An abnormal opening and metering mutation linkage identification method of an electric energy metering box

Abstract

The application discloses a kind of abnormal opening and measurement mutation linkage identification method of electric energy metering box, it is related to electric energy metering and power consumption abnormal monitoring technical field.The method includes: through box door sensor, electronic lock interface and electric energy meter acquisition interface, obtain opening, authorization and measurement monitoring data, generate synchronous sequence after time correction;Identify unauthorized effective opening event and determine reference section and influence section, establish measurement reference based on reference section;The influence section is carried out normalization deviation analysis to determine measurement mutation event, calculate opening mutation acceptance degree and measurement loop mismatch degree, exclude data supplement transmission and three-phase synchronous load change after output linkage identification result.Through opening monitoring data, authorization record and measurement monitoring data are time synchronization, reference construction, mutation acceptance analysis and loop mismatch discrimination, the linkage identification accuracy between abnormal opening and measurement mutation can be improved, and false positive caused by data supplement transmission and normal load change is reduced.

Description

Technical Field

[0001] This invention relates to the field of electricity metering and electricity consumption anomaly monitoring technology, specifically a method for linking abnormal opening of an electricity metering box with metering change detection. Background Technology

[0002] As the installation carrier for electricity meters, data acquisition terminals, and metering circuits, electricity metering boxes are typically equipped with door sensors, electronic locks, and electricity meter data acquisition interfaces. These devices acquire metering and monitoring data such as door opening and closing status, lock action records, authorized unlocking records, and voltage, current, active power, and energy consumption readings. This data allows for monitoring of box opening behavior and electricity metering status, providing a data foundation for abnormal electricity consumption, metering circuit intervention, and on-site operation and maintenance management.

[0003] In existing technologies, anomaly monitoring of electricity metering boxes is generally divided into two categories: one is based on the status of the door sensor or electronic lock to determine whether the metering box has been opened without authorization; the other is based on the data collected by the electricity meter to determine whether there are abnormal changes in the current, voltage, power, or energy readings. While these methods can detect opening anomalies or metering anomalies respectively, they typically treat the opening status and metering status as independent data, making it difficult to determine whether metering abrupt changes occur within the scope of the abnormal opening behavior, and also difficult to determine whether the metering abrupt changes conform to the electrical change characteristics after the metering circuit has been interfered with.

[0004] In practical applications, door status data, authorization records, and metering monitoring data often come from different acquisition interfaces, and the sampling timestamps, receiving timestamps, and upload delays of each data point are not consistent. Furthermore, door magnetic vibration, delayed writing of authorization records, missing metering data retransmission, and normal three-phase load synchronization changes can all lead to misjudgments of anomalies. Simply correlating abnormal door openings with metering abrupt changes can easily lead to misinterpreting normal load changes, abnormal acquisition and communication, or metering fluctuations not induced by door openings as metering abrupt changes caused by abnormal door openings.

[0005] Therefore, a method is needed to perform time correction on unpacking monitoring data, authorization records, and metering monitoring data, and to establish metering benchmarks, identify metering mutations, and determine the continuity of unpacking mutations and the mismatch of metering loops based on abnormal unpacking events, so as to achieve the linkage identification between abnormal unpacking and metering mutations. Summary of the Invention

[0006] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a method for identifying abnormal opening of an electricity metering box and sudden metering changes, so as to solve the above-mentioned technical problems.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for linking abnormal opening of an electricity metering box with metering change detection, comprising: S1: Acquire opening monitoring data, authorization records and metering monitoring data through the door sensor, electronic lock interface and energy meter acquisition interface; correct the sampling timestamp and receiving timestamp of the opening monitoring data, authorization records and metering monitoring data to generate an opening metering synchronization sequence. S2: Extract continuous state segments of door opening and closing from the box opening metering synchronization sequence, identify state segments that do not match the authorized record and whose duration meets the valid box opening conditions as abnormal box opening events, and record the box opening start and end points; S3: Use the metering and monitoring data before the start of unpacking as the benchmark segment, and the metering and monitoring data during the lag time from the start of unpacking to the end of unpacking as the influence segment; within the benchmark segment, select sampling points where the cumulative active power is consistent with the increase in electrical energy to establish a metering benchmark. S4: Perform normalized deviation calculation on the metrological monitoring data within the affected section relative to the metrological benchmark, identify sampling sections with continuous deviations in the same direction as metrological mutation events, and calculate the unpacking mutation acceptance degree based on the relative position of the starting point of the metrological mutation event within the affected section. S5: Calculate the continuous difference between the cumulative active power and the incremental power, and the continuous difference between the change in in-phase current and the change in active power within the metering abrupt event to generate the metering loop mismatch. S6: Perform exclusion judgment on missing data transmission and three-phase synchronous load changes, generate exclusion judgment results, and output linkage identification results based on the unpacking sudden change acceptance degree, metering circuit mismatch degree and exclusion judgment results.

[0008] The present invention is further configured such that the step of correcting the sampling timestamps and receiving timestamps of the unpacking monitoring data, authorization records, and metering monitoring data to generate the unpacking metering synchronization sequence includes: The door sensor acquires the door opening status, door closing status, and door status sampling time; the electronic lock interface acquires the lock action time, unlocking method, and authorization identifier; and the energy meter acquisition interface acquires voltage, current, active power, energy readings, and metering sampling time. Add data source identifier, sampling sequence number, sampling timestamp and receiving timestamp to the unpacking monitoring data, authorization record and metering monitoring data respectively, and group and sort them according to the data source identifier to generate unpacking data group, authorization data group and metering data group; The time correction amount is determined based on the correspondence between the sampling timestamp and the receiving timestamp within the same data group. For data with missing or abnormal sampling timestamps, the corrected data group is generated by supplementing and correcting the data according to the time order of adjacent sampling data in the same group. A unified sampling time axis is established based on the sampling period of the metering and monitoring data. The door status, lock action, authorization mark, voltage, current, active power and energy indication in the corrected data group are written into the corresponding sampling time to generate the opening metering synchronization sequence. Set a data gap marker for data items that are not written to the corresponding sampling time, and write the data gap marker to the open-box metering synchronization sequence.

[0009] The present invention is further configured such that, in extracting continuous state segments of door opening and closing from the unpacking metering synchronization sequence, state segments with no matching authorization records and whose duration meets the valid unpacking conditions are identified as abnormal unpacking events, and the recording of the unpacking start and end points includes: Extract the door status field, lock action field, and authorization identifier field from the box opening measurement synchronization sequence, and form a door status subsequence, lock action subsequence, and authorization status subsequence according to a unified time axis; The state subsequence of the cabinet door is stabilized by incorporating instantaneous jumps with a duration shorter than the preset anti-shake duration into adjacent stable states to generate a stable state sequence of the cabinet door. In the stable state sequence of the cabinet door, locate the first transition point from closed state to open state and the second transition point from open state to closed state, and determine the opening holding interval between the first transition point and the second transition point as the candidate open state segment; Based on the first transition point of the candidate open box state segment, the corresponding authorization identifier and unlocking action record are retrieved in the authorized state subsequence and the lock action subsequence. Based on the second transition point of the candidate open box state segment, the corresponding locking action record is retrieved in the lock action subsequence. The candidate open box state segment that matches the authorization identifier, the unlocking action record is valid, and the locking action record satisfies the closure correspondence condition is marked as the authorized open box state segment. Candidate unboxing state segments that are not marked as authorized unboxing state segments and whose duration of the open holding interval reaches the preset effective unboxing time are identified as abnormal unboxing events. The time corresponding to the first transition point is taken as the unboxing start point, and the time corresponding to the second transition point is taken as the unboxing end point to generate an abnormal unboxing event record.

[0010] The present invention is further configured such that the measurement and monitoring data before the start of unpacking is used as the baseline segment, and the measurement and monitoring data within the lag time from the start of unpacking to the end of unpacking is used as the influencing segment, includes: Locate the metering sampling positions corresponding to the start and end points of the unpacking measurement synchronization sequence; Based on the starting point of unpacking, continuous metering and monitoring data are extracted backwards, and metering sampling points with data gaps, timestamp jumps, and power consumption reading rollbacks are removed to generate a benchmark segment; Metering and monitoring data are extracted based on the opening start point, opening end point, and preset lag time. The metering and monitoring data from the opening start point to the opening end point and the metering and monitoring data within the lag time after the opening end point are merged to generate the influence segment.

[0011] The present invention is further configured such that, within the reference segment, selecting sampling points whose cumulative active power is consistent with the increase in electrical energy, and establishing a metering reference includes: Extract the active power, current and energy readings between adjacent sampling times within the reference segment; calculate the cumulative active power based on the active power at adjacent sampling times; and calculate the energy increment based on the change in energy readings at adjacent sampling times. The cumulative active power and the incremental power within the same sampling interval are compared for consistency, and the sampling points that meet the consistency conditions are marked as valid reference sampling points. Based on the effective reference sampling points, in-phase current, active power and energy increment are extracted to generate current reference, active power reference and energy increment reference respectively; The current reference, active power reference, and energy increment reference are combined into a metering reference.

[0012] The present invention is further configured such that the normalized deviation calculation of the metrological monitoring data within the affected segment relative to the metrological benchmark, and the determination of sampling segments with continuous unidirectional deviations as metrological mutation events, includes: Read the in-phase current reference, active power reference, and energy increment reference from the metering reference; extract the in-phase current, active power, and energy increment from the affected section according to the sampling time; and generate the metering sampling sequence of the affected section. The in-phase current, active power, and energy increment at each sampling time in the affected segment metering sampling sequence are processed with the corresponding benchmarks, and then normalized according to the scale of the corresponding benchmarks to generate a point-by-point deviation sequence. Based on the excess state and deviation direction of the deviation amount in the point-by-point deviation sequence, adjacent sampling points are continuously merged, and sampling segments that continuously exceed the deviation threshold and have the same deviation direction are identified as candidate mutation segments. Extract the mutation initiation point, mutation endpoint, mutation duration, mutation phase, and mutation electrical parameter type of candidate mutation segments to generate metrological mutation events.

[0013] The present invention is further configured such that calculating the open-box mutation tolerance based on the relative position of the starting point of the metering mutation event within the affected segment includes: Read the start and end points of the unpacking from the abnormal unpacking event, and read the start and end points of the impact segment from the impact segment. Divide the impact segment into the unpacking process sub-segment and the post-unpacking lag sub-segment. Match the time position of the metering mutation event start point with the start point of the affected segment, the end point of the unpacking process, and the end point of the affected segment to determine the position where the metering mutation event start point falls in the unpacking process sub-segment or the post-unpacking lag sub-segment. Based on the time distance between the starting point of the quantified mutation event and the starting point of the affected segment, normalization is performed within the duration of the affected segment to generate the open-box mutation acceptance degree. When the starting point of a measurement mutation event does not fall into the affected segment, or when the starting point of a measurement mutation event is earlier than the starting point of an abnormal unpacking event, an unpacking mutation acceptance degree that does not meet the acceptance conditions is generated.

