

[0032] Furthermore, the current waveform diagnostic module includes a waveform fitting unit and a drift judgment unit, wherein,

[0033] The waveform fitting unit is used to model the current within the standard acquisition period based on the gating behavior, construct an ideal current waveform curve through a big data model, and determine whether the actual current curve has experienced baseline drift.

[0034] Among them, the gating behavior includes the number of times the gate is opened and closed and the timestamp corresponding to each opening and closing event;

[0035] The drift judgment unit obtains the ideal current waveform based on the gating behavior, compares the actual current waveform with the ideal current waveform, and analyzes the drift category when the drift comparison result is obtained.

[0036] Furthermore, the drift determination unit includes an environmental instability drift subunit and a magnetic interference drift subunit, wherein,

[0037] The environmental instability offset subunit is used to check the current waveform drift caused by non-steady external factors in the working environment of the smart meter, so as to analyze the cause of the meter failure.

[0038] The magnetic interference offset subunit is used to perform the detection of external magnetic field interference caused by current waveform drift under non-environmental instability conditions.

[0039] Furthermore, the environmental instability offset subunit includes an airtightness detection subunit and a standard opening / closing number threshold adjustment subunit, wherein,

[0040] The airtightness detection subunit is used to detect the temperature difference between the inside and outside of the cold storage and to determine whether there is cold air leakage due to poor airtightness of the cold storage.

[0041] Compare the temperature difference between the inside and outside of the cold storage with the standard temperature difference threshold between the inside and outside of the cold storage. If the temperature difference between the inside and outside of the cold storage is less than or equal to the standard temperature difference threshold between the inside and outside of the cold storage, it indicates that the cold storage is leaking cold air and that the smart meter is experiencing temperature interference.

[0042] The standard opening and closing number threshold adjustment subunit adjusts the standard opening and closing number threshold to 80% of the original standard opening and closing number threshold and rounds it down when the temperature difference between the inside and outside environment is greater than the standard temperature difference threshold between the inside and outside environment.

[0043] Furthermore, the anomaly alarm module includes an alarm notification unit and a linkage operation and maintenance system response unit, wherein,

[0044] The alarm notification unit is used to send alarm information in a timely manner when a fault is detected in the smart meter;

[0045] The linkage operation and maintenance system response unit is used to retrieve the corresponding operation and maintenance adjustment method and feed it back to the user terminal when a fault in a smart meter or a fault in a non-smart meter is detected.

[0046] Compared with existing technologies, the advantages of this invention are as follows: the operation data acquisition module achieves unified timestamp alignment between current data and cold storage operation sensing data, improving data consistency; the jump correlation judgment module extracts current curve jump signals and synchronizes them with cold storage door opening and closing events, reducing the probability of misjudgment; the operation deviation judgment module accurately identifies the operation deviation category under jump synchronization conditions based on gate control action comparison and baseline energy consumption analysis; the current waveform diagnosis module uses ideal waveform fitting and actual waveform comparison technology to analyze the baseline drift position and determine the fault attribution; the environmental instability drift subunit detects changes in temperature difference inside and outside the cold storage, identifies cold energy leakage caused by poor airtightness, and dynamically adjusts the threshold based on the standard number of opening and closing cycles, improving the diagnostic capability for non-steady-state external interference; the abnormal alarm module integrates alarm prompts and operation and maintenance linkage functions to achieve classified response and closed-loop control of smart meters and related external faults, improving fault handling efficiency and operation and maintenance accuracy; the overall system structure is clear, the modules are closely coordinated, and it has good promotion prospects and practical application value.

[0047] Furthermore, this step involves setting up an operating current data extraction unit and an operating perception data acquisition unit to acquire smart meter current data within a standard acquisition cycle, and simultaneously collecting cold storage operating time, opening and closing events, and temperature difference information to construct multi-dimensional operating status characteristics. This provides accurate data support for judging the load changes, usage frequency, and heat exchange behavior of the cold storage during actual operation.

[0048] Furthermore, this step verifies the authenticity of smart meter jump signals by introducing an additional high-precision current sensor, effectively identifying false jump information caused by display errors, data lag, or sampling inaccuracies. Simultaneously, by combining the mapping and alignment of jump timestamps, a unified time benchmark is established, and the sampling start point is dynamically adjusted when time deviations occur, ensuring time synchronization between the smart meter and the actual operating state. This improves the accuracy of jump event identification and the reliability of fault diagnosis, providing a precise basis for system anomaly detection and significantly enhancing the stability and timeliness consistency of the data acquisition link.

