A smart meter multi-device cooperative optimization method and device for distributed energy access, equipment and medium

Abstract

The application provides a smart meter multi-device cooperative optimization method and device for distributed energy access, equipment and medium, wherein, the interpolation unified meter, inverter and energy storage data granularity build a joint time error model to calibrate the reference data to complete the multi-device time synchronization calibration; based on the reference data, the three-device operation data is used to distinguish the meter hardware failure and the multi-device cooperative anomaly by using the joint energy consumption model to complete the fault diagnosis; combined with the inverter to the meter line and the power data, the local line loss is calculated, the residual power measurement deviation is verified, the settlement data is corrected, the line loss-bidirectional metering linkage verification is completed; the fault result is associated with the metering data, the fault type and the deviation rate are long-term statistical, the line loss corresponding relationship is modeled to predict the deviation trend and early warning; the multi-device cooperative anomaly source and the relationship between the disturbance and the load interference are determined to calculate the interference coefficient to feed back the dispatching system to mark the line reconstruction priority. The above method is used to improve the distributed energy consumption efficiency and the power distribution network operation reliability.

Description

Technical Field

[0001] This invention relates to the field of smart meters, and more specifically, to a method, apparatus, device, and medium for multi-device collaborative optimization of smart meters for distributed energy access. Background Technology

[0002] With the widespread integration of distributed energy into power distribution networks, smart meters, inverters, energy storage and other equipment together constitute a key hardware cluster for the consumption and management of distributed energy. Existing technical solutions in the industry are all built around this cluster to construct a basic management system.

[0003] Current mainstream technologies generally follow the core logic of separate device function implementation and shallow data interaction: independent functional modules are configured for smart meters, inverters and energy storage respectively. Each module only focuses on the collection of operating parameters, basic status monitoring and single-point problem handling of the corresponding device. The modules only share a small amount of basic data through a common method, and do not form a multi-device linkage control logic.

[0004] However, the aforementioned existing technologies share a significant common deficiency: their management systems are centered on the functional implementation of individual devices, lacking the unified logical support and mechanism design required for the collaborative operation of multiple devices. This limitation prevents the effective mining of the correlation between data from multiple devices, making it difficult not only to accurately distinguish between device-specific problems and abnormal interactions between multiple devices, but also to ensure that load management in distribution areas cannot be carried out in a coordinated manner based on the collaborative status of multiple devices. Ultimately, existing technologies are ill-suited to the complex operational needs of distribution networks after the integration of distributed energy resources, thus hindering the improvement of distributed energy consumption efficiency and equipment management reliability. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a method, device, equipment and medium for multi-device collaborative optimization of smart meters for distributed energy access, so as to improve the efficiency of distributed energy consumption and the reliability of distribution network operation, and protect the interests of multiple stakeholders.

[0006] In a first aspect, embodiments of this application provide a multi-device collaborative optimization method for smart meters oriented towards distributed energy access, the method comprising:

[0007] By interpolating to unify the data granularity of meters, inverters and energy storage and constructing a joint time error model, calibrated unified time reference data is obtained, and time synchronization calibration of multiple devices is completed.

[0008] Based on the unified time reference data, the operation data of electricity meters, inverters and energy storage are collected. The joint energy consumption model is used to distinguish between the hardware failure of the electricity meter itself and the abnormality of multi-device coordination, and to complete the multi-dimensional collaborative fault diagnosis.

[0009] Based on the unified time reference data, combined with the line parameters and power data from the inverter output to the meter input, local line loss is calculated and residual power metering deviation is verified. The cause of the deviation is traced and the settlement data is corrected to complete the line loss-bidirectional metering linkage verification.

[0010] By linking the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-bidirectional metering linkage verification, a quantitative model is constructed by long-term statistical analysis of the correspondence between different fault types and metering deviation rate and line loss changes. Based on the model, the metering deviation trend is predicted and early warnings are sent to achieve dynamic improvement of metering accuracy.

[0011] And / or, associate the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-two-way metering linkage verification, calculate the interference coefficient by determining the interference correlation between multi-device collaborative abnormal sources and transformer area loads, feed the interference coefficient back to the power grid dispatching system and mark the priority of line transformation to achieve power grid load-side management assistance;

[0012] Among them, multi-dimensional collaborative fault diagnosis and line loss-bidirectional metering linkage verification achieve collaboration through data feedback, and multi-device time synchronization calibration provides a data foundation for the former two and subsequent related processing.

[0013] Optionally, the step of interpolating to unify the data granularity of meters, inverters, and energy storage, and constructing a joint time error model to obtain calibrated unified time reference data, thereby completing multi-device time synchronization calibration, includes:

[0014] The cubic spline interpolation method is used to unify the data from different sampling periods of the electricity meter, inverter and energy storage into interpolated data with a set granularity by setting the interpolation ratio;

[0015] Using the time of the main meter in the distribution area as the reference time, a joint time error model is constructed, which includes meter time error, inverter time error, and energy storage time error. In this model, the interpolated electricity of the main meter in the distribution area under the reference time is equal to the sum of the meter interpolated electricity, the inverter interpolated power correction electricity, the energy storage interpolated current correction electricity, and the network loss.

[0016] The joint time error model is solved by generalized least squares method to obtain the time error of each device. When the time error of a device is greater than the set value, a time calibration command is sent to it to ensure that the time error of the device after calibration is not greater than the set accuracy, and finally a unified time reference data is obtained.

[0017] Optionally, based on the unified time reference data, the operation data of the electricity meter, inverter, and energy storage are collected, and a joint energy consumption model is used to distinguish between hardware faults of the electricity meter itself and abnormalities in multi-device coordination, thereby completing multi-dimensional collaborative fault diagnosis, including:

[0018] The energy consumption values ​​of the electricity meter are collected under different module working states. The first energy consumption value is when only the display module is working, the second energy consumption value is when the display module and the communication module work together, and the third energy consumption value is when the display module, the communication module and the data processing module work together.

[0019] Based on the first energy consumption value, the second energy consumption value, and the third energy consumption value, the data processing capability coefficient of the electrical form equipment is calculated.

[0020] The output active power and power fluctuation value of the inverter are collected to calculate the power fluctuation coefficient, and the charging and discharging current and SOC value of the energy storage are collected to calculate the SOC influence factor.

[0021] The power fluctuation coefficient and the SOC impact factor are introduced as cooperative interference factors into the data processing capability coefficient of electrical equipment to construct a joint capability coefficient.

[0022] The deviation capability index is calculated based on the joint capability coefficient and the preset theoretical model. Combined with the numerical range of the cooperative interference factor, the fault type is determined.

[0023] Optionally, the fault type is determined by the numerical range of the combined cooperative interference factor, including:

[0024] When the deviation capability index is greater than the set threshold and the power fluctuation coefficient is not greater than the set value and the SOC influence factor is not greater than the set value, it is determined to be a hardware fault of the meter itself. The specific faulty module is located by calculating the edge density of each module.

[0025] When the deviation capability index is greater than the set threshold and the power fluctuation coefficient is greater than the set value or the SOC influence factor is greater than the set value, it is determined to be an abnormality in multi-device coordination, and a power stabilization control command is sent to the inverter or a charge / discharge level slow control command is sent to the energy storage.

[0026] Optionally, the step of calculating local line loss and verifying residual power metering deviation based on the unified time reference data, combined with line parameters and power data from the inverter output to the meter input, tracing the cause of the deviation and correcting the settlement data, and completing the line loss-bidirectional metering linkage verification includes:

[0027] The line length and conductor resistivity from the inverter output terminal to the meter input terminal are obtained, the inverter output current is collected, and the local line loss is calculated based on the line length, conductor resistivity, output current and set metering cycle.

[0028] Collect the total output power of the inverter, the user's own power consumption, and the reverse metering power. Subtract the user's own power consumption and local line loss from the total output power of the inverter to derive the theoretical surplus power.

[0029] Calculate the deviation rate between the reverse metered electricity and the theoretical residual electricity. When the deviation rate is greater than the set value, analyze the trend of the deviation curve through the local outlier factor algorithm to trace the cause of the deviation.

[0030] The steps for tracing the cause of the deviation using the local outlier factor algorithm include:

[0031] When the local outlier curve shows a gradual upward or downward trend, the cause of the deviation is determined to be the aging of the meter's metering chip.

[0032] When the local outlier curve shows a sudden jump trend, the cause of the deviation is determined to be a local circuit anomaly.

[0033] Optionally, the fault results output by the multi-dimensional collaborative fault diagnosis and the metering data output by the line loss-bidirectional metering linkage verification are used to construct a quantitative model by long-term statistical analysis of the correspondence between different fault types and metering deviation rates and line loss changes. Based on the model, the metering deviation trend is predicted and early warnings are sent to achieve dynamic improvement in metering accuracy, including:

[0034] Long-term correlation with the historical fault types and durations of multi-dimensional collaborative fault diagnosis outputs and the historical metering deviation rate and line loss change data of the line loss-bidirectional metering linkage verification outputs;

[0035] By statistically analyzing the correspondence between the above historical data, a quantitative model of the impact of faults on measurement accuracy is constructed. This model reflects the correlation between the duration of different fault types and the increase in measurement deviation rate and the increase in line loss.

[0036] When the multi-dimensional collaborative fault diagnosis detects that the edge density of the target module of the meter is close to the set threshold, or when the local line resistance is slowly increasing as calculated by the local line loss value through the line loss-bidirectional metering linkage verification, the quantization model is invoked.

[0037] Based on the quantitative model, the trend of metering deviation rate changes within a set period is predicted. If the predicted deviation rate exceeds the set threshold, a metering accuracy warning message is sent to the equipment maintenance terminal, and at the same time, an energy consumption data fluctuation prompt message is pushed to the user terminal, so as to realize early intervention of metering deviation.