[0014] The present invention is further configured such that, within a metering abrupt event, the calculation of the persistent difference between the cumulative active power and the incremental energy, and the persistent difference between the change in in-phase current and the change in active power, to generate the metering loop mismatch degree includes: Read the mutation start point, mutation end point, mutation phase, and mutation electrical parameter type from the metering mutation event, and extract the corresponding mutation event interval from the unpacked metering synchronization sequence to generate a mutation metering subsequence; The active power cumulative amount is formed based on the active power at adjacent sampling times within the mutation measurement subsequence, and the energy increment is formed based on the energy reading at adjacent sampling times. The active power cumulative amount and energy increment within the same sampling interval are differentiated to generate the first continuous difference sequence. Extract the in-phase current and active power corresponding to the abrupt phase from the abrupt metering subsequence, and process the changes in in-phase current and active power at adjacent sampling times to generate a second continuous difference sequence. The continuity of the first and second persistent difference sequences is checked, isolated jump points and missing transmission points are removed, and persistent difference segments are retained. The first and second persistent difference sequences that persist are normalized and aggregated to generate the metrological loop mismatch.

[0015] The present invention is further configured such that the step of generating exclusion results for data missing transmission and three-phase synchronous load changes includes: The sampling timestamp, receiving timestamp, data gap marker, and energy value within the interval corresponding to the metering mutation event are read from the metering synchronization sequence. A continuous record of energy value is formed based on the changes in energy value at adjacent sampling times. Data segments with continuous gaps, out-of-order writing, concentrated shift of receiving time, and cross-period energy value supplementation are identified to generate data missing transmission markers. When no data missing data supplementation marker is generated, the three-phase current and three-phase active power are extracted from the interval corresponding to the metering mutation event, and the current change sequence and active power change sequence are formed according to the phase respectively. The three-phase current change sequence and the three-phase active power change sequence are judged for directional consistency. Metering abrupt events in which the three-phase current and three-phase active power rise and fall synchronously without any single-phase local deviation are marked as three-phase synchronous load changes. The exclusion criteria are generated based on the missing data transmission markers and the three-phase synchronous load change markers.

[0016] The present invention is further configured such that the output of the linkage identification result based on the unpacking mutation acceptance degree, the metering loop mismatch degree, and the exclusion discrimination result includes: When excluding the data missing transmission marker in the judgment result, the abnormal unpacking event and the metrological mutation event are associated and marked as data abnormality pending verification; When the exclusion judgment result includes the three-phase synchronous load change mark, the abnormal unpacking event and the metering change event are associated and marked as load synchronous exclusion; When an abnormal unpacking event exists but no metrological mutation event is generated, the abnormal unpacking event is marked as unpacking without mutation; When a metering mutation event exists and the acceptance rate of the unpacking mutation does not meet the acceptance threshold, the metering mutation event will be marked as a mutation non-unpacking. When the acceptance of unpacking mutation meets the acceptance threshold but the mismatch of metering circuit does not meet the mismatch threshold, the abnormal unpacking event and the metering mutation event are associated and marked as insufficient mismatch for exclusion. When both the unpacking mutation acceptance rate and the metering circuit mismatch rate meet the corresponding thresholds, and the exclusion judgment results do not include data missing transmission markers and three-phase synchronous load change markers, the abnormal unpacking event and the metering mutation event are associated and marked as unpacking linkage mutation. The start point of unpacking, the end point of unpacking, the start point of metering mutation, the phase of metering mutation, the type of mutated electrical parameter, the acceptance degree of unpacking mutation, the degree of metering circuit mismatch, and the result mark are written into the same result record to generate a linkage identification result.

[0017] This invention provides a method for identifying abnormal opening and metering mutations in an electricity metering box. The method acquires opening monitoring data, authorization records, and metering monitoring data through a door sensor, an electronic lock interface, and an electricity meter acquisition interface. It corrects the sampling and receiving timestamps of the opening monitoring data, authorization records, and metering monitoring data to generate an opening metering synchronization sequence. From the opening metering synchronization sequence, it extracts continuous state segments of door opening and closing, identifying state segments without matching authorization records and whose duration meets the valid opening conditions as abnormal opening events, and records the opening start and end points. It uses the metering monitoring data before the opening start point as the baseline segment and the metering monitoring data within the lag time from the opening start point to the opening end point as the influencing segment. Based on the baseline... Within a given section, sampling points where the cumulative active power and the incremental electricity consumption are consistent are selected to establish a metering benchmark. Normalized deviation calculations are performed on the metering monitoring data within the affected section relative to the metering benchmark. Sampling sections with continuous unidirectional deviations are identified as metering abrupt events. The open-box abrupt change tolerance is calculated based on the relative position of the abrupt event's starting point within the affected section. Within the metering abrupt event, the continuous difference between the cumulative active power and the incremental electricity consumption, and the continuous difference between the change in in-phase current and the change in active power are calculated to generate the metering loop mismatch. Data missing data transmission and three-phase synchronous load changes are excluded, generating exclusion judgment results. Based on the open-box abrupt change tolerance, metering loop mismatch, and exclusion judgment results, a linkage identification result is output. The beneficial effects include: 1. Improve the time sequence consistency of unpacking and metering data: By performing time correction on unpacking monitoring data, authorization records and metering monitoring data, a unified unpacking and metering synchronization sequence is formed to avoid errors in judging the sequence of unpacking events and metering change events due to sampling delays and upload delays; 2. Improve the targeting of metrological mutation identification: Divide the benchmark segment and the affected segment by the abnormal opening start and end points, and identify continuous deviations in the same direction within the affected segment based on the metrological benchmark before opening, avoiding the reliance on fixed thresholds to judge metrological anomalies and improving the targeting of metrological mutation identification after opening. 3. Reduce the false alarm rate of linkage identification: By jointly judging the unpacking sudden change acceptance degree and the metering circuit mismatch degree, and excluding data missing transmission and three-phase synchronous load changes, it can distinguish the unpacking linkage sudden change, normal load change and abnormal data collection, thus reducing false alarms and false associations.

[0018] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 The flowchart illustrates an exemplary embodiment of the present invention, which describes a method for linking abnormal opening of an electricity metering box with metering mutation identification. Detailed Implementation

[0020] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0021] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0022] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0023] Example 1: A method for linking abnormal opening of an electricity metering box with sudden metering changes, such as... Figure 1 As shown, it includes: S1: Acquire opening monitoring data, authorization records and metering monitoring data through the door sensor, electronic lock interface and energy meter acquisition interface; correct the sampling timestamp and receiving timestamp of the opening monitoring data, authorization records and metering monitoring data to generate an opening metering synchronization sequence. S2: Extract continuous state segments of door opening and closing from the box opening metering synchronization sequence, identify state segments that do not match the authorized record and whose duration meets the valid box opening conditions as abnormal box opening events, and record the box opening start and end points; S3: Use the metering and monitoring data before the start of unpacking as the benchmark segment, and the metering and monitoring data during the lag time from the start of unpacking to the end of unpacking as the influence segment; within the benchmark segment, select sampling points where the cumulative active power is consistent with the increase in electrical energy to establish a metering benchmark. S4: Perform normalized deviation calculation on the metrological monitoring data within the affected section relative to the metrological benchmark, identify sampling sections with continuous deviations in the same direction as metrological mutation events, and calculate the unpacking mutation acceptance degree based on the relative position of the starting point of the metrological mutation event within the affected section. S5: Calculate the continuous difference between the cumulative active power and the incremental power, and the continuous difference between the change in in-phase current and the change in active power within the metering abrupt event to generate the metering loop mismatch. S6: Perform exclusion judgment on missing data transmission and three-phase synchronous load changes, generate exclusion judgment results, and output linkage identification results based on the unpacking sudden change acceptance degree, metering circuit mismatch degree and exclusion judgment results.

[0024] The present invention is further configured such that the step of correcting the sampling timestamps and receiving timestamps of the unpacking monitoring data, authorization records, and metering monitoring data to generate the unpacking metering synchronization sequence includes: The system acquires the door opening status, door closing status, and door status sampling time through a door sensor; it acquires the lock action time, unlocking method, and authorization identifier through an electronic lock interface; and it acquires voltage, current, active power, energy readings, and metering sampling time through an energy meter acquisition interface. Specifically, a door sensor, an electronic lock interface, and an energy meter acquisition interface are installed inside the energy metering box. The door sensor outputs the door opening status, door closing status, and door status sampling time; the electronic lock interface outputs the lock action time, unlocking method, and authorization identifier; and the energy meter acquisition interface outputs voltage, current, active power, energy readings, and metering sampling time. The door opening and closing statuses are used to generate open-box monitoring data; the lock action time, unlocking method, and authorization identifier are used to generate authorization records; and the voltage, current, active power, energy readings, and metering sampling time are used to generate metering monitoring data. Each data point generates a sampling timestamp at the acquisition end and a reception timestamp when received by the main station or edge terminal, ensuring that the same data point has both occurrence and reception times, providing a time basis for subsequent calibration. Data source identifiers, sampling sequence numbers, sampling timestamps, and receiving timestamps are added to the unpacking monitoring data, authorization records, and metering monitoring data, respectively. The data is then grouped and sorted according to the data source identifier to generate unpacking data, authorization data, and metering data groups. Specifically, the data source identifier distinguishes between the door sensor, electronic lock interface, and energy meter acquisition interface, while the sampling sequence number maintains the chronological order of data within the same source. All data is grouped according to the data source identifier, and then sorted within each group according to the sampling sequence number and sampling timestamp to generate unpacking data, authorization data, and metering data groups. This process ensures that the door status, lock action, and metering status form independent time series, avoiding the confusion of subsequent time correction targets caused by mixed data from different interfaces. The time correction amount is determined based on the correspondence between sampling timestamps and receiving timestamps within the same data group. For data with missing or abnormal sampling timestamps, the data is supplemented and corrected according to the time order of adjacent sampling data in the same group to generate a corrected data group. Specifically, within each data group, the sampling timestamps and receiving timestamps of the data in the same group are extracted, and the time interval between the two is compared to form the transmission delay sequence of the data group. For the current delay value in the transmission delay sequence, the difference between the current delay value and the previous delay value and the next delay value are compared respectively. If the difference between the current delay value and the previous delay value exceeds the delay fluctuation threshold, and the difference between the current delay value and the next delay value also exceeds the delay fluctuation threshold, while the difference between the previous delay value and the next delay value does not exceed the delay fluctuation threshold, then the record corresponding to the current delay value is marked as a delay anomaly. The delay fluctuation threshold is determined based on the clock resolution and sampling period of the corresponding data source interface, and can be taken as the larger of one-tenth of the sampling period and the time corresponding to the interface clock resolution. This process is used to identify situations where the upload delay of a single data item deviates significantly from the stable delay levels before and after; the interval in which the delay value of multiple consecutive records remains within the allowable fluctuation range is defined as the delay stability interval. The delay values ​​within the stable delay interval are sorted, and the median delay value after sorting is used as the time correction value for that data group. The time correction value is the basis for correction based on the actual upload delay of the same data group, and is used to revert data from different sources to a unified occurrence time. For data with missing sampling timestamps, they are filled in according to the sampling order and sampling interval of adjacent data in the same group; for data with sampling timestamps earlier than the previous sampling record or significantly later than the subsequent sampling record, they are corrected according to the time order of adjacent sampling data in the same group to generate a corrected data group. This process allows data with missing, out-of-order, or locally abnormal sampling timestamps to still enter the subsequent synchronization process. A unified sampling time axis is established based on the sampling period of the metering and monitoring data. The door status, lock action, authorization identifier, voltage, current, active power, and energy readings from the corrected data set are written into the corresponding sampling time to generate an opening metering synchronization sequence. Specifically, a unified sampling time axis is established based on the sampling period of the metering and monitoring data, with each sampling period corresponding to a synchronous sampling time. Based on the corrected time of each data point in the corrected data set, the door status, lock action, authorization identifier, voltage, current, active power, and energy readings are written into the corresponding synchronous sampling time. If a door status data point is located between two synchronous sampling times, the closest synchronous sampling time within the allowable time deviation range is written. If multiple records of the same data item exist at the same sampling time, the valid record is generated after retaining the sampling sequence number. This forms a data structure arranged according to a unified time axis, where each row corresponds to a sampling time and each column corresponds to a status or metering field. This data structure is the opening metering synchronization sequence. The opening metering synchronization sequence is the direct input for subsequent abnormal opening event identification, baseline segment extraction, and impact segment analysis. A data gap marker is set for data items not written to the corresponding sampling time. This data gap marker is then written to the opening metering synchronization sequence. Specifically, in the unified sampling timeline, if a corresponding data item is not written at a certain sampling time, a data gap marker is set for that data item, and the data gap marker is written to the opening metering synchronization sequence. The data gap marker originates from the difference between the data items that should be present in the unified timeline and the data items that are actually written, and is used to subsequently eliminate missing data and supplement the transmission. Verification is performed using a specific implementation scenario with a sampling period of 60 seconds: the door sensor collects the open state at 10:05:12, the electronic lock interface records the lock action at 10:05:10, and the energy meter acquisition interface uploads voltage, current, active power, and energy readings at 10:05:00 and 10:06:00 respectively. After time correction, the door open state, lock action, and metering monitoring data from 10:05:00 to 10:06:00 are written to adjacent sampling positions on the same timeline. If the current data at 10:07:00 is not received, the current field at that moment will be marked as missing. This synchronization result allows for the determination of the chronological relationship between changes in the unpacking status and changes in the metering status on the same timeline, while also preserving the location of missing data for subsequent exclusion and determination.