[0049] Furthermore, this step effectively distinguishes between reasonable current fluctuations during equipment operation and abnormal deviations caused by meter malfunctions by introducing a dual mechanism of current baseline offset determination and energy consumption verification, thereby improving the system's accuracy in identifying and judging baseline changes. By combining compressor operating characteristic parameters with cold storage operating status to calculate the expected offset and setting a reasonable floating range for comparison, not only is dynamic adaptation to compressor load changes achieved, but the system's fault tolerance in actual operation is also enhanced. Attached Figure Description

[0050] Figure 1This is a schematic diagram of an AI-based smart energy meter fault diagnosis system according to an embodiment of the present invention;

[0051] Figure 2 This is a connection diagram of the operation deviation judgment module in an embodiment of the present invention;

[0052] Figure 3 This is a connection diagram of the baseline energy consumption comprehensive judgment unit in an embodiment of the present invention;

[0053] Figure 4 This is a connection diagram of the current waveform diagnostic module in an embodiment of the present invention. Detailed Implementation

[0054] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are only for explaining the present invention and are not intended to limit the present invention.

[0055] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0056] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0057] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0058] Please see Figure 1 The diagram shown is a schematic of an AI-based smart energy meter fault diagnosis system according to an embodiment of the present invention. The present invention provides an AI-based smart energy meter fault diagnosis system, comprising:

[0059] Run the data acquisition module to collect current data and operating status data within each standard acquisition cycle, and align them according to a unified timestamp.

[0060] The jump correlation judgment module is connected to the running data acquisition module. It is used to acquire the jump signal in the current curve and determine whether it is synchronized with the opening and closing event of the cold storage door.

[0061] The deviation judgment module is connected to the jump correlation judgment module. It is used to determine whether the number of cold storage opening and closing events in the current standard collection cycle exceeds a set threshold when the jump signal is synchronized with the cold storage door opening and closing event. When the set threshold is exceeded, the offset category comparison result is obtained through baseline energy consumption judgment. When the offset category comparison result is the second offset category comparison result, the expected temperature judgment is performed. The offset category comparison result includes the first offset category comparison result and the second offset category comparison result.

[0062] A current waveform diagnostic module, which is connected to the operation deviation judgment module, is used to perform current waveform diagnosis when the first offset category comparison result is obtained.

[0063] An abnormal alarm module is connected to the jump correlation judgment module, the operation deviation judgment module and the current waveform diagnosis module, and is used to receive the abnormal judgment results output by each module and send corresponding alarm information.

[0064] The system's data acquisition module aligns current data with cold storage operation sensing data using unified timestamps, improving data consistency. The jump correlation judgment module extracts current curve jump signals and synchronizes them with cold storage door opening and closing events, reducing the probability of misjudgment. The operation deviation judgment module accurately identifies the type of operation deviation under jump synchronization conditions based on gate control action comparison and baseline energy consumption analysis. The current waveform diagnosis module analyzes baseline drift positions and determines fault attribution using ideal waveform fitting and actual waveform comparison techniques. The environmental instability drift subunit detects temperature differences inside and outside the cold storage, identifies cold energy leakage caused by poor airtightness, and dynamically adjusts thresholds based on standard opening and closing times, improving the diagnostic capability for unsteady external interference. The anomaly alarm module integrates alarm prompts and operation and maintenance linkage functions, enabling classified response and closed-loop control of smart meters and related external faults, improving fault handling efficiency and operation and maintenance accuracy. The system has a clear overall structure, tightly integrated modules, and good prospects for promotion and practical application value.

[0065] Specifically, the operation data acquisition module includes an operation current data extraction unit and an operation sensing data acquisition unit, wherein,

[0066] The current data extraction unit is used to obtain current data within each standard acquisition cycle, including each current value read from the smart meter, the timestamp of each current value, the cumulative power within the standard acquisition cycle, and the current curve of the current value within the standard acquisition cycle.

[0067] The operation perception data acquisition unit is used to collect operation perception data of the cold storage, including cold storage operation time, number of cold storage opening and closing events, timestamps of cold storage opening and closing events, internal temperature of the cold storage, and temperature difference between the inside and outside of the cold storage.

[0068] In this embodiment, the standard acquisition period refers to the time range during which the smart energy meter acquires current data once per second according to a custom period.

[0069] The standard acquisition cycle refers to the complete working time between the compressor restarting after one hibernation state ends and continuing to run until it enters hibernation again; in this embodiment, the standard acquisition cycle is 30 minutes.

[0070] The current curve is a continuous trend graph composed of current values ​​recorded according to the sampling frequency within the standard acquisition period, used to reflect the operating status of the electrical load.

[0071] The cold storage opening and closing event timestamp is based on the instantaneous time of door opening or closing collected by the door magnetic sensor and position switch, which is used to determine the precise time of the opening and closing action;

[0072] The temperature difference between the inside and outside of the cold storage is calculated from the real-time temperature data detected by temperature sensors located on both the inside and outside of the cold storage, and is used to measure the thermal insulation performance and airtightness of the cold storage enclosure structure.