[0038] Optionally, the fault results output by the multi-dimensional collaborative fault diagnosis and the metering data output by the line loss-bidirectional metering linkage verification are used to calculate the interference coefficient by determining the interference correlation between the multi-device collaborative anomaly source and the transformer area load. The interference coefficient is then fed back to the power grid dispatching system and the priority of line modification is marked to achieve power grid load-side management assistance, including:

[0039] The multi-device collaborative anomaly source information output by multi-dimensional collaborative fault diagnosis is correlated and matched with the local line loss value and user electricity load data output by line loss-bidirectional metering linkage verification.

[0040] Based on the matching results, the correlation between the coordinated abnormal sources and the changes in local line loss in the distribution area and the fluctuations in user electricity voltage is determined. The coordinated abnormal sources include inverter power fluctuations and high-power charging of energy storage. The interference coefficient of the coordinated abnormal sources on the distribution area load is calculated. This coefficient reflects the corresponding ratio between the degree of abnormality of the coordinated abnormal sources and the degree of interference of the distribution area load.

[0041] The interference coefficient is fed back to the power grid dispatching system of the distribution area. Based on the interference coefficient, the dispatching system adjusts the inverter power output or energy storage charging and discharging parameters during the abnormal peak period of the coordinated abnormal source, and at the same time balances the load distribution of adjacent distribution areas to avoid transformer overload or voltage abnormality in the distribution area.

[0042] Long-term monitoring of line loss - local line loss data output by bidirectional metering linkage verification. Inverter-to-meter lines with local line loss values ​​that exceed the set range for a long time are marked as priority targets for line renovation and fed back to the power grid planning terminal.

[0043] Secondly, embodiments of this application provide a multi-device collaborative optimization device for smart meters oriented towards distributed energy access, the device comprising:

[0044] The time synchronization calibration module is used to interpolate the data granularity of meters, inverters and energy storage and build a joint time error model to obtain calibrated unified time reference data and complete the time synchronization calibration of multiple devices.

[0045] The fault diagnosis module is used to collect the operating data of the meter, inverter and energy storage based on the unified time base data, and use the joint energy consumption model to distinguish between the meter's own hardware faults and the abnormality of multi-device coordination, so as to complete the multi-dimensional collaborative fault diagnosis.

[0046] The linkage verification module is used to calculate local line loss and verify residual power metering deviation based on the unified time reference data, combined with the line parameters and power data from the inverter output to the meter input, trace the cause of the deviation and correct the settlement data, and complete the linkage verification of line loss-bidirectional metering.

[0047] The deviation trend prediction module is used to correlate the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-bidirectional metering linkage verification. It constructs a quantitative model by statistically analyzing the correspondence between different fault types and metering deviation rate and line loss changes over a long period of time. Based on the model, it predicts the metering deviation trend and sends early warnings to achieve dynamic improvement in metering accuracy.

[0048] The management assistance module is used to associate the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-bidirectional metering linkage verification. It calculates the interference coefficient by determining the interference correlation between multi-device collaborative anomaly sources and transformer area loads, feeds the interference coefficient back to the power grid dispatching system, and marks the priority of line transformation to achieve power grid load-side management assistance.

[0049] Among them, the fault diagnosis module and the linkage verification module achieve collaboration through data feedback, and the time synchronization calibration module provides the data foundation for the former two and subsequent related processing.

[0050] Thirdly, embodiments of this application provide a computer device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the steps of the multi-device collaborative optimization method for smart meters oriented towards distributed energy access described in any of the optional embodiments of the first aspect above are executed.

[0051] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the multi-device collaborative optimization method for smart meters oriented towards distributed energy access as described in any of the optional embodiments of the first aspect.

[0052] The technical solution provided in this application includes, but is not limited to, the following beneficial effects:

[0053] By interpolating to unify the data granularity of meters, inverters, and energy storage, the problem of inconsistent data acquisition cycles caused by differences in manufacturer design and functional positioning of different devices is solved, achieving uniformity in data format and acquisition frequency, laying the foundation for subsequent cross-device data association. A joint time error model is constructed and a calibrated unified time reference data is obtained, breaking through the accuracy limitations of traditional single-device independent time synchronization. By associating the time deviations of multiple devices to form an overall calibration logic, data timing misalignment caused by time drift of a single device is avoided. This provides high-quality data support with consistent time dimension and unified granularity dimension for subsequent fault diagnosis, metering verification, and association processing, ensuring the accuracy of subsequent technical links.

[0054] By collecting operating data from multiple devices based on a unified time reference, the system ensures that the operating parameters of meters, inverters, and energy storage are correlated in the same time coordinate system, avoiding misjudgments of faults due to time deviations. The system uses a joint energy consumption model to distinguish between hardware faults of the meter itself and abnormalities in the coordination of multiple devices, breaking through the limitations of traditional single-device fault diagnosis. It incorporates external coordination factors such as inverter power fluctuations and energy storage charging and discharging status into the fault judgment logic, accurately identifying hardware damage of the meter itself and abnormal interactions of multiple devices, avoiding misjudgments or omissions, and improving the accuracy and pertinence of fault diagnosis.

[0055] Line loss calculation and metering deviation verification are conducted based on unified time reference data to ensure complete alignment of time periods for data such as inverter output power, user self-consumption power, and reverse metering power, avoiding line loss calculation errors caused by inconsistent time dimensions. Local line losses are calculated by combining line parameters and power data from the inverter output to the meter input, accurately locating energy consumption losses in the "inverter-meter" link, resolving the issue of unclear local link losses in traditional overall distribution area line loss calculations. By verifying surplus power metering deviations, tracing the causes of deviations, and correcting settlement data, disputes over surplus power settlement in distributed energy scenarios are resolved, clarifying the root causes of deviations and correcting data, ensuring settlement fairness between power generation users and the grid company, and reducing metering disputes.

[0056] By linking fault results from multi-dimensional collaborative fault diagnosis with metering data from line loss-bidirectional metering linkage verification, a correlation logic of "fault type-metering deviation-line loss change" is established. This breaks down the barriers between traditional independent fault handling and metering management, providing a clear basis for fault tracing to optimize metering accuracy. Through long-term statistical analysis, a quantitative model is built, transforming fault handling experience into quantifiable predictive logic, realizing a shift from passively handling metering deviations to actively predicting deviation trends. Based on the model, metering deviation trends are predicted and early warnings are sent, allowing for early intervention before metering deviations exceed allowable ranges. This avoids settlement losses or data distortion caused by accumulated deviations, maintains high metering accuracy, and meets the requirements of distributed energy scenarios for electricity metering.

[0057] By linking fault diagnosis results with metering verification data, the system accurately pinpoints the source of multi-device collaborative anomalies affecting the load of distribution areas, solving the problem in traditional load management where load fluctuations are known but the root cause is unknown. By determining the interference correlation between the anomaly source and the load of the distribution area and calculating the interference coefficient, the abstract impact of anomalies is transformed into quantifiable numerical indicators, providing clear data support for power grid dispatch and reducing the subjectivity of dispatch decisions. The interference coefficient is fed back to the dispatch system and the priority of line upgrades is marked, enabling the dispatch system to adjust equipment parameters in a targeted manner to smooth load fluctuations. At the same time, line upgrades are focused on high-priority links, avoiding resource waste, improving the adaptability of the power grid to distributed energy access, and optimizing the efficiency and reliability of load-side management.

[0058] The above steps are logically linked through "data foundation (time synchronization calibration) - intermediate processing (fault diagnosis, metering verification) - value extension (accuracy improvement, load management)" to form a multi-device collaborative optimization system: eliminating data dimension differences with a unified time benchmark, breaking through the limitations of single-device management with a collaborative model, and upgrading from passive processing to proactive prediction with data association. This solves problems such as insufficient multi-device collaboration, inaccurate fault diagnosis, insufficient metering accuracy, and inefficient load management in distributed energy access scenarios, and achieves collaborative control of smart meters, inverters, and energy storage, improving the efficiency of distributed energy consumption and the reliability of distribution network operation, and protecting the interests of multiple stakeholders.

[0059] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0060] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0061] Figure 1 The flowchart of a multi-device collaborative optimization method for smart meters for distributed energy access provided in Embodiment 1 of the present invention is shown.

[0062] Figure 2 A flowchart of a time synchronization calibration method provided in Embodiment 1 of the present invention is shown;

[0063] Figure 3 A flowchart of a fault diagnosis method provided in Embodiment 1 of the present invention is shown;

[0064] Figure 4 A flowchart of a linkage verification method provided in Embodiment 1 of the present invention is shown;

[0065] Figure 5 A flowchart of a deviation trend prediction method provided in Embodiment 1 of the present invention is shown;

[0066] Figure 6 A flowchart of a management assistance method provided in Embodiment 1 of the present invention is shown;

[0067] Figure 7 This shows a schematic diagram of the structure of a multi-device collaborative optimization device for smart meters oriented towards distributed energy access, provided in Embodiment 2 of the present invention;

[0068] Figure 8 A schematic diagram of the structure of a computer device provided in Embodiment 3 of the present invention is shown. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0070] Example 1

[0071] To facilitate understanding of this application, the following is combined with... Figure 1 The flowchart illustrating a multi-device collaborative optimization method for smart meters for distributed energy access provided in Embodiment 1 of the present invention will be described in detail for Embodiment 1 of this application.

[0072] See Figure 1 The above, Figure 1 The flowchart illustrates a multi-device collaborative optimization method for smart meters for distributed energy access provided in Embodiment 1 of the present invention, the method comprising steps S101 to S104:

[0073] S101: By interpolating to unify the data granularity of meters, inverters and energy storage and constructing a joint time error model, calibrated unified time reference data is obtained, and time synchronization calibration of multiple devices is completed.