[0025] The present invention is further configured such that, in extracting continuous state segments of door opening and closing from the unpacking metering synchronization sequence, state segments with no matching authorization records and whose duration meets the valid unpacking conditions are identified as abnormal unpacking events, and the recording of the unpacking start and end points includes: The door status field, lock action field, and authorization identifier field are extracted from the opening measurement synchronization sequence. These are then used to form door status subsequences, lock action subsequences, and authorization status subsequences according to a unified timeline. Specifically, after generating the opening measurement synchronization sequence, each sampling moment in the sequence corresponds to a door status field, lock action field, and authorization identifier field. The door status field is taken from the door sensor, with values ​​including closed and open states. The lock action field is taken from the electronic lock interface, with values ​​including unlocking, locking, and no action. The authorization identifier field is taken from the electronic lock authorization record, with values ​​including authorized and not authorized. These fields are read according to the unified timeline to sequentially form the door status subsequence, lock action subsequence, and authorization status subsequence. The unified timeline is formed by the time correction result from the previous processing step; therefore, the three subsequences have the same sampling moment arrangement, allowing for comparison of door status changes, lock actions, and authorization records at the same time point. This process avoids data from different interfaces being out of sync in time, ensuring a consistent data source for opening status judgment. The door state subsequence undergoes state stabilization processing. Instantaneous jumps with durations shorter than a preset anti-shake duration are merged into adjacent stable states to generate a stable door state sequence. Specifically, the door state subsequence is continuously scanned in chronological order. Sampling points with the same state value at adjacent sampling times are merged into a state segment, and the start time, end time, and duration of each state segment are recorded. The preset anti-shake duration is determined by the sampling period of the door sensor. For example, if the sampling period is 1 second, the anti-shake duration can be set to 3 seconds. When the duration of an opening or closing segment is shorter than the anti-shake duration, and the segments before and after it are the same opposite state, the segment is identified as an instantaneous jump and merged into the adjacent stable states. Instantaneous jumps originate from door magnetic vibration, contact vibration, or single-point sampling anomalies. After merging, a stable door state sequence is formed. The calculation logic of this step is: state segment merging—duration check—instantaneous jump merging. The generated stable door state sequence retains the continuously existing opening and closing states, reducing short-term false triggers that enter the abnormal opening identification process. In the stable state sequence of the cabinet door, the first transition point from closed to open and the second transition point from open to closed are located. The open-hold interval between the first and second transition points is determined as a candidate open state segment. Specifically, in the stable state sequence of the cabinet door, the sampling position where the state transitions from closed to open is found along the time direction, and this sampling position is determined as the first transition point. The sampling position where the state transitions from open to closed is then found, and this sampling position is determined as the second transition point. The continuous open interval between the first and second transition points is determined as a candidate open state segment. The duration of the candidate open state segment is obtained by the time difference between the time corresponding to the second transition point and the time corresponding to the first transition point. The first transition point represents the starting position of the cabinet door entering the open-hold interval, and the second transition point represents the position of the cabinet door exiting the open-hold interval; both together define the time range of the candidate open state segment. The candidate open state segment formed in this step is passed to the authorization matching process as the basis for determining authorized or abnormal opening. Based on the first transition point of the candidate unlocking state segment, the corresponding authorization identifier and unlocking action record are retrieved in the authorization state subsequence and the lock action subsequence. Based on the second transition point of the candidate unlocking state segment, the corresponding locking action record is retrieved in the lock action subsequence. Candidate unlocking state segments that match the authorization identifier, have valid unlocking action record, and whose locking action record meets the closure correspondence condition are marked as authorized unlocking state segments. Specifically, using the time corresponding to the first transition point of the candidate unlocking state segment as the retrieval center, the authorization identifier corresponding to that time is retrieved in the authorization state subsequence, and the unlocking action record adjacent to that time is retrieved in the lock action subsequence. Using the time corresponding to the second transition point of the candidate unlocking state segment as the retrieval center, the locking action record adjacent to that time is retrieved in the lock action subsequence. The authorization matching interval can be determined based on the valid time of the electronic lock authorization record. For example, if the valid time of the authorization identifier covers 60 seconds before and after the first transition point, and there is an unlocking action corresponding to the first transition point in the lock action field, the candidate unlocking state segment is considered to have matched a valid authorization. The closed corresponding interval can be determined based on the allowable time deviation between the door closing action and the electronic lock locking action. For example, if a locking action occurs within 60 seconds before or after the second transition point, the candidate opening state segment is considered to form a closed corresponding record. If an authorization identifier exists in the authorized state subsequence, but there is no valid unlocking action corresponding to the first transition point in the lock action subsequence, then the authorization identifier is not considered a valid authorization matching result. The candidate opening state segment that matches a valid authorization is marked as an authorized opening state segment, and the locking action retrieval result is written into the closed corresponding field of the state segment. This step, through the correspondence between the authorization identifier, unlocking action, and locking action, establishes a traceable association between the door opening boundary, the door closing boundary, and the lock action record, avoiding classifying candidate opening state segments as authorized opening based solely on authorization records or a single lock action. Candidate opening state segments that are not marked as authorized opening state segments and whose opening hold interval duration reaches the preset effective opening time are identified as abnormal opening events. An abnormal opening event record is generated by using the time corresponding to the first transition point as the opening start point and the time corresponding to the second transition point as the opening end point. Specifically, the duration of candidate opening state segments not marked as authorized opening state segments is determined. The preset effective opening time is determined by the sampling period of the door sensor and the minimum duration requirement of the metering box opening action. For example, when the sampling period is 1 second, the effective opening time can be set to 10 seconds. When the opening hold interval duration of a candidate opening state segment reaches this effective opening time, the candidate opening state segment is identified as an abnormal opening event. The time corresponding to the first transition point is written into the abnormal opening event record as the opening start point, and the time corresponding to the second transition point is written into the abnormal opening event record as the opening end point. Simultaneously, the opening hold interval duration, authorized matching status, and lock action matching status are written into the record to generate an abnormal opening event record. In a specific implementation scenario, the stable state sequence of the cabinet door changes from closed to open at 10:05:12 and from open to closed at 10:05:46, with the open holding interval lasting 34 seconds. The authorized state subsequence has no authorized identifier between 10:04:12 and 10:06:12, and the lock action subsequence has no valid unlocking action within the adjacent time range. Therefore, this open holding interval meets the valid opening duration but does not match authorization, generating an abnormal opening event record with an opening start at 10:05:12 and an opening end at 10:05:46. This record is passed to subsequent baseline and impact segment division steps to determine the pre-opening measurement monitoring data and the post-opening impact range.