[0073] This step involves setting up an operating current data extraction unit and an operating perception data acquisition unit to acquire smart meter current data within a standard acquisition cycle. Simultaneously, it collects cold storage operating time, opening and closing events, and temperature difference information to construct multi-dimensional operating status characteristics. This provides accurate data support for judging load changes, usage frequency, and heat exchange behavior of the cold storage during actual operation.

[0074] Specifically, the transition association judgment module includes a transition detection unit and a transition gating synchronization judgment unit, wherein,

[0075] The jump detection unit is used to determine whether there is a jump signal in the current curve read by the smart meter, and to obtain the jump signal parameters when there is a jump signal.

[0076] The transition signal parameters are the timestamp of each transition signal and the peak value of each transition signal;

[0077] The jump gate synchronization judgment unit calls the timestamp of the alignment processing of the cold storage door opening and closing event and the current jump, maps the two timestamps onto the same time axis, matches the cold storage door opening and closing event and the current jump event in chronological order, and judges whether the current jump is synchronized with the cold storage door opening and closing event based on the matching result of the timestamp of the cold storage opening and closing event and the timestamp of the jump signal.

[0078] In this embodiment, a standard matching time threshold range is set, the matching event time interval between the matching cold storage door opening and closing event and the current jump event is calculated, and compared with the standard matching event time range;

[0079] If the time interval between matching events is within the standard time interval for matching events, the smart meter is operating normally.

[0080] If the matching event time interval is not within the standard matching event time interval, obtain the actual jump current to determine whether the smart meter has a current acquisition fault.

[0081] The standard matching time threshold range is [0, 2] seconds;

[0082] This step, by setting a standard matching time threshold range and uniformly mapping and comparing the timestamps of cold storage door opening and closing events and current jump events, can accurately determine whether current jumps are synchronized with the actual operating status of the cold storage, thereby effectively identifying abnormal delays or faults in the smart meter during the data collection process, and improving the credibility of current monitoring data and the reliability of system operation.

[0083] Specifically, the current jump detection unit includes a data detection subunit, a first current jump subunit, and a second current jump subunit, wherein,

[0084] The data detection subunit is used to acquire the actual current data within each standard acquisition cycle and to determine whether the jump signal read by the smart meter is an actual jump signal.

[0085] The first current switching subunit is used to obtain the judgment result of the current acquisition fault of the smart meter when the switching signal read by the smart meter within the standard acquisition period is the actual switching signal;

[0086] The second current switching subunit is used to obtain the fault result of the smart meter display when the switching signal is not the actual switching signal, obtain the actual switching signal and the switching signal, obtain the timestamp of the two signal alignment processing, map the two timestamps onto the same time axis and obtain the time interval of the two signals, so as to dynamically adjust the time deviation of the smart meter.

[0087] In this embodiment, the jump signal not being an actual jump signal means that the jump signal recorded by the smart meter did not occur during actual current operation.

[0088] To identify and calibrate this type of fault, the system is equipped with an additional high-precision current sensor to collect actual current data in real time. By comparing the actual switching signal with the timestamp of the switching signal recorded by the smart meter, the time interval between the two is calculated and mapped to the same time axis to determine the degree of time reference offset. When the time deviation exceeds the set threshold, the system corrects the sampling sequence of the smart meter by dynamically adjusting the current data sampling start point, thereby improving the timeliness consistency of the current data and the accuracy of event discrimination, ensuring high reliability in subsequent switching signal identification and operational fault judgment.

[0089] This step verifies the authenticity of smart meter jump signals by introducing an additional high-precision current sensor, effectively identifying false jump information caused by display errors, data lag, or sampling inaccuracies. Simultaneously, by mapping and aligning jump timestamps, a unified time benchmark is established. When time deviations occur, the sampling starting point is dynamically adjusted to ensure time synchronization between the smart meter and its actual operating state. This improves the accuracy of jump event identification and the reliability of fault diagnosis, providing a precise basis for system anomaly detection and significantly enhancing the stability and timeliness of the data acquisition link.

[0090] Please see Figure 2 The diagram shown is a connection schematic of the operation deviation judgment module in an embodiment of the present invention.

[0091] Specifically, the operational deviation judgment module includes a gating action comparison unit, a baseline energy consumption comprehensive judgment unit, and a expected temperature judgment unit, wherein,

[0092] The gate control action comparison unit is used to compare the number of cold storage opening and closing events in the current collection period with the standard opening and closing number threshold and obtain the comparison result. When the comparison result is the first opening and closing comparison result, it performs baseline energy consumption comprehensive judgment, and when the comparison result is the second opening and closing comparison result, it performs current waveform diagnosis.

[0093] The baseline energy consumption comprehensive judgment unit is used to perform baseline energy consumption comprehensive judgment. The judgment process is to analyze the current waveform baseline, obtain the measured offset based on the analysis results, and perform current waveform diagnosis or expected temperature judgment based on the measured offset and offset category.

