An energy meter with self-identification of abnormal power consumption patterns and its intelligent mode switching method

Abstract

This invention discloses an energy meter with self-identification of abnormal electricity consumption patterns and its intelligent mode switching method, belonging to the field of energy meter testing technology. The energy meter integrates a database configuration module, a cyclic sampling and monitoring module, a membership analysis module, and a mode switching evaluation module. The application method is as follows: configure an electricity consumption parameter database and an electricity consumption status database, storing electricity consumption parameter sample clusters categorized by scenario and warning sample feature values ​​categorized by pattern, respectively; generate a monitoring task input set by traversing all combinations of electricity consumption parameters through a cyclic sampling model, and obtain a real-time warning feature value set; calculate membership based on the maximum / minimum difference, match the optimal electricity consumption status sample cluster, and generate a switching link label; evaluate the mode switching conditions through the cumulative deviation value and trigger the warning. This invention solves the problems of poor adaptability and delayed response caused by the reliance on fixed thresholds for mode switching in traditional energy meters, and achieves accurate identification and proactive switching of abnormal modes in multiple scenarios.

Description

Technical Field

[0001] This invention relates to the field of electricity meter testing technology, specifically to an electricity meter with self-identification of abnormal electricity consumption patterns and a method for intelligent mode switching. Background Technology

[0002] With the development of industrial intelligence and diverse electricity consumption scenarios, electricity meters need to dynamically adapt to different electricity consumption modes (such as normal, overload, short circuit, etc.) to achieve accurate early warning. Existing technologies have the following shortcomings:

[0003] Threshold rigidity: Traditional electricity meters rely on preset fixed thresholds to judge abnormal electricity consumption (such as overvoltage and overload thresholds), which cannot adapt to the dynamic changes of different scenarios (such as residential electricity consumption during off-peak / peak hours, and industrial equipment start-up and shutdown), and are prone to false alarms or missed alarms. For example, the short-term high current when a residential air conditioner starts may trigger a false alarm, while the slight current fluctuations in an industrial production line may be ignored because they do not reach the threshold.

[0004] Sampling limitations: Using fixed-period or fixed-parameter sampling makes it difficult to cover diverse combinations of power parameters (such as cross-variations of voltage, current, and power factor), potentially missing key anomalies. For example, monitoring only current cannot identify the coordinated anomalies of voltage drops and current distortions.

[0005] Passive mode switching: Manual intervention is required to switch operating modes (such as from "routine monitoring" to "troubleshooting"), resulting in delayed response. When there are sudden fluctuations in electricity load in commercial complexes, traditional electricity meters cannot automatically switch to high-frequency monitoring mode, delaying the opportunity to handle anomalies.

[0006] Weak data correlation: The correlation between power consumption parameters and abnormal patterns lacks a quantitative model. Judging the state based on a single indicator makes it difficult to distinguish between "normal fluctuations" and "abnormal trends." For example, it is impossible to identify potential faults caused by the accumulation of long-term, small-amplitude power deviations.

[0007] Existing technologies are insufficient to meet the requirements for accurate identification and dynamic adaptation of abnormal electricity consumption in various scenarios. There is an urgent need for an electricity meter and method that can automatically identify abnormal patterns and intelligently switch operating states. Summary of the Invention

[0008] The purpose of this invention is to provide an energy meter with self-identification of abnormal electricity consumption patterns and a method for intelligent mode switching, in order to solve the problems mentioned in the background art. Specifically, this invention achieves intelligent switching through the logic of "database modeling - cyclic sampling - membership matching - deviation evaluation", including:

[0009] Electricity parameter sample clusters are stored according to electricity usage scenarios (such as residential, industrial, and commercial), and early warning sample feature values ​​are stored according to electricity usage modes (such as normal, overload, and leakage). A "scenario-parameter-mode" association framework is established to lay the data foundation for multi-scenario adaptation and realize the layered design of the database.