[0026] The present invention is further configured such that the measurement and monitoring data before the start of unpacking is used as the baseline segment, and the measurement and monitoring data within the lag time from the start of unpacking to the end of unpacking is used as the influencing segment, includes: In the unpacking metering synchronization sequence, locate the metering sampling positions corresponding to the unpacking start and end points; specifically, in the unpacking metering synchronization sequence, use the metering sampling time on a unified time axis as the positioning object, and read the unpacking start point from the abnormal unpacking event record. And the end of unpacking Search for the corresponding sampling time sequence in the measurement and monitoring data. and The sampling positions with the closest time distance are used to generate the metering sampling positions for the start and end of the unpacking process. The metering sampling position refers to the sequence number of a metering monitoring data point on a unified time axis, derived from the sampling time after the electricity meter's data acquisition interface writes the metering synchronization sequence upon unpacking. In specific processing, if the metering sampling period is... The sampling time sequence is The sampling time that is closest to the start time of the unpacking process is taken as the start time position, and the sampling time that is closest to the end time of the unpacking process is taken as the end time position. When the start time of the unpacking process is between two metering sampling times, the sampling position that is no later than the start time of the unpacking process and is the closest to it is selected as the end boundary of the benchmark segment to prevent metering data after unpacking from entering the benchmark segment. Thus, the time boundary of the abnormal unpacking event is converted into the sampling boundary in the metering monitoring data, and the subsequent benchmark segment and the affected segment are clearly defined as the metering sampling points on a unified time axis. Based on the starting point of unpacking, continuous metering and monitoring data is extracted backwards. Metering sampling points with data gaps, timestamp jumps, or energy value regressions are removed to generate a benchmark segment. Specifically, continuous metering and monitoring data is extracted backwards from the metering sampling position at the start of unpacking, with the extraction length determined according to the metering sampling cycle, ensuring the benchmark segment includes several consecutive valid sampling points before unpacking. The benchmark segment refers to the metering and monitoring data interval used to establish the metering benchmark before the start of unpacking. Its input fields include voltage, current, active power, and energy value. After extraction, the metering sampling points within the benchmark segment are screened for validity: sampling points with data gaps are removed; sampling points with reversed or abnormally spaced sampling timestamps relative to adjacent sampling points are removed; sampling points where the current energy value is less than the energy value at the previous valid sampling time are identified as energy value regression points and removed. Energy value regression refers to a reverse change in the cumulative energy value without meter replacement, zeroing, or data replenishment markings, and is generally not considered a stable metering state for establishing a benchmark. After removal, the remaining consecutive valid sampling points form the benchmark segment. Using sample data with a sampling period of 60 seconds as an example, if the sampling point corresponding to the opening start is 10:05:00, then the metering and monitoring data from 10:00:00 to 10:04:00 is extracted. However, the current data at 10:02:00 has a missing marker, and the energy reading at 10:03:00 is less than the previous valid energy reading at 10:02:00; therefore, the corresponding sampling point is not included in the reference segment. The reference segment obtained through the above processing maintains time continuity and the monotonicity of energy readings, avoiding the impact of abnormal sampling points on the establishment of subsequent metering references. Metering monitoring data is extracted based on the opening start point, opening end point, and a preset lag time. The metering monitoring data from the opening start point to the opening end point, and the metering monitoring data within the lag time after the opening end point, are combined to generate an impact segment. Specifically, the impact segment is the range of metering monitoring data used to identify short-term metering changes during and after the opening process. It consists of metering monitoring data during the opening process and the delayed metering monitoring data after the opening end point. The preset lag time is derived from the sampling period of the metering monitoring data, the receiving timestamp correction result, and the data reporting interval of the electricity meter acquisition interface. It covers the time range within which the metering status is delayed in being reflected in the sampled data after the opening operation. In specific processing, the metering sampling position corresponding to the opening start point is taken as the starting point of the influence segment, and the metering sampling position corresponding to the preset lag time after the opening end point is taken as the ending point of the influence segment. Voltage, current, active power, and energy readings within this range are extracted. The metering monitoring data from the opening start point to the opening end point is determined as the opening process data, and the metering monitoring data within the lag time after the opening end point is determined as the lag observation data, and these are merged to generate the influence segment. Taking a sample data with a sampling period of 60 seconds, an opening start point corresponding to 10:05:00, an opening end point corresponding to 10:06:00, and a preset lag time of 3 sampling periods as an example, the influence segment extracts the metering monitoring data from 10:05:00 to 10:09:00, where 10:05:00 to 10:06:00 is the opening process data, and 10:07:00 to 10:09:00 is the lag observation data. The impact segment generated by this step incorporates the direct and delayed changes following the abnormal unpacking event into the same analysis interval, and serves as input for subsequent normalized deviation calculations and quantitative mutation event identification.

[0027] The present invention is further configured such that, within the reference segment, selecting sampling points whose cumulative active power is consistent with the increase in electrical energy, and establishing a metering reference includes: The active power, current, and energy readings between adjacent sampling times are extracted from the reference segment. The cumulative active power is calculated based on the active power at adjacent sampling times, and the energy increment is calculated based on the change in energy readings at adjacent sampling times. Specifically, the reference segment is read from the unpacking metering synchronization sequence. The metering monitoring data within the reference segment includes in-phase current, active power, and energy readings. Adjacent sampling times are extracted sequentially according to a unified time axis. and The active power, current, and electrical energy values ​​are displayed, where the active power is denoted as... and In-phase current is denoted as and The electrical energy reading is recorded as and , This indicates the phase difference. The cumulative active power within the sampling interval is calculated based on the active power at adjacent sampling times. The calculation method involves multiplying the active power at adjacent sampling times by the sampling interval after performing interval equivalence processing. The formula is as follows: ,in The time interval between adjacent sampling times is defined and converted to kilowatt-hours according to the unit of the electricity meter. The electricity increment is calculated based on the change in the electricity reading at adjacent sampling times. This step creates a set of comparable cumulative active power and energy increments within the same sampling interval, making the data source clear for subsequent consistency assessments. The cumulative active power and the increase in electrical energy within the same sampling interval are compared for consistency. Sampling points that meet the consistency criteria are marked as valid reference sampling points. Specifically, for each adjacent sampling interval, the cumulative active power is... With the increase in electrical energy Consistency judgment is performed. Specifically, the difference between the accumulated active power and the incremental energy is first calculated; then, the allowable error for metering accuracy is determined based on the meter's accuracy class; the allowable error for quantization of the indicated value is determined based on the minimum resolution of the energy reading; and the allowable error for time alignment is determined based on the remaining time deviation after sampling timestamp correction. The allowable error for time alignment is converted into energy by the active power and remaining time deviation within the sampling interval. The allowable errors for metering accuracy, quantization, and time alignment are combined to form the allowable difference range corresponding to the sampling interval; when the difference between the accumulated active power and the incremental energy is within the allowable difference range, the next sampling point corresponding to that sampling interval is marked as the valid reference sampling point. To avoid discrimination distortion due to an excessively small denominator under low load conditions, the power-energy consistency deviation between the two is calculated: ,in, This represents the relative difference between the cumulative active power and the increase in electrical energy. The protection quantity corresponding to the minimum resolution of the electricity meter is used. When using power-energy consistency deviation for judgment, the larger of the absolute value of the cumulative active power and the protection quantity corresponding to the minimum resolution of the electricity meter reading is used as the normalization scale. The allowable difference range is divided by this normalization scale to obtain the consistency threshold. The difference between the cumulative active power and the electricity increment is divided by the normalization scale to obtain the power-energy consistency deviation. When the power-energy consistency deviation is not greater than the consistency threshold, the corresponding sampling point is marked as a valid reference sampling point; when the power-energy consistency deviation is greater than the consistency threshold, it indicates that the cumulative active power and the electricity increment within the sampling interval do not meet the reference establishment requirements, and the corresponding sampling point is not included in the metering reference establishment process. Therefore, the consistency threshold is obtained by converting the electricity meter accuracy level, the electricity reading resolution, and the sampling time deviation through the electricity scale within the same sampling interval, which can cover normal metering errors, reading quantization errors, and sampling alignment errors. This processing can eliminate sampling points where the power accumulation and electricity reading changes are inconsistent, so that the data entering the metering reference establishment process maintains a basic consistent metering relationship. Based on the effective reference sampling points, in-phase current, active power, and energy increment are extracted to generate current reference, active power reference, and energy increment reference, respectively. Specifically, after marking the effective reference sampling points, in-phase current, active power, and energy increment are extracted from these points. For the same phase... The in-phase currents in the effective reference sampling points are processed centrally to generate a current reference. The active power in the effective reference sampling points is centrally processed to generate an active power reference. The energy increments corresponding to the valid reference sampling points are centrally processed to generate an energy increment reference. Centralized processing can use the median value or a representative value after removing extreme values, where the median value is used to reduce the impact of a single fluctuating sampling point on the benchmark result. If the effective benchmark current of phase A in the benchmark section is 4.8A, 4.9A, 5.0A, and 4.9A, then the benchmark current of phase A can be taken as 4.9A; if the effective active power is concentrated around 3.0kW, then the active power benchmark takes the corresponding concentrated value. Through this step, the metering status before unpacking is organized into benchmark quantities that can be used for comparison of affected sections; The current reference, active power reference, and energy increment reference are combined into a metering reference. Specifically, the current reference, active power reference, and energy increment reference for each phase are combined according to phase and metering field to generate the metering reference. ,in, Corresponding to the current reference of each phase, Corresponding to active power benchmark, The corresponding energy increment benchmark is written into the subsequent normalized deviation calculation process for the affected section. This benchmark is used to calculate the deviation of in-phase current, active power, and energy increment within the affected section relative to the state before unpacking. Taking verification data with a sampling period of 60 seconds as an example, the cumulative active power in the five sampling intervals within the benchmark section are 0.050, 0.051, 0.050, 0.049, and 0.052 kWh, corresponding to energy increments of 0.050, 0.051, 0.049, 0.080, and 0.052 kWh, respectively. The fourth sampling interval exceeds the consistency threshold and is not marked as a valid benchmark sampling point; the remaining sampling points are used to form the current benchmark, active power benchmark, and energy increment benchmark. The metering benchmark obtained through this process can serve as a reference input for identifying metering abrupt changes in the affected section.