[0094] The expected temperature judgment unit is used to acquire temperature values ​​and temperature curves of the temperature values ​​within a standard acquisition period in the cold storage; it calculates the real-time temperature change rate of the temperature curve at different time periods and determines whether the direction of change is negative; it also determines the relationship between the absolute value of the slope of the average temperature change rate of the curve within the standard acquisition period and the preset cooling rate threshold; if the slope is negative but its absolute value is lower than the set threshold, the temperature drop rate within the standard acquisition period is insufficient, indicating an abnormal cooling performance fault.

[0095] In this embodiment, the number of cold storage opening and closing events within the current data collection period is compared with the standard opening and closing frequency threshold.

[0096] If the number of cold storage opening and closing events in the current collection period is greater than or equal to the standard opening and closing number threshold, the first opening and closing ratio result is obtained. At this time, the cold storage opening and closing events are frequent, and in this case, current waveform diagnosis is further performed.

[0097] If the number of cold storage opening and closing events in the current collection period is less than the standard opening and closing event threshold, the cold storage door opening and closing events are not frequent. The second opening and closing ratio result is obtained, and the current waveform diagnosis is further performed.

[0098] The standard threshold for the number of opening and closing operations is set at 10.

[0099] Real-time temperature change rate, which is the change in temperature per second;

[0100] The absolute value of the slope of the average temperature change rate of the curve within the standard acquisition period is obtained by performing a first-order linear fit on the temperature curve within the standard acquisition period.

[0101] The initial temperature value is the real-time temperature value at the start of the standard acquisition cycle.

[0102] The absolute value of the slope of the average temperature change rate is the ratio of the difference between the initial temperature value and the expected temperature to the time interval when the initial temperature value drops to the expected temperature; if this time interval is greater than the standard acquisition period, then the absolute value of the slope of the average temperature change rate is the ratio of the difference between the initial temperature value and the real-time temperature at the end of the standard acquisition period to the standard acquisition period.

[0103] The ratio to the standard acquisition cycle;

[0104] The absolute value of the slope of the average temperature change rate is compared with the cooling rate threshold.

[0105] If the absolute value of the slope of the average temperature change rate is greater than or equal to the cooling rate threshold, the cooling performance is normal, but there is a fault in the current acquisition of the smart meter.

[0106] If the absolute value of the slope of the average temperature change rate is less than the cooling rate threshold, there is an abnormal refrigeration performance fault.

[0107] This step introduces a gating action comparison mechanism to dynamically determine the operating status of the cold storage based on the number of cold storage opening and closing events. This effectively identifies potential operational deviations caused by abnormal equipment start-up and shutdown frequencies, improving the proactiveness and accuracy of fault identification. Combining baseline energy consumption analysis and expected temperature change judgment as dual criteria, it can further distinguish between abnormal refrigeration performance and abnormal smart meter data acquisition, avoiding false alarms or missed alarms and improving the accuracy of fault tracing. At the same time, by performing linear fitting analysis on the temperature change curve, it achieves a quantitative assessment of refrigeration efficiency, enabling the system to quickly locate potential fault sources within the sampling period, effectively supporting intelligent diagnosis and operation and maintenance decisions for cold storage.

[0108] Please see Figure 3 As shown, it is a connection diagram of the baseline energy consumption comprehensive judgment unit in an embodiment of the present invention;

[0109] Specifically, the baseline energy consumption comprehensive judgment unit includes a baseline offset judgment subunit and an energy consumption verification subunit, wherein,

[0110] The baseline offset determination subunit is used to analyze the baseline change of the current waveform, identify whether there is an offset, and if so, record the offset as the measured offset.

[0111] The energy consumption verification subunit is used to perform energy consumption verification to determine the offset category and obtain the offset category comparison result. When the first offset category comparison result is obtained, the current waveform diagnosis is performed, and when the second offset category comparison result is obtained, the expected temperature is determined.

[0112] In this embodiment, the compressor operating status is determined. When there is a deviation in the current baseline, the actual cold storage operation perception data is extracted and the expected deviation of the current baseline is calculated by combining the compressor operating coefficient. Based on the expected deviation of the current baseline, the expected deviation fluctuation range is obtained.

[0113] Compare the measured offset with the expected offset fluctuation range.

[0114] If the measured offset is within the expected offset fluctuation range, the first offset category comparison result is obtained. If it is a reasonable offset caused by the normal operation of the compressor, the current waveform diagnosis is performed.

[0115] If the measured offset is not within the expected offset fluctuation range, the second offset category comparison result is obtained. If it is a fault offset caused by compressor failure or smart meter failure, the expected temperature judgment is performed.

[0116] The baseline indicates the static current value of the device under no-load operating conditions;

[0117] The system uses the static current value as a reference and calculates the difference between the baseline of the actual current and the baseline within the standard acquisition period to obtain the measured offset.