[0074] Specifically, the electricity meter includes a single-phase smart IoT energy meter. The data acquisition process is achieved through a dedicated transformer acquisition terminal type III (4G wireless public network): the dedicated transformer acquisition terminal type III (4G wireless public network) establishes a communication connection with the single-phase smart IoT energy meter, inverter and energy storage through the 4G wireless public network, and receives the raw data of the three types of devices in real time (the sampling period of the single-phase smart IoT energy meter is 15 minutes, the inverter is 10 minutes, and the energy storage is 5 minutes).

[0075] The interpolation adopts the cubic spline interpolation method, and the interpolation ratio is determined based on the shortest sampling period (5 minutes for energy storage). The data of single-phase smart IoT energy meters and inverters are unified into interpolated data with a granularity of 5 minutes. The joint time error model takes the total time of the distribution area meter as the sole reference. The model includes three core variables: time error Δt1 of single-phase smart IoT energy meters, time error Δt2 of inverters, and time error Δt3 of energy storage. The correlation between variables is constructed by using historical data from the past 24 hours. The final calibrated unified time reference data must meet the requirement that the time errors of single-phase smart IoT energy meters, inverters, and energy storage are all ≤ ±0.5 seconds, providing a consistent data time dimension for subsequent fault diagnosis and metering verification.

[0076] S102: Based on the unified time reference data, collect the operating data of the electricity meter, inverter and energy storage, and use the joint energy consumption model to distinguish between the hardware failure of the electricity meter itself and the abnormality of multi-device coordination, and complete the multi-dimensional collaborative fault diagnosis.

[0077] Specifically, the operating data includes: the energy consumption values ​​of the meter's display module, communication module, and data processing module under different states (sampling frequency 1 time / minute, continuously collected for 10 minutes and averaged to reduce the impact of fluctuations), the real-time output active power of the inverter (sampling frequency 1 time / 10 seconds), the charging and discharging current of the energy storage (sampling frequency 1 time / 10 seconds), and the SOC value (sampling frequency 1 time / minute).

[0078] The joint energy consumption model uses "electricity meter device energy consumption characteristics + multi-device collaborative interference factor" as its core framework. The electricity meter device energy consumption characteristics are determined by the energy consumption values ​​under different module operating states. The multi-device collaborative interference factor is derived from inverter power fluctuations and energy storage SOC changes. The model output is a "fault type judgment result", which can clearly distinguish between "electricity meter hardware failure" (such as abnormally high energy consumption of the display module) and "multi-device collaborative abnormality" (such as sudden changes in inverter power causing fluctuations in electricity meter data), avoiding misjudgments from single device data diagnosis.

[0079] S103: Based on the unified time reference data, combined with the line parameters and power data from the inverter output to the meter input, calculate the local line loss and verify the residual power metering deviation, trace the cause of the deviation and correct the settlement data, and complete the line loss-bidirectional metering linkage verification.

[0080] Specifically, the line parameters include line length (the measured length from the inverter output terminal to the meter input terminal, accurate to 0.1 meters) and conductor resistivity (determined based on the conductor material, e.g., 0.017 Ω·mm² for copper conductors). 2 / m, aluminum conductor is 0.028Ω·mm 2 / m) and conductor cross-sectional area (obtained from the line construction archives, accurate to 0.1mm). 2 ).

[0081] The power data includes the total output power of the inverter (the cumulative output power recorded by the inverter metering module, in kWh with an accuracy of 0.01 kWh), the user's self-consumption power (the cumulative power consumption recorded by the user-side load meter with an accuracy of 0.01 kWh), and the power consumption measured by the meter in reverse (the cumulative surplus power recorded by the meter and fed into the grid with an accuracy of 0.01 kWh).

[0082] Local line loss is determined by the formula P_line_loss = I_line_loss 2 Rt is calculated (I is the average output current of the inverter, R is the line resistance, and t is the set metering period, such as 1 hour). The deviation of the residual electricity metering is calculated by "deviation rate = |electricity meter reverse metering amount - theoretical residual electricity amount| / theoretical residual electricity amount × 100%" (theoretical residual electricity amount = total output electricity of the inverter - electricity consumed by the user - local line loss). When the deviation rate is > 5%, deviation tracing is initiated, and the final settlement data is corrected based on the "theoretical residual electricity amount" to ensure the accuracy of the residual electricity settlement.

[0083] S104: Associate the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-bidirectional metering linkage verification. Construct a quantitative model by statistically analyzing the correspondence between different fault types and metering deviation rate and line loss changes over a long period of time. Based on the model, predict the metering deviation trend and send early warnings to achieve dynamic improvement in metering accuracy.

[0084] Specifically, the long-term statistical period is more than 30 consecutive days, and the statistical objects include: the daily fault type (such as meter data processing module failure, inverter power fluctuation abnormality), the maximum value of the metering deviation rate on the day, and the average value of local line loss on the day; the quantitative model is constructed using a multiple linear regression algorithm, with "fault duration (days)" as the independent variable and "metering deviation rate increase (%)" and "line loss increase (%)" as the dependent variables. The model formula is as follows: ΔE=k1×t1+k2×t2+C (ΔE is the metering deviation rate increase, t1 is the duration of meter hardware failure, t2 is the duration of multi-device collaborative abnormality, k1 and k2 are influence coefficients, and C is a constant term). The coefficients are obtained by fitting historical statistical data (e.g., k1=0.2, which means that for every day the meter hardware failure lasts, the metering deviation rate increases by 0.2%).

[0085] The warning is triggered when “the metering deviation rate is predicted to exceed 8% within the next 7 days”. The warning information is sent to the equipment maintenance terminal (including the faulty module and the expected deviation value) via the Internet of Things, and is also pushed to the user terminal via SMS / APP to indicate that the energy consumption data may fluctuate.

[0086] S105: And / or, associate the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-two-way metering linkage verification, calculate the interference coefficient by determining the interference correlation between multi-device collaborative abnormal sources and transformer area loads, feed the interference coefficient back to the power grid dispatching system and mark the priority of line transformation to achieve power grid load-side management assistance.

[0087] Specifically, the multi-device collaborative anomaly source information includes inverter power fluctuation amplitude (the difference between the maximum and minimum active power within 10 minutes, in kW) and the proportion of energy storage charging and discharging current exceeding the rated value ((actual charging and discharging current - rated charging and discharging current) / rated charging and discharging current × 100%).

[0088] The distribution area load data includes the real-time value of the total load of the distribution area (unit: kW, sampling frequency: 1 time / minute) and the voltage fluctuation value of the user (the difference between the actual voltage and the rated voltage of 380V / 220V, unit: V); the interference coefficient is calculated by the formula η=α×(ΔP / Pn)+β×(ΔI / In) (ΔP is the inverter power fluctuation amplitude, Pn is the inverter rated power, ΔI is the energy storage current exceeding the rated value, In is the rated energy storage current, α=0.6, β=0.4, calibrated through the distribution area load influence experiment).

[0089] The priority for line renovation is determined by whether the local line loss value consistently exceeds the theoretical line loss value by 20% (the theoretical line loss value is calculated based on line parameters). If it consistently exceeds this value, the line is marked as a "Level 1 Renovation Target," and the line length, current line loss value, and suggested renovation plan (e.g., reducing the conductor cross-sectional area from 2.5mm²) are specified. 2 Upgraded to 4mm 2 Feedback is sent to the power grid planning terminal.

[0090] In one feasible implementation plan, see Figure 2 As shown, Figure 2 The flowchart of a time synchronization calibration method provided in Embodiment 1 of the present invention is shown. The method involves interpolating to unify the data granularity of the meter, inverter, and energy storage, and constructing a joint time error model to obtain calibrated unified time reference data, thereby completing the time synchronization calibration of multiple devices. This includes steps S201-S203.

[0091] S201: The cubic spline interpolation method is used to unify the data from different sampling periods of the electricity meter, inverter and energy storage into interpolated data with a set granularity by setting the interpolation ratio.

[0092] Specifically, the logic for determining the interpolation factor is to "take the shortest sampling period among the three types of equipment as the benchmark and calculate the ratio of the sampling period of other equipment to the benchmark period" (e.g., if the energy storage sampling period is 5 minutes as the benchmark, the interpolation factor is 2 if the inverter is 10 minutes and the interpolation factor is 3 if the meter is 15 minutes), ensuring that the interpolated data can completely retain the trend of the original data.

[0093] The implementation steps of cubic spline interpolation are as follows: First, obtain the original sampling data of each device (including timestamps and corresponding parameter values). Then, using the timestamp of the reference period as the interpolation node, fit the parameter change curve of each device through the cubic spline function, and finally generate interpolated data with a set granularity (such as 5 minutes). The error of the interpolated data must be ≤ 5% of the accuracy of the original data to avoid introducing additional deviations in the interpolation process.

[0094] S202: Using the time of the main meter in the distribution area as the reference time, a joint time error model is constructed, which includes meter time error, inverter time error, and energy storage time error. In this model, the interpolated electricity of the main meter in the distribution area under the reference time is equal to the sum of the meter interpolated electricity, the inverter interpolated power correction electricity, the energy storage interpolated current correction electricity, and the network loss.

[0095] Specifically, the mathematical expression of the joint time error model is: Etotal = Emeter + (Preverse × t - Δt2 × Preverse × k2) + (Istorage × Ustorage × t - Δt3 × Istorage × Ustorage × k3) + Enetwork loss, where Etotal is the interpolated electricity amount of the total meter in the distribution area (kWh), Emeter is the interpolated electricity amount of the meter (kWh), Preverse is the interpolated power of the inverter (kW), t is the metering cycle (h), Δt2 is the inverter time error (h), k2 is the inverter power correction coefficient (taken as 1.02, based on inverter efficiency calibration), Istorage... Let Ustore be the interpolated energy storage current (A), Ustore be the energy storage terminal voltage (V, taken as rated voltage), Δt3 be the energy storage time error (h), k3 be the energy storage current correction coefficient (taken as 1.01, based on energy storage charge and discharge efficiency calibration), and Enetwork loss be the network loss (kWh, estimated at 2% of the total interpolated electricity of the distribution area, which can be calibrated through historical data). The time errors Δt1 (meter), Δt2, and Δt3 in the model are variables to be solved. The system of equations is constructed by substituting multiple sets of historical interpolated data (such as one set of data per hour in the past 24 hours).