[0028] The present invention is further configured such that the normalized deviation calculation of the metrological monitoring data within the affected segment relative to the metrological benchmark, and the determination of sampling segments with continuous unidirectional deviations as metrological mutation events, includes: The system reads the in-phase current benchmark, active power benchmark, and energy increment benchmark from the metering reference. It then extracts the in-phase current, active power, and energy increment from the affected section according to the sampling time, generating a metering sampling sequence for the affected section. Specifically, in the embodiment of abnormal opening linkage identification of the energy metering box, the monitoring objects are the three-phase energy meter, the box door status sensor, and the electronic lock control unit installed inside the metering box. The energy meter acquisition interface uses RS485, HPLC, or carrier communication interface to read the in-phase current, active power, and active energy readings, with a sampling period set to 60 seconds. The box door status sensor and electronic lock control unit are used to determine the start and end points of the abnormal opening. The affected section is the metering data interval from the start of the abnormal opening to the end point after a preset lag time. For example, if the start point of the opening is T0, the end point is T1, and the lag time is 5 minutes, then the affected section is from T0 to T1+5 minutes. The metering references are derived from reference segment data that has undergone consistency screening before the start of unpacking. Among them, the in-phase current reference represents the current reference value of the same phase under normal conditions, the active power reference represents the power reference value of the same phase under normal conditions, and the energy increment reference represents the normal increase in active energy readings within adjacent sampling periods. After reading the metering references, the data is processed according to the sampling time. , ... The current, active power, and active energy increment values ​​of phases A, B, and C are extracted sequentially to form a metering sampling sequence for the affected section. This process ensures that subsequent mutation judgments have a unified time axis and a unified phase object, avoiding the mixing and comparison of data from different sampling times or different phases, and improving the traceability of metering mutation location after abnormal unpacking. For each sampling moment in the affected section's metering sampling sequence, the in-phase current, active power, and energy increment are subtracted from the corresponding benchmark. These values ​​are then normalized according to the benchmark's scale to generate a point-by-point deviation sequence. Specifically, for each sampling point in the affected section's metering sampling sequence, missing value filtering, duplicate sampling removal, and timestamp alignment are performed. If any one of the current, active power, or active energy readings is missing from the same sampling point, that sampling point is marked as invalid and does not participate in the abrupt change continuity determination. For valid sampling points, the in-phase current, active power, and energy increment of the current phase are subtracted from the corresponding benchmark, and then normalized according to the benchmark's scale to form a point-by-point deviation sequence. (The last sentence appears to be incomplete and possibly refers to a separate process.) Sampling time Electrical parameter types Represents current Active power or increase in electrical energy Then the normalized deviation can be expressed as: ,in, To affect the phase distinction within the segment exist The measured electrical parameter values ​​at time [time]. For the same phase and the same electrical parameter type, To prevent abnormal amplification when the reference value approaches zero, a scale protection quantity is used. The point-by-point deviation sequence consists of the phase, time, electrical parameter type, deviation direction, and deviation magnitude for each sampling point. The deviation direction is determined by… The sign is determined, and the deviation range is determined by Characterization. Through normalization, the increments of current, power, and electrical energy with different dimensions are converted into comparable relative deviations, which can avoid unreasonable amplification of a certain type of electrical parameter in the judgment due to different units or magnitudes; Based on the exceeding state and direction of deviation in the point-by-point deviation sequence, adjacent sampling points are continuously merged, and sampling segments that continuously exceed the deviation threshold and have the same deviation direction are identified as candidate mutation segments. Specifically, after the point-by-point deviation sequence is generated, continuous scanning is performed according to phase and electrical parameter type. Let the deviation threshold be θ, which can be determined based on the normal fluctuation range within the reference segment. For example, the upper bound of the normalized deviation in the reference segment can be extended to the threshold, or a relative deviation range of 0.20 to 0.30 can be taken in engineering implementation. For each phase and each electrical parameter type, if a sampling point satisfies... If the deviation direction of adjacent excess points is the same, and there are no consecutive invalid points or direction reversal points in between, then the adjacent excess points are merged into the same continuous sampling segment. To reduce false alarms caused by single-point noise, the continuous sampling segment must also meet a minimum duration condition, such as a duration of no less than 2 sampling cycles; continuous sampling segments that meet this condition are identified as candidate mutation segments. For example, if the active power reference for phase A is 1.10kW, and the active power of four consecutive sampling points within the affected segment is 0.78kW, 0.74kW, 0.73kW, and 0.75kW respectively, after normalization, the deviation direction is negative, the deviation amplitude exceeds 0.25, and it lasts for 4 minutes, then this segment is merged into a negative candidate mutation segment for phase A active power. This continuous merging logic can retain the time continuity characteristics of abnormal changes and reduce the interference of instantaneous communication jitter and single-point sampling errors on the metering mutation identification results. The measurement mutation event is generated by extracting the mutation initiation point, mutation endpoint, mutation duration, mutation phase, and mutation electrical parameter type of candidate mutation segments. Specifically, for each candidate mutation segment, the first out-of-limit sampling point within the segment is read as the mutation initiation point, and the last consecutive out-of-limit sampling point is read as the mutation endpoint. The mutation duration is calculated based on the mutation initiation point and the mutation endpoint. The mutation phase is recorded according to the phase to which the candidate mutation segment belongs, and the mutation electrical parameter type is recorded according to the electrical parameter type to which the candidate mutation segment belongs. This generates a measurement mutation event, and the event content includes at least the event number, mutation initiation point, mutation endpoint, mutation duration, mutation phase, mutation electrical parameter type, deviation direction, and maximum normalized deviation. To verify this, consider the following example: In the baseline section, the current baseline for phase A is 5.00A, the active power baseline for phase A is 1.10kW, and the energy increment baseline per minute is 0.018kWh. During the abnormal opening of the box, the active power of phase A decreases to 0.73kW, 0.74kW, 0.73kW, and 0.75kW respectively for four consecutive sampling cycles starting from T0+2min. The normalized deviation ranges from 0.29 to 0.34, exceeding the deviation threshold of 0.25. The deviation direction is negative. Therefore, the output metering abrupt event is as follows: abrupt start point T0+2min, abrupt end point T0+6min, abrupt duration of 4min, abrupt phase A, abrupt electrical parameter type active power, and a deviation direction negative. Compared to using only single-point over-limit judgment, this embodiment uses a closed-loop process of benchmark reading—normalized deviation—continuous merging in the same direction—event feature extraction to enable the measurement mutation result to simultaneously have time location, phase location, and electrical parameter type location, which facilitates subsequent continuity calculation and linkage identification with abnormal unpacking events.

[0029] The present invention is further configured such that calculating the open-box mutation tolerance based on the relative position of the starting point of the metering mutation event within the affected segment includes: The process involves retrieving the start and end points of the unpacking process from the abnormal unpacking event, and the start and end points of the affected segment from the affected segment. The affected segment is then divided into an unpacking process sub-segment and a post-unpacking lag sub-segment. Specifically, after identifying the metrological mutation event, the input objects include the abnormal unpacking event, the affected segment, and the metrological mutation event. The start point of the unpacking process is recorded in the abnormal unpacking event. And the end of unpacking The affected segment records the starting point of the affected segment. and the end point of the affected section The starting point of the affected segment corresponds to the starting point of the unpacking process, and the ending point of the affected segment is obtained by adding a preset lag time to the ending point of the unpacking process. Based on the above time boundaries, the affected segment is divided into unpacking process sub-segments. and the delayed sub-segment after unpacking The unpacking process segment corresponds to the metering changes while the box door remains open, and the post-unpacking lag segment corresponds to the delayed changes in metering data after the box door is closed. This step converts the time range of abnormal unpacking events into a time window for subsequent judgment, giving a clear reference for the location of metering abrupt events. The starting point of the metering mutation event is matched with the starting point of the affected segment, the end point of the unpacking process, and the end point of the affected segment in terms of time position to determine the position where the starting point of the metering mutation event falls within the unpacking process sub-segment or the post-unpacking delayed sub-segment; specifically, the mutation starting point is read from the metering mutation event. ,Will sequentially with the starting point of the affected segment End of unpacking and the end point of the affected section Perform time and location matching. When When, the starting point of the measurement mutation event is marked as falling into the unpacking process sub-segment; when When, the starting point of the measurement mutation event is marked as falling into the post-opening hysteresis sub-segment; when or At that time, the starting point of the measurement mutation event is marked as not falling into the affected segment. The falling position mark is written into the measurement mutation event record to distinguish whether the mutation occurred during the unpacking operation or the lagged observation period, and is passed to the unpacking mutation acceptance degree calculation process; Based on the time distance between the starting point of the quantified mutation event and the starting point of the affected segment, a normalization process is performed within the duration of the affected segment to generate the open-box mutation acceptance degree. Specifically, when the starting point of the quantified mutation event falls into the affected segment, normalization is performed based on the time distance between the mutation starting point and the starting point of the affected segment. Let the duration of the affected segment be... The time distance between the mutation initiation point and the initiation point of the affected segment is Then the acceptance of sudden changes upon unpacking It can be generated using the following formula: ,in, The value range is limited to between 0 and 1. The larger the value, the closer the starting point of the measurement mutation event is to the starting position of the abnormal unpacking event; The smaller the value, the closer the starting point of the measured mutation event is to the end of the affected segment. This scaling process unifies the mutation locations under different open-box durations and different lag times to the same dimension, facilitating subsequent comparison with the acceptance threshold. The generated open-box mutation acceptance degree, along with the measured mutation event number, mutation starting point, and landing position marker, is written into the event record. When the starting point of a measurement mutation event does not fall within the affected segment, or when the starting point of the measurement mutation event is earlier than the opening start point of an abnormal opening event, an opening mutation acceptance degree that does not meet the acceptance conditions is generated. Specifically, when the starting point of a measurement mutation event does not fall within the affected segment, or when the starting point of the measurement mutation event is earlier than the opening start point of an abnormal opening event, the opening mutation acceptance degree is set to a value that does not meet the acceptance conditions, and the flag indicating that the acceptance conditions are not met is written into the measurement mutation event record. As a verification example, if the opening start point is 10:05:00, the opening end point is 10:06:00, the preset lag time is 3 sampling periods, and each sampling period is 60 seconds, then the end point of the affected segment is 10:09:00, and the duration of the affected segment is 240 seconds; if the starting point of the measurement mutation event is 10:06:00, then =60 seconds, unpacking mutation tolerance is If the event falls near the end of the unpacking process and forms a valid acceptance record, then the event is considered valid. If the starting point of the metering mutation event is 10:10:00, then this starting point exceeds the end point of the affected section, generating an unpacking mutation acceptance degree that does not meet the acceptance conditions. The above results are transmitted to the linkage identification process and participate in the linkage threshold determination together with the metering loop mismatch degree.