[0118] Next, the energy consumption verification subunit, in conjunction with the actual operating conditions of the cold storage, calls the compressor operating coefficient to assess the current consumption trend that the cold storage compressor should have under normal energy consumption. Among them, the compressor operating coefficient takes into account factors that affect the refrigeration load, such as the opening and closing frequency of the cold storage door, and is used to estimate the reasonable current baseline deviation caused by the compressor load fluctuation.

[0119] The formula for estimating the compressor's additional current increment is:

[0120]

[0121] in,

[0122] ΔI(t) represents the additional current increment of the compressor;

[0123] N door (t) represents the number of times the cold storage door was opened within the standard data collection period t;

[0124] k is an empirical coefficient that takes into account the impact of frequent compressor starts caused by door opening and closing. In this embodiment, it is taken as 0.1.

[0125] V room For cold storage volume;

[0126] C compressor The compressor's refrigeration power coefficient reflects the average current consumption caused by refrigeration per unit volume.

[0127] The formula for estimating the additional current increment of the compressor is used to estimate the additional current load required by the compressor to maintain the low temperature due to heat entering the cold storage caused by frequent door openings, which is manifested as a shift in the current baseline.

[0128] The formula for the expected offset of the current baseline is:

[0129] I base (t)=I base_0 (t)+ΔI(t)I base (t) represents the expected offset of the current baseline;

[0130] I base_0 (t) represents the current baseline of the compressor under normal standby conditions;

[0131] ΔI(t) represents the additional current increment of the compressor;

[0132] Based on the expected offset of the current baseline, and setting the floating tolerance range to ±5% of the expected offset;

[0133] The formula for expected deviation of current baseline reflects the expected normal current level of the compressor under high-frequency start-up and shut-down events, and can be used to compare with the measured value to determine whether there is a deviation or fault.

[0134] The measured offset is compared with the expected offset fluctuation range. If the measured offset is within the fluctuation range, the first offset category comparison result is obtained. It is determined to be a reasonable offset under normal equipment operation. The system enters the current waveform diagnosis process to confirm whether there is a baseline offset under external interference.

[0135] If the measured offset exceeds the expected range, the second offset category comparison result is obtained, and the expected temperature judgment is performed to further check whether there are problems such as abnormal cooling performance or sensor acquisition error.

[0136] This step effectively distinguishes between reasonable current fluctuations during equipment operation and abnormal deviations caused by meter malfunctions by introducing a dual mechanism of current baseline offset determination and energy consumption verification, thereby improving the system's accuracy in identifying and judging baseline changes. By combining compressor operating characteristic parameters with cold storage operating status to calculate the expected offset and setting a reasonable floating range for comparison, it not only achieves dynamic adaptation to compressor load changes but also enhances the system's fault tolerance in actual operation.

[0137] Please see Figure 4 The diagram shown is a connection schematic of the current waveform diagnostic module according to an embodiment of the present invention.

[0138] Specifically, the current waveform diagnostic module includes a waveform fitting unit and a drift judgment unit, wherein,

[0139] The waveform fitting unit is used to model the current within the standard acquisition period based on the gating behavior, construct an ideal current waveform curve through a big data model, and determine whether the actual current curve has experienced baseline drift.

[0140] Among them, the gating behavior includes the number of times the gate is opened and closed and the timestamp corresponding to each opening and closing event;

[0141] The drift judgment unit obtains the ideal current waveform based on the gating behavior, compares the actual current waveform with the ideal current waveform, and analyzes the drift category when the drift comparison result is obtained.

[0142] In this embodiment, gate control behavior refers to the opening and closing actions of the door that affect the heat exchange process inside the cold storage. In this invention, gate control behavior includes not only the number of times the door is opened and closed per unit time, but also the timestamp of each opening and closing and its distribution within the standard collection period.

[0143] The ideal current waveform curve refers to the expected waveform of compressor current change under standard cold storage operating environment and specific gating behavior input conditions; this waveform is not actually measured, but is a simulated output result constructed based on big data modeling, and is used as a reference benchmark for the actual current waveform.

[0144] The big data model is built based on historical operating data. By statistically analyzing the mapping relationship between various gating behaviors and current responses in a large number of operating cycles, a mapping model with gating parameters as input and waveform prediction as output is realized.

[0145] In this embodiment, a time-series prediction model based on an LSTM network is preferably used to achieve accurate fitting of the ideal current waveform;

[0146] To determine the degree of deviation between the actual current waveform and the ideal current waveform, the system uses a waveform similarity comparison algorithm;

[0147] In this embodiment, the waveform similarity comparison algorithm is DTW, which can effectively handle the situation where two curves have local rate changes in the time dimension and calculate a similarity score.

[0148] The maximum similarity score is 1.0, and the similarity threshold is set at 0.85.

[0149] The similarity score is compared with the similarity threshold.