[0096] S203: Solve the joint time error model using the generalized least squares method to obtain the time error of each device. When the time error of a device is greater than the set value, send a time calibration command to it to ensure that the time error of the device after calibration is not greater than the set accuracy, and finally obtain unified time reference data.

[0097] Specifically, the solution steps of the generalized least squares method are as follows: First, the joint time error model is transformed into a set of error equations (with time error as the variable). Then, weights are assigned according to the reliability of each device's data (e.g., 0.6 for the main meter data of the distribution area, 0.2 for the meter data, 0.1 for the inverter data, and 0.1 for the energy storage data). The interpolated data of the past 24 hours are substituted to calculate the time error of each device. The set value is ±1 second (determined based on the metering accuracy requirements of the distribution area to ensure that the time deviation will not cause the metering error to exceed 0.1%), and the set accuracy is ±0.5 seconds.

[0098] The time calibration command is sent to the corresponding device via a wireless communication module (such as LoRa or 4G). After receiving the command, the device adjusts its local clock based on the time of the station's master meter. After calibration is completed, the calibration result is fed back. After the system verifies that the error meets the standard, it generates unified time reference data (including the timestamps and corresponding parameter values ​​of each device after calibration).

[0099] In one feasible implementation plan, see Figure 3 As shown, Figure 3 The flowchart of a fault diagnosis method provided in Embodiment 1 of the present invention is shown. The method involves collecting operational data from the meter, inverter, and energy storage based on the unified time reference data, using a joint energy consumption model to distinguish between hardware faults in the meter itself and multi-device collaborative anomalies, and completing multi-dimensional collaborative fault diagnosis, including steps S301-S305:

[0100] S301: Collects the energy consumption value of the meter under different module working states, where the first energy consumption value is when only the display module is working, the second energy consumption value is when the display module and the communication module work together, and the third energy consumption value is when the display module, the communication module and the data processing module work together.

[0101] Specifically, the communication module is a communication unit (single-phase / dual-mode / 2020 version smart energy meter). This communication unit supports 2G / 4G dual-mode communication and is the core module for data interaction between the single-phase smart IoT energy meter and the outside world. When collecting the second energy consumption value, it is necessary to control the communication unit (single-phase / dual-mode / 2020 version smart energy meter) to be in the actual working state of "4G network transmission of real-time energy consumption data of single-phase smart IoT energy meter" to avoid collection errors caused by static standby energy consumption.

[0102] The data acquisition process is triggered by the control unit built into the single-phase smart IoT energy meter: first, the display module is started separately (acquiring the first energy consumption value for 10 minutes, with a sampling frequency of 1 time / minute); then, the display module + communication unit (single-phase / dual-mode / 2020 version smart energy meter) is started (acquiring the second energy consumption value for 10 minutes); finally, the display module + communication unit + data processing module is started (acquiring the third energy consumption value for 10 minutes). The average value of each set of energy consumption values ​​is taken as the average of 10 samples to ensure data stability, and the acquisition accuracy needs to reach ±2%.

[0103] S302: Calculate the data processing capability coefficient of the electrical form device based on the first energy consumption value, the second energy consumption value, and the third energy consumption value.

[0104] Specifically, the calculation formula for the data processing capability coefficient K of the meter's data processing module is: K=(E3-E2) / E0, where E3-E2 is the energy consumption value when the data processing module works alone (since E3 is the total energy consumption of the three modules and E2 is the total energy consumption of the two modules, the difference is the energy consumption of the data processing module), and E0 is the rated energy consumption value of the meter's data processing module (obtained from the meter product manual, such as E0=5Wh / 10 minutes); the normal range of coefficient K is 0.8~1.2. When K is within this range, it means that the data processing capability of the meter's data processing module is normal. If it exceeds this range, it indicates that the meter's data processing module may have a fault (e.g., K>1.2 means that the energy consumption of the data processing module is too high, which may indicate that the hardware is aging).

[0105] S303: Collects the output active power and power fluctuation value of the inverter to calculate the power fluctuation coefficient, and collects the charging and discharging current and SOC value of the energy storage to calculate the SOC influence factor.

[0106] Specifically, the calculation logic for the power fluctuation coefficient C1 is as follows: collect the output active power data of the inverter within 10 minutes (a total of 60 data points, sampling frequency 1 time / 10 seconds), calculate the standard deviation σ of the active power within this time period, and then divide it by the rated power Pn of the inverter, i.e., C1=σ / Pn. The normal range of C1 is 0~0.1 (representing power fluctuation ≤10%). The calculation logic for the SOC impact factor C2 is as follows: collect the current SOC value S of the energy storage. If S is within the rated SOC range (e.g., 20%~80%), then C2=0; if S<20%, then C2=(20%-S) / 20%; if S>80%, then C2=(S-80%) / 20%. The value range of C2 is 0~1. The larger the value, the stronger the interference of the energy storage SOC on the coordination of multiple devices.

[0107] S304: The power fluctuation coefficient and the SOC influence factor are introduced as cooperative interference factors into the data processing capability coefficient of the electrical form equipment to construct the joint capability coefficient.

[0108] Specifically, the cooperative interference factor is introduced in a "weighted attenuation" manner. The calculation formula for the joint capability coefficient Kconnected is: Kconnected = K × (1 - C1 × 0.4 - C2 × 0.6), where 0.4 and 0.6 are the weights of the power fluctuation coefficient and the SOC influence factor, respectively. This is calibrated through multi-device cooperative experiments in the distribution area (experiments show that the impact of inverter power fluctuations on meter data is slightly less than that of energy storage SOC changes). The normal range of Kconnected is consistent with K (0.8~1.2). If Kconnected exceeds this range, it is necessary to further combine C1 and C2 to determine whether it is a single-device problem or a multi-device cooperative problem.

[0109] S305: Based on the joint capability coefficient and the preset theoretical model, the deviation capability index is calculated, and the fault type is determined by combining the numerical range of the cooperative interference factor.

[0110] Specifically, the preset theoretical model is the "theoretical range model of joint capability coefficient", that is, theoretically Kjoint should be within 0.8~1.2, and the calculation formula of deviation capability index D is: D=|Kjoint-Ktheoretical| / Ktheoretical, where Ktheoretical is the theoretical intermediate value of the joint capability coefficient (taken as 1.0); the numerical range judgment criteria of the cooperative interference factor are: C1≤0.1 and C2≤0.2 is "low interference", C1>0.1 or C2>0.2 is "high interference"; when D>0.2 (deviation capability index exceeds 20%) and "low interference", it is judged as a hardware failure of the meter itself; when D>0.2 and "high interference", it is judged as an abnormality of multi-device coordination.

[0111] In a feasible implementation, the fault type is determined by the numerical range of the combined cooperative interference factor, including:

[0112] When the deviation capability index is greater than the set threshold and the power fluctuation coefficient is not greater than the set value and the SOC influence factor is not greater than the set value, it is determined to be a hardware fault of the meter itself. The specific faulty module is located by calculating the edge density of each module.

[0113] Specifically, the set threshold is 0.2 (i.e., fault determination is triggered when the deviation capability index D > 0.2), the power fluctuation coefficient is set to 0.1, and the SOC influence factor is set to 0.2. The edge density is calculated as follows: using the energy consumption values ​​of each module of the meter during normal operation in the past 30 days as samples (a total of 900 data points, 30 per day), the kernel density estimation method is used to construct the probability density distribution curve of the energy consumption of each module. The currently collected E1, E2, and E3 are substituted into the distribution curve of the corresponding module to obtain the edge density value of each module (display module density ρ1, communication module density ρ2, data processing module density ρ3). The edge density is set to a threshold of 0.3. When the density value of a module is < 0.3, the module is determined to be a specific fault module (e.g., if ρ3 < 0.3, the data processing module is faulty).

[0114] When the deviation capability index is greater than the set threshold and the power fluctuation coefficient is greater than the set value or the SOC influence factor is greater than the set value, it is determined to be an abnormality in multi-device coordination, and a power stabilization control command is sent to the inverter or a charge / discharge level slow control command is sent to the energy storage.

[0115] Specifically, the set threshold, power fluctuation coefficient set value, and SOC impact factor set value are consistent with those mentioned above. If the power fluctuation coefficient C1 > 0.1, a power stabilization control command is sent to the inverter. The command includes "controlling the power fluctuation amplitude within 10% of the rated power" and "adjusting the inverter's PID control parameters (e.g., adjusting the proportional coefficient from 2.0 to 1.5 and the integral coefficient from 0.5 to 0.3)". If the SOC impact factor C2 > 0.2, a charge / discharge level easing control command is sent to the energy storage system. The command includes "when SOC < 20%, reduce the charging current to 50% of the rated current" and "when SOC > 80%, reduce the discharging current to 50% of the rated current". The command is sent through the wireless communication module. After the device executes the command, it provides feedback on the execution result. The system continuously monitors C1 and C2 until they return to within the set values.

[0116] In one feasible implementation plan, see Figure 4 As shown, Figure 4 The flowchart of a linkage verification method provided in Embodiment 1 of the present invention is shown, wherein the step of completing the line loss-bidirectional metering linkage verification based on unified time reference data includes steps S401 to S405:

[0117] S401: Obtain line parameters and power data from the inverter output terminal to the meter input terminal, collect the inverter output current, and calculate the local line loss based on the line length, conductor resistivity, output current and set metering cycle.