[0030] The present invention is further configured such that, within a metering abrupt event, the calculation of the persistent difference between the cumulative active power and the incremental energy, and the persistent difference between the change in in-phase current and the change in active power, to generate the metering loop mismatch degree includes: The process involves extracting the mutation start point, mutation end point, mutation phase, and mutation electrical parameter type from the metering mutation event. Then, the corresponding mutation event interval is extracted from the open-box metering synchronization sequence to generate a mutation metering subsequence. Specifically, after identifying the metering mutation event, the mutation event is used as input to extract the mutation start point, mutation end point, mutation phase, and mutation electrical parameter type from the event record. The mutation phase is used to limit the extraction range of subsequent in-phase currents, and the mutation electrical parameter type is used to determine one or more metering fields involved in the event, such as current, active power, or energy increment. Based on the mutation start point and mutation end point, metering monitoring data within the corresponding time range is extracted from the open-box metering synchronization sequence. The active power, energy readings, in-phase current corresponding to the mutation phase, sampling timestamps, and data gap markers at adjacent sampling times are retained to generate the mutation metering subsequence. The mutation metering subsequence is the basic data block for calculating the metering loop mismatch. Its time boundary originates from the metering mutation event, and its data fields originate from the energy meter acquisition interface and the open-box metering synchronization sequence, ensuring that subsequent difference calculations are limited to the identified mutation events. The active power cumulativeity is generated based on the active power at adjacent sampling times within the mutation metering subsequence, and the energy increment is generated based on the energy readings at adjacent sampling times. The active power cumulativeity and energy increment within the same sampling interval are differentiated to generate the first continuous difference sequence. Specifically, within the mutation metering subsequence, adjacent sampling times are read sequentially according to a unified time axis. and active power , and power consumption readings , The cumulative active power within the same sampling interval is formed based on the active power at adjacent sampling times. Specifically, this can be achieved using... ,in The energy value is converted to kilowatt-hours according to the electricity meter's unit of measurement; the energy increment is formed based on the energy readings at adjacent sampling times. Within the same sampling interval, and Perform differentiation processing to obtain the first difference quantity. ,in The protection quantity, taken from the minimum resolution of the energy meter reading, is used for denominator protection under low load or small energy increment conditions. The first difference quantities are arranged in chronological order of sampling time to form the first persistent difference sequence. This sequence reflects the degree of persistent deviation between the cumulative active power result and the actual change in the energy meter reading within a sudden event, providing a quantitative basis for energy accumulation link anomalies or power-energy inconsistencies. The in-phase current and active power corresponding to the abrupt phase are extracted from the abrupt metering subsequence. The changes in in-phase current and active power at adjacent sampling times are correlated to generate a second persistent difference sequence. Specifically, the in-phase current corresponding to the abrupt phase is read from the abrupt metering subsequence. , and active power at adjacent sampling times , ,in This is to distinguish the abrupt changes in the metering abrupt event record. Current changes are generated based on the in-phase current at adjacent sampling times, and power changes are generated based on the active power at adjacent sampling times. The in-phase current and active power references from the pre-unpacking metering standards are then used for scaling. Specifically, the current change can be normalized to... Normalize the change in active power to ,in For the current reference corresponding to the abrupt phase change, As the active power reference, and These are taken from the minimum measurable current and the minimum measurable power, respectively. and The difference is generated as the second difference quantity The sampling sequence is used to form a second persistent difference sequence. This sequence reflects the corresponding deviation between the current change and the active power change under abrupt phase changes, and is used to identify metering loop mismatch situations where the current loop change is not synchronously reflected by the active power, or the active power change is not consistent with the in-phase current change. The continuity of the first and second persistent difference sequences is verified, eliminating isolated jump points and missing transmission points, and retaining persistent difference segments. Specifically, continuity verification is performed on the first and second persistent difference sequences respectively. Continuity verification includes three processes: First, reading the data gap markers and transmission markers in the abrupt measurement subsequence and removing the corresponding sampling intervals from the difference sequence; second, identifying isolated jump points that appear only in a single sampling interval and do not form the same type of difference in the preceding and following sampling intervals, and removing them as single-point fluctuations; third, retaining the difference amounts that continuously reach a preset number of continuity points as persistent difference segments. The preset number of continuity points is determined based on the measurement sampling period. For example, when the sampling period is 60 seconds, two consecutive sampling intervals can be used as the continuity requirement; when the sampling period is shorter, the number of continuity points can be increased. After this step, the first persistent difference sequence retains segments where power-energy is continuously inconsistent, and the second persistent difference sequence retains segments where in-phase current-active power is continuously inconsistent. This process separates missing transmissions, isolated spikes, and persistent mismatches, giving the difference segments entering the mismatch aggregation a continuous basis. The first and second persistent difference sequences are normalized and aggregated separately to generate the metrological loop mismatch. Specifically, the first and second persistent difference segments are normalized and aggregated separately to generate the metrological loop mismatch. The normalization aggregation can use the average level of the difference within the persistent segment to form the first mismatch component. and the second mismatch component Then, the metering loop mismatch degree is formed by assigning equal weights. When a certain type of persistent difference segment is absent, the mismatch component of that type is recorded as zero; when both types of persistent difference segments exist, the metering loop mismatch degree includes both power-energy mismatch and in-phase current-power mismatch. The generated metering loop mismatch degree is written into the metering abrupt event record and transmitted to the linkage identification process, participating in threshold comparison together with the open-box abrupt event acceptance degree. As a verification example, the metering abrupt event interval contains 3 adjacent sampling intervals, with a sampling period of 60 seconds, a reference active power of 3.0kW, and a reference A-phase current of 5.0A; if the cumulative active power of the first sampling interval is 0.050kWh and the energy increment is 0.020kWh, then the first difference is 0.60; if the normalized change of the A-phase current within the same interval is -0.40 and the normalized change of the active power is -0.03, then the second difference is 0.37. If similar differences continue to appear in subsequent sampling intervals, they are retained as persistent difference segments. When the metering loop mismatch obtained after aggregation exceeds the mismatch threshold, it indicates that the metering mutation event has loop mismatch characteristics and enters the linkage identification result output process.

[0031] The present invention is further configured such that the step of generating exclusion results for data missing transmission and three-phase synchronous load changes includes: The sampling timestamp, receiving timestamp, data gap marker, and energy value within the interval corresponding to the metering mutation event are read from the metering synchronization sequence. A continuous energy value record is formed based on the changes in energy values ​​at adjacent sampling times. Data segments with continuous gaps, out-of-order writing, concentrated shifts in receiving time, and cross-period energy value supplementation are identified to generate data missing transmission markers. Specifically, after identifying the metering mutation event, the interval corresponding to the metering mutation event is used as the exclusion criterion. The sampling timestamp, receiving timestamp, data gap marker, and energy value for each sampling time within that interval are read from the metering synchronization sequence. The sampling points are arranged according to their sampling timestamps, and a continuous energy value record is formed based on the changes in energy values ​​at adjacent sampling times. This continuous energy value record includes the time sequence of adjacent sampling points, the increase / decrease status of energy values, and the status of energy value changes across sampling intervals. If, within the interval corresponding to a metering mutation event, there are consecutive missing sampling points, inverted sampling timestamps, data from multiple different sampling times concentrated in the same receiving period, or the energy readings after the missing period are all filled with the cumulative change corresponding to the previous missing period, then the corresponding data segment is marked as a data missing data retransmission segment. The data missing data retransmission marker originates from the data missing markers in the unpacked metering synchronization sequence, the sampling timestamp order, the receiving timestamp distribution, and the energy reading continuity record, and is used to indicate the existence of data quality anomalies within the metering mutation event interval. After this marker is generated, subsequent three-phase synchronous load change exclusion will no longer be performed to avoid synchronous changes caused by centralized writing of retransmission data being misjudged as actual load changes. When no data loss supplementation marker is generated, the three-phase current and three-phase active power are extracted from the interval corresponding to the metering abrupt event, and current change sequences and active power change sequences are formed according to each phase. Specifically, under the condition that no data loss supplementation marker is generated for the interval corresponding to the metering abrupt event, the currents and active power of phases A, B, and C in that interval are read from the unpacked metering synchronization sequence. Current change sequences and active power change sequences are formed according to each phase, where the current change sequence consists of the change in current of the same phase at adjacent sampling times, and the active power change sequence consists of the change in active power of the same phase at adjacent sampling times. For each phase... The change in current between adjacent sampling times can be denoted as... The change in active power is denoted as ,in Choose phase A, phase B, or phase C. All the above change sequences are generated within the same abrupt event interval and correspond one-to-one with a unified sampling time axis, ensuring that the synchronous changes in three-phase current and three-phase active power have the same time reference. The three-phase current change sequence and the three-phase active power change sequence are evaluated for directional consistency. Metering abrupt events where the three-phase current and three-phase active power rise and fall synchronously without any single-phase local deviation are marked as synchronous three-phase load changes. Specifically, the directional consistency of the three-phase current change sequence and the three-phase active power change sequence is evaluated. For the same sampling interval, first, it is determined whether the current change directions of phases A, B, and C are consistent, and then it is determined whether the active power change directions of phases A, B, and C are consistent. When both the three-phase current and the three-phase active power continuously increase or continuously decrease in the same direction, it is further determined whether there is a single-phase local deviation. Single-phase local deviation refers to a change in amplitude in one phase that is significantly higher or lower than that in the other two phases. It is typically determined by whether the difference in the three-phase change amplitudes exceeds the inter-phase allowable deviation. When determining the directional consistency of three-phase current change sequences and three-phase active power change sequences, the current change in each phase is first normalized according to the current reference of the corresponding phase, and the active power change in each phase is normalized according to the active power reference of the corresponding phase, resulting in the normalized three-phase change amplitude. For the same sampling interval, the differences between the normalized change amplitudes of phases A, B, and C are compared, and the largest inter-phase difference is taken as the inter-phase deviation for that sampling interval. The inter-phase allowable deviation is determined based on the normal operating state of the three phases within the reference segment. Specifically, the inter-phase deviation of each effective sampling interval is calculated in the same way within the reference segment. The inter-phase deviations within the reference segment are sorted by value, and the effective inter-phase deviation closest to the upper limit is taken as the upper limit of the reference inter-phase fluctuation. This, combined with the current sampling accuracy and active power sampling accuracy, forms the inter-phase allowable deviation. If the phase-to-phase deviation within the abrupt change event interval does not exceed the phase-to-phase allowable deviation, then no single-phase local deviation is considered to have occurred; if the phase-to-phase deviation within the abrupt change event interval exceeds the phase-to-phase allowable deviation, then a single-phase local deviation is considered to have occurred. When the difference between the normalized change amplitudes of the three phases does not exceed the phase-to-phase allowable deviation, and the directional consistency state continues to reach the preset number of continuous sampling points, the metering abrupt change event is marked as a three-phase synchronous load change. The preset number of continuous sampling points for the directional consistency state is determined based on the minimum continuous observation time and the metering sampling period. Specifically, the minimum continuous observation time is divided by the metering sampling period and rounded up to obtain the preset number of continuous sampling points, and the preset number of continuous sampling points is not less than two sampling points. If the sampling period is 60 seconds and the minimum continuous observation time is 120 seconds, then the preset number of continuous sampling points is 2. Only when the synchronous rise and fall states of the three-phase current and the three-phase active power continuously reach the preset number of continuous sampling points, and no single-phase local deviation occurs, is the metering abrupt change event marked as a three-phase synchronous load change. This process distinguishes normal three-phase load fluctuations from single-phase metering circuit anomalies, preventing simultaneous changes in three-phase load from being included in the unpacking linkage sudden change judgment. The exclusion judgment result is generated based on the data missing transmission marker and the three-phase synchronous load change marker. Specifically, the exclusion judgment result is generated based on the data missing transmission marker and the three-phase synchronous load change marker. If a data missing transmission marker has been generated, the exclusion judgment result is written to the data anomaly pending verification category; if no data missing transmission marker has been generated but a three-phase synchronous load change marker has been generated, the exclusion judgment result is written to the load synchronization exclusion category; if neither type of marker has been generated, the exclusion judgment result is written as no exclusion marker, and is passed to the linkage judgment process of the unpacking sudden change acceptance degree and the metering circuit mismatch degree. As a verification example, the metering sudden change event interval is from 10:06:00 to 10:09:00. If there are missing markers in the metering data at 10:07:00 and 10:08:00, and the data from multiple sampling times is concentratedly written at the receiving time at 10:09:00, and the power consumption reading shows cumulative supplementation across the missing time period at 10:09:00, then a data missing transmission marker is generated, and the exclusion judgment result is data anomaly pending verification. If there are no gaps or supplementary transmissions in the interval, and the currents of phases A, B, and C all decrease continuously, the active power of the three phases also decreases synchronously, and the change amplitudes of each phase are similar, then a three-phase synchronous load change flag is generated, and the exclusion judgment result is load synchronization exclusion. If the sudden change is concentrated in a single phase, and the direction and amplitude of the three-phase change do not meet the synchronization conditions, then no exclusion flag is generated, and the relevant event continues to enter the linkage identification result output process.