[0150] If the similarity score is greater than or equal to the similarity threshold, the waveform has not drifted;

[0151] If the similarity score is less than the similarity threshold, the waveform drifts, and further drift judgment is performed to determine the drift type;

[0152] This step involves setting up a current waveform diagnostic module, constructing an ideal current waveform curve by combining gating behavior, and comparing it with the actual current waveform using a waveform similarity comparison algorithm. This effectively identifies whether the compressor's current baseline has drifted within the standard acquisition cycle, improving the system's ability to detect abnormal energy consumption and meter malfunctions. Compared to traditional methods that rely solely on numerical thresholds, this invention introduces time-series prediction based on an LSTM model and DTW similarity comparison to achieve dynamic modeling and offset identification of current change trends, resulting in higher adaptability and accuracy.

[0153] Specifically, the drift determination unit includes an environmental instability drift subunit and a magnetic interference drift subunit, wherein,

[0154] The environmental instability offset subunit is used to check the current waveform drift caused by non-steady external factors in the working environment of the smart meter.

[0155] The magnetic interference offset subunit is used to perform the detection of external magnetic field interference caused by current waveform drift under non-environmental instability conditions;

[0156] In this embodiment, the operating ambient temperature of the smart meter is obtained and compared with a standard operating ambient temperature threshold.

[0157] If the operating environment temperature of the smart meter is greater than or equal to the standard operating environment temperature threshold, it is further determined that the environmental instability deviation is caused by cold leakage due to poor air tightness of the cold storage or by a drop in the operating environment temperature of the smart meter due to an excessive number of standard opening and closing times.

[0158] If the operating temperature of the smart meter is lower than the standard operating temperature threshold, it indicates interference from an external magnetic field, which is a magnetic interference fault in the smart meter.

[0159] This step, by setting a temperature threshold for the smart meter's operating environment, effectively distinguishes the sources of abnormal current waveforms. On one hand, it can identify environmental temperature instability caused by poor airtightness in cold storage or frequent door control, thus determining whether there are operational anomalies such as cold energy leakage. On the other hand, when the temperature is below the threshold but the waveform has drifted, environmental influences are excluded, directly pointing to external magnetic field interference, thereby accurately identifying abnormalities caused by magnetic interference in the smart meter. This method improves the accuracy of fault tracing and the system's adaptive judgment capability, reduces the false judgment rate, and facilitates intelligent response in subsequent operation and maintenance and early warning mechanisms.

[0160] Specifically, the environmental instability offset subunit includes an airtightness detection subunit and a standard opening / closing number threshold adjustment subunit, wherein,

[0161] The airtightness detection subunit is used to detect the temperature difference between the inside and outside of the cold storage door seam to determine whether there is cold leakage caused by poor airtightness of the cold storage.

[0162] The method involves comparing the temperature difference between the inside and outside of the cold storage door seam with the standard threshold temperature difference between the inside and outside of the door seam.

[0163] If the temperature difference between the inside and outside of the door gap is less than or equal to the standard threshold for the temperature difference between the inside and outside of the door gap, the cold storage will leak cold air, causing the working environment temperature of the smart meter to drop and the image to drift. This is a temperature interference fault of the smart meter.

[0164] If the temperature difference between the inside and outside of the door gap is greater than the standard threshold for the temperature difference between the inside and outside of the door gap, and the cold storage has not leaked cold energy, the problem is that the operating temperature of the smart meter has dropped due to the excessive standard opening and closing times threshold, causing the image to drift, which is a magnetic interference fault of the smart meter.

[0165] The standard opening and closing frequency threshold is reduced to 80% of the original standard opening and closing frequency threshold and rounded down;

[0166] In this embodiment, the standard threshold for the temperature difference between the inside and outside environment at the door seam is set to 20 degrees Celsius, and the temperature difference between the inside and outside environment at the cold storage door seam is the absolute value of the difference.

[0167] Obtain the temperature difference between the inside and outside environment at the door seam within a standard collection period, compare the temperature difference between the inside and outside environment with the standard temperature difference threshold between the inside and outside environment and obtain the comparison result, and reduce the standard opening and closing number threshold to 8 times when the temperature difference between the inside and outside environment is greater than the standard temperature difference threshold between the inside and outside environment.

[0168] This step introduces a detection mechanism for the temperature difference between the inside and outside of the cold storage room at the door seam, enabling accurate assessment of the cold storage's airtightness. This effectively identifies cold air leakage caused by poor airtightness, thus providing early warning and preventing abnormal energy consumption or temperature control failures. Simultaneously, by combining the analysis of the temperature difference at the door seam with the opening and closing frequency, not only can the cause of drift be identified, but the threshold for the number of opening and closing times can also be adaptively adjusted. This effectively reduces the impact of excessively high opening and closing frequencies on the environmental stability of the smart meter, improving the accuracy of current image monitoring and the targeted nature of system operation and maintenance responses.