[0118] Specifically, the electricity meter includes energy meters (single-phase and three-phase). The acquisition of line parameters needs to be determined in conjunction with the installation location of the energy meter (single-phase and three-phase): For single-phase power consumption scenarios (such as residential users), the line length from the inverter output terminal to the single-phase energy meter input terminal is measured by a laser rangefinder (accurate to 0.1 meters from the inverter terminal to the single-phase energy meter terminal).

[0119] For three-phase power applications (such as industrial and commercial users), the line length is the measured distance from the inverter output to the three-phase energy meter input. The conductor resistivity is determined based on the conductor material (copper / aluminum) (0.017 Ω•mm for copper conductors). 2 / m, aluminum conductor is 0.028Ω•mm 2 / m); the inverter output current acquisition needs to be synchronized with the metering cycle of the energy meter (single-phase, three-phase) (set the metering cycle to 1 hour), and the average current within 1 hour is taken as the calculated current I. The formula for calculating local line loss is P line loss = I. 2 Rt (where R is the line resistance and t is the metering period) is calculated. The result needs to be matched with the metering data of single-phase and three-phase energy meters respectively, so as to provide a basis for subsequent residual power metering deviation verification of different types of energy meters.

[0120] S402: Collect the total output power of the inverter, the user's self-consumption power, and the reverse metering power. Subtract the user's self-consumption power and local line losses from the total output power of the inverter to derive the theoretical surplus power.

[0121] Specifically, the total output power of the inverter, Einverted_total, is the cumulative output power (in kWh, with an accuracy of 0.01 kWh) recorded by the inverter metering module within one hour; the user's self-consumption power, Eself-consumption, is the cumulative power consumption (in kWh, with an accuracy of 0.01 kWh) recorded by the user's load metering meter within one hour; and the meter-reverse metering power, Emeter-reverse, is the cumulative surplus power fed into the grid within one hour recorded by the meter (in kWh, with an accuracy of 0.01 kWh).

[0122] The theoretical surplus electricity Eliquid is derived as Eliquid = Ereverse total - Eself-use - Pline loss. During the derivation process, it is necessary to ensure that the time range of all electricity data is completely consistent with the set metering cycle (1 hour) (based on unified time reference data alignment) to avoid time misalignment leading to derivation errors.

[0123] S403: Calculate the deviation rate between the reverse metered electricity and the theoretical remaining electricity. When the deviation rate is greater than the set value, analyze the trend of the deviation curve through the local outlier factor algorithm to trace the cause of the deviation.

[0124] Specifically, the formula for calculating the deviation rate δ is δ=|Etable_reverse-Erational_surplus| / Erational_surplus×100%, with a set value of 5% (determined based on the accuracy requirements of residual electricity metering in the transformer area; a deviation rate ≤5% is considered normal). The implementation steps of the local outlier factor algorithm are as follows: using the deviation rate of each hour in the past 24 hours as data points (a total of 24), construct a deviation rate time series curve, set the sliding window size to 5 (i.e., analyze 5 consecutive data points each time), calculate the local outlier factor (LOF value) of each data point relative to other data points in the window, and determine the outlier point if the LOF value is >1.5. The cause of the deviation is traced through the distribution trend (gradual change or sudden change) of the outlier points.

[0125] The steps for tracing the cause of the deviation using the local outlier algorithm include: when the local outlier curve shows a gradual upward or downward trend, the cause of the deviation is determined to be aging of the metering chip; when the local outlier curve shows a sudden jump trend, the cause of the deviation is determined to be a local circuit abnormality.

[0126] Specifically, the quantitative standard for the gradual trend is: the slope of the local outlier curve for 5 consecutive data points is ≤0.05% / hour (i.e., the change in LOF value per hour does not exceed 0.05%), and the overall trend is upward or downward (e.g., the LOF value gradually increases from 1.2 to 1.8 over 5 consecutive hours); the criterion for judging the aging of the metering chip is: the aging of the metering chip will cause the metering accuracy to gradually drift, which is reflected in the deviation rate as a slow change rather than a sudden jump. This trend is consistent with the physical characteristics of chip aging, and the cause of the deviation can be verified by replacing the metering chip (the deviation rate should be restored to within 5% after replacement).

[0127] The quantitative standard for the sudden jump trend is: the change in the local outlier factor value of a single data point compared to the previous data point is >10% (e.g., the LOF value of the previous data point is 1.2, and the current data point suddenly rises to 1.5), and there is no obvious gradual change process in the early stage; local line anomalies include poor line contact (leading to increased resistance), line short circuit (leading to a sudden increase in line loss), etc. Such anomalies will cause sudden changes in local line loss, which will in turn cause the deviation rate to jump suddenly. This can be verified by line inspection (e.g., checking whether the wiring terminals are oxidized or whether the wires are damaged. After repair, the deviation rate should be restored to within 5%).

[0128] In one feasible implementation plan, see Figure 5 As shown, Figure 5 The flowchart of a deviation trend prediction method provided in Embodiment 1 of the present invention is shown. The method involves associating fault results from multi-dimensional collaborative fault diagnosis with metering data from line loss-bidirectional metering linkage verification. A quantitative model is constructed by statistically analyzing the correspondence between different fault types and metering deviation rates and line loss changes over a long period. Based on this model, the method predicts metering deviation trends and sends early warnings to achieve dynamic improvement in metering accuracy. The method includes steps S501-S504.

[0129] S501: Historical fault types and durations of long-term associated multi-dimensional collaborative fault diagnosis outputs, and historical measurement deviation rates and line loss change data of line loss-bidirectional measurement linkage verification outputs.

[0130] Specifically, "long-term" refers to continuous statistics for more than 30 days. The daily recorded data includes: the type of fault on the day (such as meter chip failure, abnormal inverter power fluctuation, or "none" if there is no fault), the duration of the fault on the day (in hours, or 0 if there is no fault), the maximum metering deviation rate (%) on the day, the average local line loss on the day (kWh), and the line loss change rate (the difference between the average line loss on the day and the average line loss on the previous day / the average line loss on the previous day × 100%). The data association method is to store the fault data and metering data in the database through a unified timestamp (accurate to the hour) to form a "fault-metering" associated dataset, ensuring the integrity of the data for subsequent modeling.

[0131] S502: Statistically analyze the correspondence of the above historical data and construct a quantitative model of the impact of faults on measurement accuracy. This model reflects the correlation between the duration of different fault types and the increase in measurement deviation rate and the increase in line loss.

[0132] Specifically, the statistical analysis employs a combination of correlation analysis and regression fitting: First, the correlation strength between the duration of different fault types and the increase in metering deviation rate and the proportion of increased line loss is calculated using the Pearson correlation coefficient (correlation strength > 0.7 is considered a strong correlation), and strongly correlated fault types (such as metering chip failure and poor contact in local lines) are screened out; then, with "fault duration (t, hours)" as the independent variable and "increase in metering deviation rate (Δδ, %)" as the dependent variable, a linear regression fitting is used to obtain a quantitative model formula, such as Δδ = 0.01 × t + 0.5 (representing that for every hour the metering chip failure lasts, the metering deviation rate increases by 0.01%, with a base deviation of 0.5%); the goodness of fit R is... 2 A value ≥ 0.8 is required to ensure that the model accurately reflects the correlation patterns.

[0133] S503: When the multi-dimensional collaborative fault diagnosis detects that the edge density of the target module of the meter is close to the set threshold, or when the local line resistance is slowly increasing as calculated by the local line loss value through the line loss-bidirectional metering linkage verification, the quantization model is invoked.

[0134] Specifically, the target module of the meter is the data processing module (because the failure of this module has the greatest impact on metering accuracy), and the edge density threshold is set at 0.35 (close to but not reaching the fault judgment threshold of 0.3, providing an early warning); the judgment criterion for the slow increase trend of local line resistance is: within 7 consecutive days, the daily calculated line resistance value (R=ρ×L / S, based on the line loss and current data of the day) increases by more than 1% compared with the previous day, and there is no obvious sudden change (excluding sudden line failure).

[0135] When any of the above conditions are met, the system automatically retrieves the corresponding quantization model from the database (e.g., if the edge density of the meter data processing module is detected to be close to the threshold, the "meter hardware failure - metering deviation" model is called).

[0136] S504: Based on a quantitative model, predict the trend of metering deviation rate changes within a set period in the future. If the predicted deviation rate will exceed the set threshold, send a metering accuracy warning to the equipment maintenance terminal and push energy consumption data fluctuation prompts to the user terminal to achieve early intervention in metering deviation.

[0137] Specifically, the future set period is 7 days, and the prediction process is as follows: Substitute the current fault duration (e.g., it has lasted for 24 hours) into the quantitative model to calculate the daily prediction measurement deviation rate for the next 7 days (e.g., if the current deviation rate is 2%, it will increase by 0.01% × 24 = 0.24% each day for the next 7 days, and the deviation rate after 7 days = 2% + 0.24% × 7 = 3.68%); set the threshold to 5%, and if the deviation rate is predicted to exceed 5% after 7 days (e.g., if the current deviation rate is 4.5%, it is expected to exceed 5% after 3 days), then an early warning will be triggered.

[0138] When early warning information is sent to the equipment maintenance terminal, it must include "predicted deviation rate, expected time of exceeding threshold, and recommended maintenance measures (such as replacing the meter data processing module)". When it is pushed to the user terminal, it must include "the reason for possible fluctuations in energy consumption data, the expected fluctuation time, and contact information (for user consultation)" to ensure that both maintenance personnel and users can respond in a timely manner.

[0139] In one feasible implementation plan, see Figure 6 As shown, Figure 6 The flowchart of a management assistance method provided in Embodiment 1 of the present invention is shown. The method involves associating the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-bidirectional metering linkage verification. By determining the interference correlation between multi-device collaborative anomaly sources and transformer loads, an interference coefficient is calculated. This interference coefficient is then fed back to the power grid dispatching system, and line modification priorities are marked to achieve power grid load-side management assistance. The method includes steps S601-S604.