[0010] A cyclic sampling model is constructed to traverse all combinations of power consumption parameters, generating a unique input set for monitoring tasks. This ensures comprehensive real-time data collection without blind spots and avoids feature omissions caused by fixed sampling. For example, in industrial scenarios, it can simultaneously cover the monitoring of cross-combinations of parameters such as voltage, current, and frequency.

[0011] By calculating the membership degree of real-time early warning feature values ​​to the power consumption status sample cluster using the maximum / minimum difference, the current power consumption mode can be accurately located. Compared with the "black and white" judgment of traditional thresholds, the membership degree can quantify the degree of matching (e.g., a membership degree of 0.8 indicates a high match to the "overload mode"), solving the adaptability problem of rigid thresholds;

[0012] The necessity of mode switching is determined based on the cumulative deviation between historical and real-time data. When the cumulative deviation exceeds a threshold, mode switching is automatically triggered (e.g., switching from "normal mode" to "anomaly tracking mode").

[0013] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0014] An energy meter with self-identification of abnormal power consumption patterns includes: a database configuration module, used to configure an electricity consumption parameter database and an electricity consumption status database in a memory, and to store an electricity consumption parameter sample cluster and an electricity consumption status sample cluster respectively;

[0015] The cyclic sampling and monitoring module is used to build a cyclic sampling model, generate a monitoring task input set and execute tests, and output a real-time early warning feature value set;

[0016] The membership analysis module is used to compare the real-time early warning feature value set with the power consumption status sample cluster, calculate the membership degree and determine the sample cluster with the highest matching degree, and establish switching link labels;

[0017] The mode switching evaluation module calculates the cumulative matching deviation value of different monitoring task input sets based on the switching link label, and determines whether to trigger a state switching warning based on the threshold.

[0018] Furthermore, the database configuration module includes:

[0019] The power consumption parameter configuration unit creates a power consumption parameter sample cluster based on the power consumption scenario and stores the power consumption parameters under the corresponding scenario.

[0020] The power consumption status configuration unit creates a power consumption status sample cluster based on the power consumption mode, and stores the early warning sample feature values ​​and associated early warning item indicators for the corresponding mode.

[0021] Furthermore, the cyclic sampling and monitoring module includes:

[0022] The sampling model construction unit is used to build a cyclic sampling model, which cyclically selects electricity parameters from the electricity parameter sample cluster to form the monitoring task input set, ensuring that the input set is not repeated each time.

[0023] The real-time monitoring unit is used to perform tests with the monitoring task input set, collect and output the real-time warning feature values ​​corresponding to each warning item indicator, and form a real-time warning feature value set.

[0024] Furthermore, the membership analysis module includes:

[0025] The feature comparison unit is used to calculate the difference between the real-time warning feature value and the warning sample feature value in the power consumption status sample cluster, and to determine the maximum and minimum difference.

[0026] The membership calculation unit calculates the membership degree of the real-time early warning feature value set to each power consumption status sample cluster based on the difference, and selects the power consumption status sample cluster with the largest membership degree.

[0027] The link tag creation unit is used to establish switching link tags between the power consumption status sample cluster and the real-time early warning feature value set and the monitoring task input set.

[0028] Furthermore, the mode switching evaluation module includes:

[0029] The deviation calculation unit calculates the cumulative value of the matching deviation between the benchmark monitoring task input set and the input set to be switched, based on the switching link label.

[0030] The threshold judgment unit is used to compare the cumulative deviation value with the preset threshold. If it is less than or equal to the threshold, the switching warning condition is determined to be met; otherwise, it is not met.

[0031] A method for intelligent switching of electricity meter modes with self-identification of abnormal electricity consumption patterns includes:

[0032] Step S1: Configure an electricity consumption parameter database and an electricity consumption status database in the memory integrated into the electricity meter. The electricity consumption parameter database stores electricity consumption parameters for different electricity consumption scenarios, and the electricity consumption status database stores early warning sample feature values ​​for different electricity consumption modes.