[0169] Specifically, the anomaly alarm module includes an alarm notification unit and a linkage operation and maintenance system response unit, wherein,

[0170] The alarm notification unit is used to send alarm information in a timely manner when a fault is detected in the smart meter;

[0171] The linkage operation and maintenance system response unit is used to retrieve the corresponding operation and maintenance adjustment method and feed it back to the user terminal when a smart meter fault or a non-smart meter fault is detected.

[0172] In this embodiment, the alarm information includes smart meter display failure, smart meter current acquisition failure, smart meter temperature interference failure, and smart meter magnetic interference failure.

[0173] Once the above alarm information is identified, the joint operation and maintenance system will execute a response operation.

[0174] The coordinated operation and maintenance system response includes fault response for smart meters, fault response for cold storage facilities, and interference response.

[0175] Among them, the fault response for smart meters is the smart meter display fault and the smart meter current acquisition failure fault; the fault response for cold storage is the response to the abnormal refrigeration performance fault; and the interference response is the smart meter temperature interference fault and the smart meter magnetic interference fault.

[0176] If the fault response is related to smart meters, a maintenance response message will be sent to promptly repair or replace the smart meter.

[0177] If it is a cold storage-related fault response, send an operation and maintenance response message to the operation and maintenance personnel to diagnose the cold storage refrigeration system;

[0178] If it is an interference response, then an operational response message will be issued targeting the corresponding source of interference, including:

[0179] If the fault is due to temperature interference, an operation and maintenance response message will be sent to check the temperature difference between the inside and outside of the cold storage and the installation environment of the smart meter.

[0180] If the fault is magnetic interference, a maintenance response message will be sent to check whether there are high-intensity magnetic field interference sources around the smart meter.

[0181] This step categorizes and identifies abnormal alarm information and provides matching response strategies through a coordinated operation and maintenance system, enabling precise location and targeted handling of different types of faults. Compared to traditional single alarm mechanisms, this invention not only achieves rapid response to smart meter and cold storage faults, but also provides detailed intervention paths for external influencing factors such as temperature interference and magnetic interference, improving fault handling efficiency and system stability. It is particularly suitable for vaccine cold storage scenarios with high requirements for temperature control and data acquisition accuracy.

[0182] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0183] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An AI-based intelligent energy meter fault diagnosis system, characterized in that, include, The data acquisition module is used to collect current data and operating status data within each standard acquisition cycle, and align them according to a unified timestamp. The current data includes the current values ​​read from the smart meter, the timestamp of each current value, and the current curve of the current values ​​within the standard acquisition cycle. The operating status data includes the number of cold storage opening and closing events and the timestamp of the cold storage opening and closing events. The jump correlation judgment module is connected to the running data acquisition module. It is used to acquire the jump signal in the current curve, compare the jump signal with the actual jump signal, identify the smart meter display fault or the smart meter current acquisition failure fault, and determine whether the jump signal is synchronized with the cold storage opening and closing event. The operation deviation judgment module, which is connected to the jump correlation judgment module, includes a gating action comparison unit, a baseline energy consumption comprehensive judgment unit, and a expected temperature judgment unit. The gate action comparison unit is used to compare the number of cold storage opening and closing events in the current collection period with the standard opening and closing number threshold when the jump signal is synchronized with the cold storage opening and closing event and obtain the comparison result. When the comparison result is the first opening and closing comparison result, the baseline energy consumption comprehensive judgment is performed, and when the comparison result is the second opening and closing comparison result, the current waveform diagnosis is performed. The baseline energy consumption comprehensive judgment unit includes a baseline offset judgment subunit and an energy consumption verification subunit, wherein, The baseline offset determination subunit is used to analyze the changes in the current waveform baseline, identify whether there is an offset, and if so, record the offset as the measured offset. The energy consumption verification subunit is used to perform energy consumption verification, determine the offset category based on the measured offset, obtain the offset category comparison result, perform current waveform diagnosis when the first offset category comparison result is obtained, and perform expected temperature judgment when the second offset category comparison result is obtained. The expected temperature judgment unit is used to obtain the temperature values ​​and temperature curves of the temperature values ​​within the standard collection period in the cold storage. By calculating the real-time temperature change rate of the temperature curve at different time periods and determining whether its change direction is negative, and by determining the relationship between the absolute value of the slope of the average temperature change rate of the curve within the standard collection period and the preset cooling rate threshold, it is determined whether there is a fault in the current collection of the smart meter or a fault in the refrigeration performance. The current waveform diagnostic module is connected to the operating deviation judgment module. It is used to perform current waveform diagnosis to determine whether there is an abnormality in cold storage cold energy leakage, smart meter temperature interference fault, or smart meter magnetic interference fault. An abnormal alarm module is connected to the jump correlation judgment module, the operation deviation judgment module and the current waveform diagnosis module. It is used to receive the abnormal judgment results output by each module and send corresponding alarm information.