[0140] S601: The multi-device collaborative anomaly source information output by the multi-dimensional collaborative fault diagnosis is associated and matched with the local line loss value and user power load data output by the line loss-bidirectional metering linkage verification.

[0141] Specifically, the multi-device collaborative anomaly source information includes the anomaly source type (such as inverter power fluctuation, energy storage charging and discharging anomaly), anomaly parameter values ​​(such as inverter power fluctuation amplitude of 5kW, energy storage charging current exceeding the rated value by 20%), and anomaly occurrence time (accurate to the minute); the local line loss value is the average line loss value within the anomaly occurrence period (such as the line loss average value for 10 minutes if the anomaly occurs between 9:00 and 9:10); the user electricity load data is the average total load of the transformer area (unit: kW) and the average voltage at the user end (unit: V) within the anomaly occurrence period; the association matching method is to align the three types of data based on the "anomaly occurrence time", and a time deviation of ≤1 minute is considered a successful match, ensuring that the data can accurately reflect the impact of the anomaly source on the load.

[0142] S602: Based on the matching results, determine the correlation between the coordinated abnormal source and the local line loss change in the distribution area and the voltage fluctuation of the user's electricity consumption. The coordinated abnormal source includes inverter power fluctuation and high-power charging of energy storage. Calculate the interference coefficient of the coordinated abnormal source on the distribution area load. This coefficient reflects the corresponding ratio between the abnormality degree of the coordinated abnormal source and the interference degree of the distribution area load.

[0143] Specifically, the correlation is determined through "comparison before and after the anomaly": comparing the local line loss change rate (ΔP line loss / P line loss before × 100%) and voltage fluctuation change rate (ΔU / U before × 100%) 10 minutes before the anomaly (e.g., 8:50-9:00) and the time of the anomaly (9:00-9:10). If the change rate is positively correlated with the abnormal parameter value (e.g., the larger the power fluctuation amplitude, the larger the line loss change rate and voltage fluctuation change rate), then a correlation is determined to exist; the interference coefficient η is calculated by the formula η=0.5×(ΔP reverse / Pn)+0.5×(ΔI storage / In), where ΔP reverse is the inverter power fluctuation amplitude (kW), Pn is the inverter rated power (kW), ΔI storage is the energy storage charging and discharging current exceeding the rated value (A), and In is the energy storage rated charging and discharging current (A). The coefficient range is 0~1. η>0.3 indicates strong interference, which needs to be fed back to the dispatch system.

[0144] S603: The interference coefficient is fed back to the power grid dispatching system of the distribution area. Based on the interference coefficient, the dispatching system adjusts the inverter power output or energy storage charging and discharging parameters during the abnormal peak period of the coordinated abnormal source, and balances the load distribution of adjacent distribution areas to avoid transformer overload or voltage abnormality in the distribution area.

[0145] Specifically, the load distribution balancing process involves three-phase smart IoT energy meters. Commands are issued through a local communication unit (Concentrator Type I / Dual Mode): the local communication unit (Concentrator Type I / Dual Mode) supports power line carrier (PLC) and low-power wireless dual-mode communication. One end connects to the power grid dispatching system of the distribution area, and the other end covers the three-phase smart IoT energy meters, single-phase smart IoT energy meters, and inverters within the distribution area. When the interference coefficient η > 0.3 (set threshold) and the abnormal source is inverter power fluctuation, the dispatching system generates a command to "reduce the upper limit of inverter power output by 10%", which is sent to the inverter through the PLC channel of the local communication unit (Concentrator Type I / Dual Mode).

[0146] Meanwhile, for three-phase electricity users (such as small industrial and commercial users) within the distribution area, the dispatch system sends a "lower priority for power supply to non-essential loads" instruction to the three-phase smart IoT energy meter through the local communication unit (concentrator type I / dual mode). After receiving the instruction, the three-phase smart IoT energy meter automatically adjusts its internal load control module to prioritize power supply to necessary production equipment, reducing the impact on the total load of the distribution area. The entire instruction transmission delay is ≤5 seconds, ensuring a rapid response during abnormal peak periods (such as 14:00-16:00) and preventing the distribution area transformer load rate from exceeding 80% and the user-end voltage from deviating from the rated range of 380V±5%.

[0147] S604: Long-term monitoring of line loss - local line loss data output by bidirectional metering linkage verification. For inverter-to-meter lines where the local line loss value exceeds the set range for a long period of time, the line is marked as a priority object for line renovation and fed back to the power grid planning terminal.

[0148] Specifically, the long-term monitoring period is 30 consecutive days, with a set range of "120% of the theoretical line loss value" (the theoretical line loss value is calculated based on line parameters, such as a theoretical line loss of 0.5 kWh / hour, with a set range of 0.6 kWh / hour); the criteria for determining the priority of line renovation are: if the local line loss value exceeds the set range for more than 20 days within 30 consecutive days, it is marked as a "Level 1 renovation target"; if it exceeds the range for 10-19 days, it is marked as a "Level 2 renovation target"; if it exceeds the range for less than 10 days, it is marked as an "observation target"; the information fed back to the power grid planning terminal includes "line number, current line loss value, theoretical line loss value, number of days exceeding the range, priority level, and suggested renovation plan (e.g., 2.5 mm)". 2 The copper wire has been upgraded to 4mm. 2 "Copper conductors are expected to reduce line loss by 30% after the upgrade," providing accurate data support for power grid planning.

[0149] Example 2

[0150] See Figure 7 As shown, Figure 7The diagram shows a structural schematic of a multi-device collaborative optimization device for smart meters oriented towards distributed energy access, provided in Embodiment 2 of the present invention. The device includes:

[0151] The time synchronization calibration module 701 is used to interpolate the data granularity of meters, inverters and energy storage and build a joint time error model to obtain calibrated unified time reference data and complete the time synchronization calibration of multiple devices.

[0152] The fault diagnosis module 702 is used to collect the operating data of the electricity meter, inverter and energy storage based on the unified time base data, and use the joint energy consumption model to distinguish between the hardware fault of the electricity meter itself and the abnormality of multi-device coordination, so as to complete the multi-dimensional collaborative fault diagnosis.

[0153] The linkage verification module 703 is used to calculate local line loss and verify residual power metering deviation based on the unified time reference data, combined with the line parameters and power data from the inverter output terminal to the meter input terminal, trace the cause of the deviation and correct the settlement data, and complete the linkage verification of line loss-bidirectional metering.

[0154] The deviation trend prediction module 704 is used to associate the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-bidirectional metering linkage verification. It constructs a quantitative model by statistically analyzing the correspondence between different fault types and metering deviation rate and line loss changes over a long period of time. Based on the model, it predicts the metering deviation trend and sends early warnings to achieve dynamic improvement in metering accuracy.

[0155] The management auxiliary module 705 is used to associate the fault results output by multi-dimensional collaborative fault diagnosis with the metering data output by line loss-bidirectional metering linkage verification. It calculates the interference coefficient by determining the interference correlation between multi-device collaborative abnormal sources and transformer area loads, feeds the interference coefficient back to the power grid dispatching system, and marks the priority of line transformation to achieve power grid load-side management assistance.

[0156] Among them, the fault diagnosis module and the linkage verification module achieve collaboration through data feedback, and the time synchronization calibration module provides the data foundation for the former two and subsequent related processing.

[0157] In a feasible implementation, the process of interpolating to unify the data granularity of meters, inverters, and energy storage, and constructing a joint time error model to obtain calibrated unified time reference data, thereby completing multi-device time synchronization calibration, includes:

[0158] The cubic spline interpolation method is used to unify the data from different sampling periods of the electricity meter, inverter and energy storage into interpolated data with a set granularity by setting the interpolation ratio;

[0159] Using the time of the main meter in the distribution area as the reference time, a joint time error model is constructed, which includes meter time error, inverter time error, and energy storage time error. In this model, the interpolated electricity of the main meter in the distribution area under the reference time is equal to the sum of the meter interpolated electricity, the inverter interpolated power correction electricity, the energy storage interpolated current correction electricity, and the network loss.

[0160] The joint time error model is solved by generalized least squares method to obtain the time error of each device. When the time error of a device is greater than the set value, a time calibration command is sent to it to ensure that the time error of the device after calibration is not greater than the set accuracy, and finally a unified time reference data is obtained.

[0161] In one feasible implementation, based on the unified time base data, operational data from meters, inverters, and energy storage are collected. A joint energy consumption model is used to distinguish between meter-specific hardware faults and multi-device collaborative anomalies, completing multi-dimensional collaborative fault diagnosis, including:

[0162] The energy consumption values ​​of the electricity meter are collected under different module working states. The first energy consumption value is when only the display module is working, the second energy consumption value is when the display module and the communication module work together, and the third energy consumption value is when the display module, the communication module and the data processing module work together.

[0163] Based on the first energy consumption value, the second energy consumption value, and the third energy consumption value, the data processing capability coefficient of the electrical form equipment is calculated.

[0164] The output active power and power fluctuation value of the inverter are collected to calculate the power fluctuation coefficient, and the charging and discharging current and SOC value of the energy storage are collected to calculate the SOC influence factor.

[0165] The power fluctuation coefficient and the SOC impact factor are introduced as cooperative interference factors into the data processing capability coefficient of electrical equipment to construct a joint capability coefficient.

[0166] The deviation capability index is calculated based on the joint capability coefficient and the preset theoretical model. Combined with the numerical range of the cooperative interference factor, the fault type is determined.

[0167] In a feasible implementation, the fault type is determined by the numerical range of the combined cooperative interference factor, including:

[0168] When the deviation capability index is greater than the set threshold and the power fluctuation coefficient is not greater than the set value and the SOC influence factor is not greater than the set value, it is determined to be a hardware fault of the meter itself. The specific faulty module is located by calculating the edge density of each module.