[0033] Step S2: Construct a cyclic sampling model, which forms a monitoring task input set by cyclically selecting samples of different electricity consumption parameters, tests the electricity meter and obtains a real-time early warning feature value set, until the sampling of all combinations of electricity consumption parameters is completed;

[0034] Step S3: Compare the real-time early warning feature value set with the power consumption status sample cluster in the power consumption status database, calculate the membership degree to determine the power consumption status sample cluster with the highest matching degree, and establish a switching link label between the power consumption status sample cluster and the real-time data.

[0035] Step S4: Based on the switching link label, evaluate the cumulative value of the matching deviation between different monitoring task input sets, and determine whether the switching warning conditions of the power consumption status sample cluster are met according to the preset threshold.

[0036] Furthermore, the specific implementation process of step S1 includes:

[0037] The electricity consumption parameter database contains several electricity consumption parameter sample clusters. The electricity consumption parameter sample clusters are configured based on electricity consumption scenarios and are used to store electricity consumption parameters. Each electricity consumption scenario corresponds to one electricity consumption parameter sample cluster. The electricity consumption parameters refer to the electrical energy parameters collected under the electricity consumption scenario.

[0038] The electricity consumption status database contains several electricity consumption status sample clusters. The electricity consumption status sample clusters are configured based on electricity consumption patterns and are used to store the characteristic values ​​of electricity consumption pattern warning samples. One electricity consumption pattern corresponds to one electricity consumption status sample cluster. The characteristic values ​​of the electricity consumption pattern warning samples are configured by initialization, and one characteristic value of the electricity consumption pattern warning samples corresponds to one warning item indicator.

[0039] Furthermore, the specific implementation process of step S2 includes:

[0040] Construct a cyclic sampling model, selecting one electricity consumption parameter from each electricity consumption parameter sample cluster to form the input set for the monitoring task in the m-th cyclic sampling, denoted as . Where I represents the total number of power consumption parameter sample clusters, This represents the x-th electricity parameter selected from the i-th electricity parameter sample cluster; the input set of the monitoring task for the (m+1)-th cyclic sampling is denoted as... ,and ;

[0041] One cycle of sampling is considered as one monitoring task, and the monitoring task input set is used. When testing an electricity meter, the meter outputs real-time characteristic values ​​of the electricity consumption mode corresponding to each warning item, forming a real-time warning characteristic value set, denoted as... ,and ,in, This represents the real-time characteristic value of the power consumption mode warning corresponding to the y-th warning item indicator, where Y represents the total number of warning items indicators;

[0042] like Then the cyclic sampling stops, where, This represents the i-th power consumption parameter sample cluster. This indicates the number of electricity parameters included in the electricity parameter sample cluster.

[0043] Furthermore, the specific implementation process of step S3 includes:

[0044] The real-time early warning feature set is compared with sample clusters of different power consumption states to analyze the real-time early warning feature set. The power consumption status sample cluster with the highest membership degree:

[0045] Let the j-th electricity consumption status sample cluster be denoted as ,and ,in, This represents the feature value of the electricity consumption pattern warning sample corresponding to the y-th warning item indicator;

[0046] Real-time early warning feature set For power consumption status sample clusters membership degree In the formula, Indicates the characteristic value of the power consumption pattern early warning sample. Real-time characteristic values ​​of power consumption pattern early warning The absolute value of the difference between them, max{} and min{} represent the maximum and minimum value functions, respectively;

[0047] By filtering function Return the power consumption state sample cluster with the highest membership degree. In the formula, J represents the total number of power consumption status sample clusters;

[0048] And in the returned power consumption status sample cluster With real-time early warning feature set Monitoring task input set The transition links are formed between the tags, denoted as .