[0032] The present invention is further configured such that the output of the linkage identification result based on the unpacking mutation acceptance degree, the metering loop mismatch degree, and the exclusion discrimination result includes: When the exclusion judgment result includes a data missing transmission marker, the abnormal unpacking event and the metering mutation event are associated and marked as data abnormality pending verification. Specifically, in the linkage identification output stage, the abnormal unpacking event record, the metering mutation event record, the unpacking mutation acceptance degree, the metering circuit mismatch degree, and the exclusion judgment result are read. The exclusion judgment result comes from the previous data missing transmission and three-phase synchronous load change exclusion judgment process. The data missing transmission marker is generated by data segments with continuous gaps, out-of-order writing, concentrated shift of receiving time, and cross-period supplementation of energy readings. If the exclusion judgment result includes a data missing transmission marker, the current event is directly marked as data abnormality pending verification, and the corresponding unpacking start point, unpacking end point, metering mutation start point, data missing transmission marker, and result marker are written to the result cache. This processing prioritizes data quality as the discrimination condition, so that the supplemented data or out-of-order data does not enter the final linkage classification of unpacking mutation acceptance degree and metering circuit mismatch degree. When the exclusion judgment result includes a three-phase synchronous load change flag, the abnormal unpacking event and the metering mutation event are associated and marked as load synchronization exclusion. Specifically, if the exclusion judgment result does not include a data missing transmission flag, the three-phase synchronous load change flag is read. The three-phase synchronous load change flag is derived from the directional consistency judgment of the three-phase current change sequence and the three-phase active power change sequence, and is generated only when the current and active power of phases A, B, and C rise and fall synchronously, and no single-phase local deviation occurs. If the exclusion judgment result includes a three-phase synchronous load change flag, the current event is marked as load synchronization exclusion, and the three-phase synchronous load change flag, the involved time interval, and the result flag are written to the result cache. This step separates the overall three-phase load rise and fall from single-phase or local metering circuit anomalies, so that subsequent linkage outputs retain the identification object of local metering circuit anomalies. When an abnormal unpacking event exists but no metering mutation event is generated, the abnormal unpacking event is marked as "unpacking without mutation." Specifically, excluding cases where the judgment result does not include data missing transmission markers and three-phase synchronous load change markers, the abnormal unpacking event record and the metering mutation event record are read. If the abnormal unpacking event record exists and no metering mutation event is generated within the unpacking-affected segment, the abnormal unpacking event is marked as "unpacking without mutation." The unpacking start and end points in the abnormal unpacking event record are written to the result record, while the metering mutation start point, mutation phase, and mutation electrical parameter type are written as null values ​​or "not generated." This step is used to form the complete output branch of the abnormal unpacking event, ensuring that unpacking events that do not cause a continuous deviation in the metering state do not enter subsequent mutation association classifications. When a metering mutation event exists and the acceptance rate of the unpacking mutation does not meet the acceptance threshold, the metering mutation event is marked as a mutation without unpacking. Specifically, under the condition of a metering mutation event, the acceptance rate of the unpacking mutation is read and compared with the acceptance threshold. The acceptance rate of the unpacking mutation is obtained by normalizing the relative position of the starting point of the metering mutation event within the affected section. The acceptance threshold is determined based on the duration of the affected section, the duration of the abnormal unpacking, and the allowable lag time. The duration of the affected section is determined by the time length between the starting point and the ending point of the affected section; the duration of the abnormal unpacking is determined by the time length between the starting point and the ending point of the unpacking; the allowable lag time is determined based on the sampling period of the metering monitoring data, the remaining time deviation after time correction, and the data reporting interval of the energy meter acquisition interface. During processing, the duration of abnormal opening and the allowable lag time are combined as the effective acceptance time. This effective acceptance time is then converted to a unified scale corresponding to the duration of the affected segment to obtain the acceptance threshold. Specifically, the duration of the affected segment is the time length between the start and end points of the affected segment; the effective acceptance time is the sum of the duration of abnormal opening and the allowable lag time. The duration of abnormal opening is the time length between the start and end points of the opening, and the allowable lag time is determined based on the metering sampling cycle, the remaining time deviation after time correction, and the data reporting interval of the energy meter's data acquisition interface. During processing, the effective acceptance time is divided by the duration of the affected segment to obtain the proportion of the effective acceptance time within the affected segment. This proportion is then subtracted from one to obtain the acceptance threshold. Since the open-box mutation acceptance rate is normalized according to the time distance between the start of the metering mutation event and the start of the affected section, the closer the start of the metering mutation event is to the start of the affected section, the greater the open-box mutation acceptance rate. When the start of the metering mutation event is within the effective acceptance time range, the open-box mutation acceptance rate reaches the acceptance threshold; when the start of the metering mutation event exceeds the effective acceptance time range, the open-box mutation acceptance rate is lower than the acceptance threshold. Therefore, the closer the metering mutation event is to the start of the open-box, the easier it is for the open-box mutation acceptance rate to meet the acceptance threshold; the closer the metering mutation event is to the end of the affected section, the less likely it is to meet the acceptance threshold. If a metering mutation event exists, but the open-box mutation acceptance rate does not reach the acceptance threshold, or if the previous processing has generated an open-box mutation acceptance rate that does not meet the acceptance conditions, the metering mutation event is marked as a mutation without an open-box. This result is recorded with the metering mutation start point, mutation phase, mutation electrical parameter type, and open-box mutation acceptance rate, used to indicate insufficient time acceptance between the metering mutation and the abnormal open-box event. When the acceptance degree of the unpacking mutation meets the acceptance threshold but the mismatch degree of the metering circuit does not meet the mismatch threshold, the abnormal unpacking event and the metering mutation event are associated and marked as insufficient mismatch for exclusion. Specifically, under the condition that the acceptance degree of the unpacking mutation reaches the acceptance threshold, the metering circuit mismatch degree is read and compared with the mismatch threshold. The metering circuit mismatch degree is obtained by normalizing and aggregating the continuous difference between the cumulative active power and the incremental energy, and the continuous difference between the change in in-phase current and the change in active power. The mismatch threshold is formed based on the normal metering relationship within the reference segment. Specifically, within the reference segment, following the same data processing method as the metering circuit mismatch degree, the reference difference between the cumulative active power and the incremental energy, and the reference difference between the change in in-phase current and the change in active power are extracted to form a reference mismatch sequence. The reference difference between the cumulative active power and the incremental energy is calculated using the same method as the generation method of the first continuous difference sequence in the metering circuit mismatch degree, i.e. The results obtained from the baseline data are applied; the baseline difference between in-phase current variation and active power variation is calculated using the same method as the second persistent difference sequence generation, i.e. The results obtained from the baseline data are applied. For the same valid sampling interval, the larger of the two baseline differences is taken as the baseline mismatch value for that sampling interval. The baseline mismatch values ​​of each valid sampling interval are arranged in order of sampling time to form a baseline mismatch sequence. The baseline mismatch sequence is used to record the natural fluctuation levels between power and energy, and current and power under normal conditions before unpacking. Subsequently, the concentration level and fluctuation level are extracted from the baseline mismatch sequence, and combined with the energy meter accuracy, current sampling accuracy, active power sampling accuracy, and energy display resolution to form a mismatch threshold. This mismatch threshold serves as the basis for comparing the degree of mismatch in the metering circuit within a metering mutation event. If the degree of acceptance of the unpacking mutation reaches the acceptance threshold, but the degree of mismatch in the metering circuit does not reach the mismatch threshold, the current event is marked as insufficient mismatch and excluded. This step excludes metering fluctuations that accept abnormal unpacking in time but do not form a continuous mismatch in electrical correspondence from the unpacking linkage mutation. When both the unpacking mutation acceptance rate and the metering circuit mismatch rate meet their corresponding thresholds, and the exclusion judgment result does not include data missing transmission markers and three-phase synchronous load change markers, the abnormal unpacking event and the metering mutation event are associated and marked as an unpacking-linked mutation. Specifically, under the condition that the exclusion judgment result does not include data missing transmission markers and three-phase synchronous load change markers, when the unpacking mutation acceptance rate reaches the acceptance threshold and the metering circuit mismatch rate reaches the mismatch threshold, the abnormal unpacking event and the metering mutation event are associated, and the associated event is marked as an unpacking-linked mutation. In specific implementation, the linkage judgment state can be recorded as follows: ,when , And excluding cases where the judgment result is empty, let Corresponding to the sudden change upon unpacking; among which To ensure the tolerance for sudden changes upon unpacking, To accept the threshold, For metering loop mismatch, This step sets the mismatch threshold. Events that simultaneously meet both the time continuity condition and the metering loop mismatch condition are grouped into the same linked event, generating the classification tags required for subsequent result recording. The opening start point, opening end point, metering mutation start point, metering mutation phase, mutation electrical parameter type, opening mutation acceptance rate, metering circuit mismatch rate, and result flag are written into the same result record to generate a linkage identification result. Specifically, the result record can be stored according to the event number, which is formed by associating the abnormal opening event number and the metering mutation event number. As a verification example, the opening start point of the abnormal opening event is 10:05:00, the opening end point is 10:06:00, the metering mutation start point is 10:06:00, the mutation phase is phase A, and the mutation electrical parameter type is current and active power; the opening mutation acceptance rate is 0.75, the acceptance threshold is 0.50, the metering circuit mismatch rate is 0.62, and the mismatch threshold is 0.30. Excluding the judgment result that does not include the data missing transmission mark and the three-phase synchronous load change mark, the generated result is marked as "opening linkage mutation". If the same data shows a concentrated shift in reception time and a power consumption reading that is updated across time periods, then a data anomaly pending verification will be generated first; if the three-phase current and three-phase active power rise and fall synchronously without any single-phase local deviation, then a "load synchronization exclusion" will be generated.

[0033] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for linking abnormal opening of an electricity metering box with sudden metering changes, characterized in that, include: S1: Acquire opening monitoring data, authorization records and metering monitoring data through the door sensor, electronic lock interface and energy meter acquisition interface; correct the sampling timestamp and receiving timestamp of the opening monitoring data, authorization records and metering monitoring data to generate an opening metering synchronization sequence. S2: Extract continuous state segments of door opening and closing from the box opening metering synchronization sequence, identify state segments that do not match the authorized record and whose duration meets the valid box opening conditions as abnormal box opening events, and record the box opening start and end points; S3: The measurement and monitoring data before the start of unpacking is used as the baseline segment, and the measurement and monitoring data during the lag time from the start of unpacking to the end of unpacking is used as the influence segment. Within the benchmark section, sampling points whose cumulative active power and incremental electrical energy are consistent are selected to establish a metering benchmark; S4: Perform normalized deviation calculation on the metrological monitoring data within the affected section relative to the metrological benchmark, identify sampling sections with continuous deviations in the same direction as metrological mutation events, and calculate the unpacking mutation acceptance degree based on the relative position of the starting point of the metrological mutation event within the affected section. S5: Calculate the continuous difference between the cumulative active power and the incremental power, and the continuous difference between the change in in-phase current and the change in active power within the metering abrupt event to generate the metering loop mismatch. S6: Perform exclusion judgment on missing data transmission and three-phase synchronous load changes, generate exclusion judgment results, and output linkage identification results based on the unpacking sudden change acceptance degree, metering circuit mismatch degree and exclusion judgment results.