[0169] When the deviation capability index is greater than the set threshold and the power fluctuation coefficient is greater than the set value or the SOC influence factor is greater than the set value, it is determined to be an abnormality in multi-device coordination, and a power stabilization control command is sent to the inverter or a charge / discharge level slow control command is sent to the energy storage.

[0170] In a feasible implementation, based on the unified time reference data, combined with the line parameters and power consumption data from the inverter output to the meter input, local line losses are calculated and residual power metering deviations are verified. The cause of the deviation is traced and the settlement data is corrected, thus completing the line loss-bidirectional metering linkage verification, including:

[0171] The line length and conductor resistivity from the inverter output terminal to the meter input terminal are obtained, the inverter output current is collected, and the local line loss is calculated based on the line length, conductor resistivity, output current and set metering cycle.

[0172] Collect the total output power of the inverter, the user's own power consumption, and the reverse metering power. Subtract the user's own power consumption and local line loss from the total output power of the inverter to derive the theoretical surplus power.

[0173] Calculate the deviation rate between the reverse metered electricity and the theoretical residual electricity. When the deviation rate is greater than the set value, analyze the trend of the deviation curve through the local outlier factor algorithm to trace the cause of the deviation.

[0174] The steps for tracing the cause of the deviation using the local outlier factor algorithm include:

[0175] When the local outlier curve shows a gradual upward or downward trend, the cause of the deviation is determined to be the aging of the meter's metering chip.

[0176] When the local outlier curve shows a sudden jump trend, the cause of the deviation is determined to be a local circuit anomaly.

[0177] In a feasible implementation, the fault results output by the multi-dimensional collaborative fault diagnosis and the metering data output by the line loss-bidirectional metering linkage verification are used to construct a quantitative model by long-term statistical analysis of the correspondence between different fault types and metering deviation rates and line loss changes. Based on the model, the metering deviation trend is predicted and early warnings are sent to achieve dynamic improvement in metering accuracy, including:

[0178] Long-term correlation with the historical fault types and durations of multi-dimensional collaborative fault diagnosis outputs and the historical metering deviation rate and line loss change data of the line loss-bidirectional metering linkage verification outputs;

[0179] By statistically analyzing the correspondence between the above historical data, a quantitative model of the impact of faults on measurement accuracy is constructed. This model reflects the correlation between the duration of different fault types and the increase in measurement deviation rate and the increase in line loss.

[0180] When the multi-dimensional collaborative fault diagnosis detects that the edge density of the target module of the meter is close to the set threshold, or when the local line resistance is slowly increasing as calculated by the local line loss value through the line loss-bidirectional metering linkage verification, the quantization model is invoked.

[0181] Based on the quantitative model, the trend of metering deviation rate changes within a set period is predicted. If the predicted deviation rate exceeds the set threshold, a metering accuracy warning message is sent to the equipment maintenance terminal, and at the same time, an energy consumption data fluctuation prompt message is pushed to the user terminal, so as to realize early intervention of metering deviation.

[0182] In a feasible implementation, the fault results output by the multi-dimensional collaborative fault diagnosis and the metering data output by the line loss-bidirectional metering linkage verification are used to calculate the interference coefficient by determining the interference correlation between the multi-device collaborative anomaly source and the transformer area load. The interference coefficient is then fed back to the power grid dispatching system and the line modification priority is marked to achieve power grid load-side management assistance, including:

[0183] The multi-device collaborative anomaly source information output by multi-dimensional collaborative fault diagnosis is correlated and matched with the local line loss value and user electricity load data output by line loss-bidirectional metering linkage verification.

[0184] Based on the matching results, the correlation between the coordinated abnormal sources and the changes in local line loss in the distribution area and the fluctuations in user electricity voltage is determined. The coordinated abnormal sources include inverter power fluctuations and high-power charging of energy storage. The interference coefficient of the coordinated abnormal sources on the distribution area load is calculated. This coefficient reflects the corresponding ratio between the degree of abnormality of the coordinated abnormal sources and the degree of interference of the distribution area load.

[0185] The interference coefficient is fed back to the power grid dispatching system of the distribution area. Based on the interference coefficient, the dispatching system adjusts the inverter power output or energy storage charging and discharging parameters during the abnormal peak period of the coordinated abnormal source, and at the same time balances the load distribution of adjacent distribution areas to avoid transformer overload or voltage abnormality in the distribution area.

[0186] Long-term monitoring of line loss - local line loss data output by bidirectional metering linkage verification. Inverter-to-meter lines with local line loss values ​​that exceed the set range for a long time are marked as priority targets for line renovation and fed back to the power grid planning terminal.

[0187] Example 3

[0188] Based on the same application concept, see [link / reference] Figure 8 As shown, Figure 8 A schematic diagram of the structure of a computer device provided in Embodiment 3 of the present invention is shown, wherein, as Figure 8 As shown, the computer device 800 provided in Embodiment 3 of this application includes:

[0189] The system includes a processor 801, a memory 802, and a bus 803. The memory 802 stores machine-readable instructions executable by the processor 801. When the computer device 800 is running, the processor 801 and the memory 802 communicate via the bus 803. When the machine-readable instructions are executed by the processor 801, they perform the steps of the multi-device collaborative optimization method for smart meters oriented towards distributed energy access as shown in Embodiment 1 above.

[0190] Example 4

[0191] Based on the same concept, this application also provides a computer-readable storage medium storing a computer program, which, when run by a processor, executes the steps of the multi-device collaborative optimization method for smart meters oriented towards distributed energy access as described in any of the above embodiments.

[0192] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and apparatus described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0193] The computer program product for multi-device collaborative optimization of smart meters for distributed energy access provided in this embodiment of the invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0194] The multi-device collaborative optimization device for smart meters oriented towards distributed energy access provided in this embodiment of the invention can be specific hardware on the device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in this embodiment of the invention are the same as those in the foregoing method embodiments. For the sake of brevity, any parts not mentioned in the device embodiments can be referred to the corresponding content in the foregoing method embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can all be referred to the corresponding processes in the above method embodiments, and will not be repeated here.

[0195] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0196] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0197] In addition, the functional units in the embodiments provided by the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0198] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0199] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0200] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. All should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for multi-device collaborative optimization of smart meters oriented to distributed energy access, characterized in that, The method comprises: The method comprises: Based on the unified time reference data, the operation data of the electric meter, the inverter and the energy storage are collected, the joint energy consumption model is used to distinguish the electric meter hardware failure and the multi-device collaborative anomaly, and multi-dimensional collaborative fault diagnosis is completed; Based on the unified time reference data, the line loss is calculated and the residual power measurement deviation is checked by combining the line parameters from the inverter output end to the electric meter input end and the power data, the deviation reason is traced to correct the settlement data, and the line loss-bidirectional measurement linkage verification is completed; Correlate the fault results output by the multi-dimensional collaborative fault diagnosis with the measurement data output by the line loss-bidirectional measurement linkage verification, build a quantitative model by long-term statistics of the corresponding relationship between different fault types and measurement deviation rate and line loss change, predict the measurement deviation trend based on the model and send an early warning to realize dynamic improvement of measurement accuracy; And / or, correlate the fault results output by the multi-dimensional collaborative fault diagnosis with the measurement data output by the line loss-bidirectional measurement linkage verification, calculate the interference coefficient by determining the interference correlation between the multi-device collaborative anomaly source and the transformer area load, feed back the interference coefficient to the power grid dispatching system and mark the line reconstruction priority to realize power grid load side management assistance; Wherein, the multi-dimensional collaborative fault diagnosis and the line loss-bidirectional measurement linkage verification realize cooperation through data feedback, and the multi-device time synchronization calibration provides data basis for the former two and subsequent correlation processing; The method comprises: A cubic spline interpolation method is used to unify the data of different sampling periods of the electric meter, the inverter and the energy storage into interpolation data of a set granularity by setting an interpolation multiple; A joint time error model including the electric meter time error, the inverter time error and the energy storage time error is constructed based on the transformer area total meter time as the reference time, and in the model, the transformer area total meter interpolation power under the reference time is equal to the sum of the electric meter interpolation power, the inverter interpolation power corrected power, the energy storage interpolation current corrected power and the network loss; The joint time error model is solved by the generalized least squares method to obtain the time error of each device, and when the device time error is greater than a set value, a time calibration instruction is sent to it to ensure that the calibrated device time error is not greater than a set accuracy, and finally the unified time reference data is obtained; The method comprises: The energy consumption values of the electric meter under different module working states are collected, wherein only the first energy consumption value is displayed when the module works, the second energy consumption value is displayed when the communication module and the communication module work together, and the third energy consumption value is displayed when the display module, the communication module and the data processing module work together; The single device data processing capacity coefficient of the electric meter is calculated according to the first energy consumption value, the second energy consumption value and the third energy consumption value; The output active power and power fluctuation value of the inverter are collected to calculate a power fluctuation coefficient, and the charging and discharging current and SOC value of the energy storage are collected to calculate an SOC influence factor; The power fluctuation coefficient and the SOC influence factor are introduced as a synergistic interference factor into the meter single device data processing capability coefficient to construct a joint capability coefficient; Based on the joint capability coefficient and a preset theoretical model, a deviation capability index is calculated, and a fault type is determined according to the numerical range of the synergistic interference factor; The fault result output by the multi-dimensional synergistic fault diagnosis is associated with the metering data output by the line loss-bidirectional metering linkage verification, a quantitative model is constructed by long-term statistics of the corresponding relationship between different fault types and the metering deviation rate and the line loss change, the metering deviation trend is predicted based on the model, and a warning is sent to realize dynamic improvement of the metering accuracy, including: The historical fault type and fault duration output by the multi-dimensional synergistic fault diagnosis are associated with the historical metering deviation rate and line loss change data output by the line loss-bidirectional metering linkage verification; The corresponding relationship of the historical data is counted, a quantitative model of the influence of faults on metering accuracy is constructed, and the model reflects the correlation law of the duration of different fault types and the rising amplitude of the metering deviation rate and the increasing proportion of the line loss; When the multi-dimensional synergistic fault diagnosis detects that the edge density of the meter target module is close to the set threshold, or the local line resistance shows a slow increasing trend through the local line loss value calculated by the line loss-bidirectional metering linkage verification, the quantitative model is called; Based on the quantitative model, the metering deviation rate change trend in the future set period is predicted, if the predicted deviation rate will exceed the set threshold, the metering accuracy warning information is sent to the device maintenance terminal, and the energy consumption data fluctuation prompt information is pushed to the user end, so that the metering deviation is intervened in advance.