2. The AI-based smart energy meter fault diagnosis system according to claim 1, characterized in that, The operation data acquisition module includes an operation current data extraction unit and an operation sensing data acquisition unit, wherein... The current data extraction unit is used to obtain current data within each standard acquisition cycle, including each current value read from the smart meter, the timestamp of each current value, the cumulative power within the standard acquisition cycle, and the current curve of the current value within the standard acquisition cycle. The operation sensing data acquisition unit is used to collect the operation status data of the cold storage, including the cold storage operation time, the number of cold storage opening and closing events, the timestamp of the cold storage opening and closing events, the temperature inside the cold storage, and the temperature difference between the inside and outside of the cold storage.

3. The AI-based smart energy meter fault diagnosis system according to claim 1, characterized in that, The transition association judgment module includes a transition detection unit and a transition gating synchronization judgment unit, wherein... The jump detection unit is used to determine whether there is a jump signal in the current curve read by the smart meter, and to obtain the jump signal parameters when there is a jump signal. The transition signal parameters are the timestamp of each transition signal and the peak value of each transition signal; The jump gate synchronization judgment unit calls the timestamp of the alignment processing of the cold storage opening and closing event and the current jump, maps the two timestamps onto the same time axis, matches the cold storage opening and closing event and the current jump event in chronological order, and judges whether the current jump is synchronized with the cold storage opening and closing event based on the matching result of the timestamp of the cold storage opening and closing event and the timestamp of the jump signal.

4. The AI-based smart energy meter fault diagnosis system according to claim 3, characterized in that, The current jump detection unit includes a data detection subunit, a first current jump subunit, and a second current jump subunit, wherein... The data detection subunit is used to acquire the actual current data within each standard acquisition cycle and to determine whether the jump signal read by the smart meter is an actual jump signal. The first current switching subunit is used to obtain the judgment result of the current acquisition fault of the smart meter when the switching signal read by the smart meter within the standard acquisition period is the actual switching signal; The second current switching subunit is used to obtain the fault result of the smart meter display when the switching signal is not the actual switching signal, obtain the actual switching signal and the switching signal, obtain the timestamp of the two signal alignment processing, map them onto the same time axis and obtain the time interval of the two signals, so as to dynamically adjust the time deviation of the smart meter.

5. The AI-based smart energy meter fault diagnosis system according to claim 2, characterized in that, The current waveform diagnostic module includes a waveform fitting unit and a drift judgment unit, wherein... The waveform fitting unit is used to model the current within the standard acquisition period based on the gating behavior, construct an ideal current waveform curve through a big data model, and determine whether the actual current curve has experienced baseline drift. Among them, the gating behavior includes the number of times the gate is opened and closed and the timestamp corresponding to each opening and closing event; The drift judgment unit obtains the ideal current waveform based on the gating behavior, compares the actual current waveform with the ideal current waveform for waveform similarity, calculates the similarity score and compares it with the similarity threshold. If the similarity score is greater than or equal to the similarity threshold, the waveform is determined not to have drifted. If the similarity score is less than the similarity threshold, the waveform is determined to have drifted. Then, a drift judgment is performed to obtain the drift type.

6. The AI-based smart energy meter fault diagnosis system according to claim 5, characterized in that, The drift determination unit further includes an environmental instability drift subunit and a magnetic interference drift subunit, wherein... The environmental instability drift subunit is used to check the current waveform drift caused by non-steady external factors in the working environment of the smart meter, so as to analyze the cause of the meter failure. The magnetic interference drift subunit is used to perform the detection of external magnetic field interference caused by current waveform drift under non-environmental instability conditions.

7. The AI-based smart energy meter fault diagnosis system according to claim 6, characterized in that, The environmental instability drift subunit includes an airtightness detection subunit and a standard opening / closing number threshold adjustment subunit, wherein... The airtightness detection subunit is used to detect the temperature difference between the inside and outside of the cold storage and to determine whether there is cold air leakage due to poor airtightness of the cold storage. Compare the temperature difference between the inside and outside of the cold storage with the standard temperature difference threshold between the inside and outside of the cold storage. If the temperature difference between the inside and outside of the cold storage is less than or equal to the standard temperature difference threshold between the inside and outside of the cold storage, it indicates that the cold storage is leaking cold air and that the smart meter is experiencing temperature interference. The standard opening and closing number threshold adjustment subunit adjusts the standard opening and closing number threshold to 80% of the original standard opening and closing number threshold and rounds it down when the temperature difference between the inside and outside environment is greater than the standard temperature difference threshold between the inside and outside environment.

8. The AI-based smart energy meter fault diagnosis system according to claim 1, characterized in that, The anomaly alarm module includes an alarm notification unit and a linkage operation and maintenance system response unit, wherein... The alarm notification unit is used to send alarm information in a timely manner when a fault is detected in the smart meter; The linkage operation and maintenance system response unit is used to retrieve the corresponding operation and maintenance adjustment method and feed it back to the user terminal when a fault in a smart meter or a fault in a non-smart meter is detected.

Images