2. The method for linking abnormal opening and metering change identification of an electricity metering box according to claim 1, characterized in that, The sampling timestamps and receiving timestamps of the unpacking monitoring data, authorization records, and metering monitoring data are corrected to generate an unpacking metering synchronization sequence, including: The door sensor acquires the door opening status, door closing status, and door status sampling time; the electronic lock interface acquires the lock action time, unlocking method, and authorization identifier; and the energy meter acquisition interface acquires voltage, current, active power, energy readings, and metering sampling time. Add data source identifier, sampling sequence number, sampling timestamp and receiving timestamp to the unpacking monitoring data, authorization record and metering monitoring data respectively, and group and sort them according to the data source identifier to generate unpacking data group, authorization data group and metering data group; The time correction amount is determined based on the correspondence between the sampling timestamp and the receiving timestamp within the same data group. For data with missing or abnormal sampling timestamps, the corrected data group is generated by supplementing and correcting the data according to the time order of adjacent sampling data in the same group. A unified sampling time axis is established based on the sampling period of the metering and monitoring data. The door status, lock action, authorization mark, voltage, current, active power and energy indication in the corrected data group are written into the corresponding sampling time to generate the opening metering synchronization sequence. Set a data gap marker for data items that are not written to the corresponding sampling time, and write the data gap marker to the open-box metering synchronization sequence.

3. The method for linking abnormal opening and metering change identification of an electricity metering box according to claim 2, characterized in that, Extract continuous state segments of door opening and closing from the unpacking metering synchronization sequence. State segments without matching authorization records but whose duration meets the valid unpacking conditions are identified as abnormal unpacking events. Record the start and end points of the unpacking, including: Extract the door status field, lock action field, and authorization identifier field from the box opening measurement synchronization sequence, and form a door status subsequence, lock action subsequence, and authorization status subsequence according to a unified time axis; The state subsequence of the cabinet door is stabilized by incorporating instantaneous jumps with a duration shorter than the preset anti-shake duration into adjacent stable states to generate a stable state sequence of the cabinet door. In the stable state sequence of the cabinet door, locate the first transition point from closed state to open state and the second transition point from open state to closed state, and determine the opening holding interval between the first transition point and the second transition point as the candidate open state segment; Based on the first transition point of the candidate open box state segment, the corresponding authorization identifier and unlocking action record are retrieved in the authorized state subsequence and the lock action subsequence. Based on the second transition point of the candidate open box state segment, the corresponding locking action record is retrieved in the lock action subsequence. The candidate open box state segment that matches the authorization identifier, the unlocking action record is valid, and the locking action record satisfies the closure correspondence condition is marked as the authorized open box state segment. Candidate unboxing state segments that are not marked as authorized unboxing state segments and whose duration of the open holding interval reaches the preset effective unboxing time are identified as abnormal unboxing events. The time corresponding to the first transition point is taken as the unboxing start point, and the time corresponding to the second transition point is taken as the unboxing end point to generate an abnormal unboxing event record.

4. The method for linking abnormal opening and metering change identification of an electricity metering box according to claim 1, characterized in that, The measurement and monitoring data before the start of unpacking is taken as the baseline segment, and the measurement and monitoring data during the lag time from the start of unpacking to the end of unpacking is taken as the influencing segment, including: Locate the metering sampling positions corresponding to the start and end points of the unpacking measurement synchronization sequence; Based on the starting point of unpacking, continuous metering and monitoring data are extracted backwards, and metering sampling points with data gaps, timestamp jumps, and power consumption reading rollbacks are removed to generate a benchmark segment; Metering and monitoring data are extracted based on the opening start point, opening end point, and preset lag time. The metering and monitoring data from the opening start point to the opening end point and the metering and monitoring data within the lag time after the opening end point are merged to generate the influence segment.

5. The method for linking abnormal opening and metering change identification of an electricity metering box according to claim 4, characterized in that, Within the benchmark range, sampling points whose cumulative active power and incremental electricity consumption are consistent are selected to establish a metering benchmark, including: Extract the active power, current and energy readings between adjacent sampling times within the reference segment; calculate the cumulative active power based on the active power at adjacent sampling times; and calculate the energy increment based on the change in energy readings at adjacent sampling times. The cumulative active power and the incremental power within the same sampling interval are compared for consistency, and the sampling points that meet the consistency conditions are marked as valid reference sampling points. Based on the effective reference sampling points, in-phase current, active power and energy increment are extracted to generate current reference, active power reference and energy increment reference respectively; The current reference, active power reference, and energy increment reference are combined into a metering reference.

6. The method for linking abnormal opening of an electricity metering box with metering change identification according to claim 5, characterized in that, The normalized deviation of the metrological monitoring data within the affected section relative to the metrological benchmark is calculated, and sampling sections with continuous deviations in the same direction are identified as metrological abrupt events, including: Read the in-phase current reference, active power reference, and energy increment reference from the metering reference; extract the in-phase current, active power, and energy increment from the affected section according to the sampling time; and generate the metering sampling sequence of the affected section. The in-phase current, active power, and energy increment at each sampling time in the affected segment metering sampling sequence are processed with the corresponding benchmarks, and then normalized according to the scale of the corresponding benchmarks to generate a point-by-point deviation sequence. Based on the excess state and deviation direction of the deviation amount in the point-by-point deviation sequence, adjacent sampling points are continuously merged, and sampling segments that continuously exceed the deviation threshold and have the same deviation direction are identified as candidate mutation segments. Extract the mutation initiation point, mutation endpoint, mutation duration, mutation phase, and mutation electrical parameter type of candidate mutation segments to generate metrological mutation events.

7. The method for linking abnormal opening of an electricity metering box with metering change identification according to claim 6, characterized in that, The calculation of the open-box mutation tolerance based on the relative position of the starting point of the metric mutation event within the affected segment includes: Read the start and end points of the unpacking from the abnormal unpacking event, and read the start and end points of the impact segment from the impact segment. Divide the impact segment into the unpacking process sub-segment and the post-unpacking lag sub-segment. Match the time position of the metering mutation event start point with the start point of the affected segment, the end point of the unpacking process, and the end point of the affected segment to determine the position where the metering mutation event start point falls in the unpacking process sub-segment or the post-unpacking lag sub-segment. When the starting point of a measurement mutation event falls into the affected segment, the open-box mutation acceptance degree is generated by normalizing the time distance between the starting point of the measurement mutation event and the starting point of the affected segment within the duration of the affected segment. If the starting point of a measurement mutation event does not fall within the affected segment, or if the starting point of a measurement mutation event is earlier than the starting point of an abnormal unpacking event, the failure to meet the acceptance condition will be marked and written into the measurement mutation event record.

8. The method for linking abnormal opening of an electricity metering box with metering change identification according to claim 6, characterized in that, Within a metering abrupt event, the metering loop mismatch is generated by calculating the persistent difference between the cumulative active power and the incremental energy, and the persistent difference between the change in in-phase current and the change in active power. Read the mutation start point, mutation end point, mutation phase, and mutation electrical parameter type from the metering mutation event, and extract the corresponding mutation event interval from the unpacked metering synchronization sequence to generate a mutation metering subsequence; The active power cumulative amount is formed based on the active power at adjacent sampling times within the mutation measurement subsequence, and the energy increment is formed based on the energy reading at adjacent sampling times. The active power cumulative amount and energy increment within the same sampling interval are differentiated to generate the first continuous difference sequence. Extract the in-phase current and active power corresponding to the abrupt phase from the abrupt metering subsequence, and process the changes in in-phase current and active power at adjacent sampling times to generate a second continuous difference sequence. The continuity of the first and second persistent difference sequences is checked, isolated jump points and missing transmission points are removed, and persistent difference segments are retained. The first and second persistent difference sequences that persist are normalized and aggregated to generate the metrological loop mismatch.

9. The method for linking abnormal opening and metering change identification of an electricity metering box according to claim 2, characterized in that, The process of identifying and eliminating missing data and changes in three-phase synchronous load generates the following exclusion results: The sampling timestamp, receiving timestamp, data gap marker, and energy value within the interval corresponding to the metering mutation event are read from the metering synchronization sequence. A continuous record of energy value is formed based on the changes in energy value at adjacent sampling times. Data segments with continuous gaps, out-of-order writing, concentrated shift of receiving time, and cross-period energy value supplementation are identified to generate data missing transmission markers. When no data missing data supplementation marker is generated, the three-phase current and three-phase active power are extracted from the interval corresponding to the metering mutation event, and the current change sequence and active power change sequence are formed according to the phase respectively. The three-phase current change sequence and the three-phase active power change sequence are judged for directional consistency. Metering abrupt events in which the three-phase current and three-phase active power rise and fall synchronously without any single-phase local deviation are marked as three-phase synchronous load changes. The exclusion criteria are generated based on the missing data transmission markers and the three-phase synchronous load change markers.

10. The method for linking abnormal opening of an electricity metering box with metering change identification according to claim 9, characterized in that, Based on the unpacking mutation acceptance rate, metering loop mismatch rate, and exclusion judgment results, the linked identification results are output as follows: When excluding the data missing transmission marker in the judgment result, the abnormal unpacking event and the metrological mutation event are associated and marked as data abnormality pending verification; When the exclusion judgment result includes the three-phase synchronous load change mark, the abnormal unpacking event and the metering change event are associated and marked as load synchronous exclusion; When an abnormal unpacking event exists but no metrological mutation event is generated, the abnormal unpacking event is marked as unpacking without mutation; When a metering mutation event exists and the acceptance rate of the unpacking mutation does not meet the acceptance threshold, the metering mutation event will be marked as a mutation non-unpacking. When the acceptance of unpacking mutation meets the acceptance threshold but the mismatch of metering circuit does not meet the mismatch threshold, the abnormal unpacking event and the metering mutation event are associated and marked as insufficient mismatch for exclusion. When both the unpacking mutation acceptance rate and the metering circuit mismatch rate meet the corresponding thresholds, and the exclusion judgment results do not include data missing transmission markers and three-phase synchronous load change markers, the abnormal unpacking event and the metering mutation event are associated and marked as unpacking linkage mutation. The start point of unpacking, the end point of unpacking, the start point of metering mutation, the phase of metering mutation, the type of mutated electrical parameter, the acceptance degree of unpacking mutation, the degree of metering circuit mismatch, and the result mark are written into the same result record to generate a linkage identification result.

Images