2. The method of claim 1, wherein, The numerical range of the combined synergistic interference factor is used to determine the fault type, including: When the deviation capability index is greater than the set threshold, and the power fluctuation coefficient is not greater than the set value, and the SOC influence factor is not greater than the set value, it is determined that the meter itself hardware fault, and the specific fault module is located by calculating the edge density of each module; When the deviation capability index is greater than the set threshold, and the power fluctuation coefficient is greater than the set value or the SOC influence factor is greater than the set value, it is determined that the multi-device synergistic anomaly, and the power stabilization control instruction is sent to the inverter or the charging and discharging gentle control instruction is sent to the energy storage.

3. The method of claim 1, wherein, Based on the unified time reference data, the line loss is calculated and the residual electricity metering deviation is verified by combining the line parameters and the electric quantity data from the inverter output end to the meter input end, the deviation reason is traced to correct the settlement data, the line loss-bidirectional metering linkage verification is completed, including: The line length and wire resistivity from the inverter output end to the meter input end are obtained, the inverter output current is collected, and the local line loss is calculated based on the line length, wire resistivity, output current and set metering period; The total inverter output electric quantity, user self-use electric quantity and meter reverse metering electric quantity are collected, the total inverter output electric quantity is subtracted by the user self-use electric quantity and the local line loss, and the theoretical residual electric quantity is derived. The deviation rate of the reverse metering electric quantity and the theoretical residual electric quantity is calculated, when the deviation rate is greater than a set value, the deviation curve trend is analyzed by a local outlier factor algorithm, and the deviation reason is traced back to the source; The step of tracing back to the source of the deviation by the local outlier factor algorithm comprises: When the local outlier factor curve presents a gradual rising or falling trend, it is determined that the deviation reason is the aging of the metering chip of the electric meter; When the local outlier factor curve presents a sudden jump trend, it is determined that the deviation reason is a local line abnormality.

4. The method of claim 1, wherein, The fault result output by the multi-dimensional collaborative fault diagnosis is associated with the metering data output by the line loss-bidirectional metering linkage verification, an interference coefficient is calculated by determining the interference correlation between the multi-device collaborative abnormal source and the substation load, the interference coefficient is fed back to the power grid dispatching system and the line reconstruction priority is marked to realize power grid load side management assistance, comprising: The multi-device collaborative abnormal source information output by the multi-dimensional collaborative fault diagnosis is associated and matched with the local line loss value and the user power consumption load data output by the line loss-bidirectional metering linkage verification; Based on the matching result, the correlation between the collaborative abnormal source and the local line loss change of the substation and the user power consumption voltage fluctuation is determined, wherein the collaborative abnormal source includes inverter power fluctuation and large power charging of energy storage, and the interference coefficient of the collaborative abnormal source to the substation load is calculated, which reflects the corresponding proportion of the abnormal degree of the collaborative abnormal source and the interference degree of the substation load; The interference coefficient is fed back to the substation power grid dispatching system, and the dispatching system adjusts the inverter power output or the energy storage charging and discharging parameters during the abnormal peak period of the collaborative abnormal source based on the interference coefficient, balances the load distribution of adjacent substations, and avoids substation transformer overload or voltage abnormality; Long-term monitoring of the local line loss data output by the line loss-bidirectional metering linkage verification, the inverter-to-electric meter line with a local line loss value that is long-term beyond the set range is marked as a line reconstruction priority object and fed back to the power grid planning terminal.

5. A smart meter multi-device collaborative optimization device for distributed energy access, characterized in that, The device comprises: A time synchronization calibration module is configured to unify the data granularity of the electric meter, the inverter and the energy storage by interpolation and construct a joint time error model to obtain calibrated unified time reference data, and complete multi-device time synchronization calibration; A fault diagnosis module is configured to collect the operation data of the electric meter, the inverter and the energy storage based on the unified time reference data, distinguish the electric meter hardware fault and the multi-device collaborative abnormality by using a joint energy consumption model, and complete multi-dimensional collaborative fault diagnosis; A linkage verification module is configured to calculate the local line loss and verify the residual electric quantity metering deviation based on the unified time reference data, trace back to the source of the deviation and correct the settlement data by combining the line parameters and the electric quantity data from the inverter output end to the electric meter input end, and complete line loss-bidirectional metering linkage verification; A deviation trend prediction module is configured to associate the fault result output by the multi-dimensional collaborative fault diagnosis with the metering data output by the line loss-bidirectional metering linkage verification, construct a quantitative model by long-term statistics of the corresponding relationship between different fault types and metering deviation rate and line loss change, predict the metering deviation trend based on the model and send an early warning to realize dynamic improvement of metering accuracy. The management auxiliary module is used for associating fault results of multi-dimensional collaborative fault diagnosis output with metering data of line loss-bidirectional metering linkage verification output, calculating an interference coefficient by determining an interference correlation between multi-device collaborative abnormal sources and a load of a transformer area, feeding back the interference coefficient to a power grid dispatching system and marking a line reconstruction priority to realize power grid load side management assistance; The fault diagnosis module and the linkage verification module are cooperated through data feedback, and the time synchronization calibration module provides a data basis for the former two and subsequent associated processing; The data granularity of the electric meter, the inverter and the energy storage is unified through interpolation, and a joint time error model is constructed to obtain calibrated unified time reference data, and multi-device time synchronization calibration is completed, including: A cubic spline interpolation method is used to unify the data of different sampling periods of the electric meter, the inverter and the energy storage into interpolation data of a set granularity by setting an interpolation multiple; A joint time error model including time errors of the electric meter, the inverter and the energy storage is constructed with the transformer area total meter time as the reference time, and in the model, the transformer area total meter interpolation electric quantity at the reference time is equal to the sum of the electric meter interpolation electric quantity, the inverter interpolation power corrected electric quantity, the energy storage interpolation current corrected electric quantity and the network loss; The joint time error model is solved by the generalized least squares method to obtain the time errors of each device, and when the device time error is greater than a set value, a time calibration instruction is sent to it to ensure that the calibrated device time error is not greater than a set precision, and finally the unified time reference data is obtained; Based on the unified time reference data, the operation data of the electric meter, the inverter and the energy storage are collected, and the joint energy consumption model is used to distinguish the electric meter hardware fault and the multi-device collaborative abnormality, and multi-dimensional collaborative fault diagnosis is completed, including: The energy consumption values of the electric meter under different module working states are collected, wherein the first energy consumption value is only displayed when the module is working, the second energy consumption value is displayed when the display module and the communication module are working together, and the third energy consumption value is displayed when the display module, the communication module and the data processing module are working together; The electric meter single device data processing capability coefficient is calculated according to the first, second and third energy consumption values; The output active power and power fluctuation value of the inverter are collected to calculate the power fluctuation coefficient, and the charge and discharge current and SOC value of the energy storage are collected to calculate the SOC influence factor; The power fluctuation coefficient and the SOC influence factor are introduced as collaborative interference factors into the electric meter single device data processing capability coefficient to construct the joint capability coefficient; Based on the joint capability coefficient and the preset theoretical model calculation deviation capability index, the fault type is judged in combination with the numerical range of the collaborative interference factor; The association of the fault results of multi-dimensional collaborative fault diagnosis output and the metering data of line loss-bidirectional metering linkage verification output is constructed, a quantitative model is constructed by long-term statistics of the corresponding relationship between different fault types and metering deviation rate and line loss change, the metering deviation trend is predicted based on the model, and a warning is sent to realize dynamic improvement of metering accuracy, including: The historical fault type and fault duration of multi-dimensional collaborative fault diagnosis output are associated with the historical metering deviation rate and line loss change data of line loss-bidirectional metering linkage verification output for a long time. Corresponding relationship of the historical data is counted, a quantitative model of influence of fault on metering accuracy is constructed, and the model reflects the correlation law of duration of different fault types and rising amplitude of metering deviation rate and increasing proportion of line loss; When the multi-dimensional collaborative fault diagnosis detects that the edge density of the target module of the electric meter is close to the set threshold, or the local line resistance shows a slow increasing trend through the local line loss value calculated by the line loss-bidirectional metering linkage verification, the quantitative model is called; Based on the quantitative model, the change trend of the metering deviation rate in the future set period is predicted, if the predicted deviation rate will exceed the set threshold, the metering accuracy early warning information is sent to the equipment maintenance terminal, and the energy consumption data fluctuation prompt information is pushed to the user end, so that the early intervention of the metering deviation is realized.

6. A computer device, comprising: It comprises a processor, a memory and a bus, the memory stores machine readable instructions executable by the processor, when the computer equipment runs, the processor and the memory communicate through the bus, and the machine readable instructions are executed by the processor to execute the steps of the multi-device collaborative optimization method of the smart electric meter facing the distributed energy access as claimed in any one of claims 1 to 4. The computer readable storage medium stores a computer program, and the computer program is executed by the processor to execute the steps of the multi-device collaborative optimization method of the smart electric meter facing the distributed energy access as claimed in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, ​

Images