[0049] Furthermore, the specific implementation process of step S4 includes:

[0050] Based on switching link tags, to monitor the task input set As the baseline task, the nth monitoring task input set For the matching task to be switched, evaluate the monitoring task input set. With monitoring task input set Cumulative value of matching deviation between ;

[0051] A preset cumulative deviation threshold is set; if the cumulative deviation value matches... If the value is less than or equal to the cumulative deviation threshold, then the monitoring task input set is determined. With monitoring task input set If the switching warning for the power consumption status sample cluster is met, the monitoring task input set is determined to be invalid. With monitoring task input set Early warning for switching between sample clusters that do not meet the power consumption status requirements.

[0052] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: The energy meter with self-identification of abnormal electricity consumption patterns and its intelligent mode switching method provided by this invention integrates a database configuration module, a cyclic sampling and monitoring module, a membership analysis module, and a mode switching evaluation module within the energy meter. The application method of the energy meter is as follows: Configure an electricity consumption parameter database and an electricity consumption status database to store electricity consumption parameter sample clusters classified by scenario and warning sample feature values ​​classified by mode, respectively; generate a monitoring task input set by traversing all combinations of electricity consumption parameters through a cyclic sampling model to obtain a real-time warning feature value set; calculate the membership degree based on the maximum / minimum difference, match the optimal electricity consumption status sample cluster, and generate a switching link label; evaluate the mode switching conditions through the cumulative deviation value and trigger the warning. This invention solves the problems of poor adaptability and delayed response caused by the reliance on fixed thresholds for mode switching in traditional energy meters, and achieves accurate identification and proactive switching of abnormal modes in multiple scenarios. Attached Figure Description

[0053] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0054] Figure 1 This is a schematic diagram illustrating the steps of an intelligent switching method for electricity meter modes that self-identifies abnormal electricity usage patterns according to the present invention. Detailed Implementation

[0055] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] In this first embodiment: a power meter with self-identification of abnormal power consumption patterns is provided, including: a database configuration module, used to configure a power consumption parameter database and a power consumption status database in a memory, and to store power consumption parameter sample clusters and power consumption status sample clusters respectively;

[0057] The database configuration module includes:

[0058] The power consumption parameter configuration unit creates a power consumption parameter sample cluster based on the power consumption scenario and stores the power consumption parameters under the corresponding scenario.

[0059] The power consumption status configuration unit creates a power consumption status sample cluster based on the power consumption mode, and stores the early warning sample feature values ​​and associated early warning item indicators for the corresponding mode.

[0060] The cyclic sampling and monitoring module is used to build a cyclic sampling model, generate a monitoring task input set and execute tests, and output a real-time early warning feature value set;

[0061] The cyclic sampling and monitoring module includes:

[0062] The sampling model construction unit is used to build a cyclic sampling model, which cyclically selects electricity parameters from the electricity parameter sample cluster to form the monitoring task input set, ensuring that the input set is not repeated each time.

[0063] The real-time monitoring unit is used to perform tests with the monitoring task input set, collect and output the real-time warning feature values ​​corresponding to each warning item indicator, and form a real-time warning feature value set.

[0064] The membership analysis module is used to compare the real-time early warning feature value set with the power consumption status sample cluster, calculate the membership degree and determine the sample cluster with the highest matching degree, and establish switching link labels;

[0065] The membership analysis module includes:

[0066] The feature comparison unit is used to calculate the difference between the real-time warning feature value and the warning sample feature value in the power consumption status sample cluster, and to determine the maximum and minimum difference.

[0067] The membership calculation unit calculates the membership degree of the real-time early warning feature value set to each power consumption status sample cluster based on the difference, and selects the power consumption status sample cluster with the largest membership degree.

[0068] The link tag creation unit is used to establish switching link tags between the power consumption status sample cluster and the real-time early warning feature value set and the monitoring task input set;

[0069] The mode switching evaluation module calculates the cumulative matching deviation value of different monitoring task input sets based on the switching link label, and determines whether to trigger a state switching warning based on the threshold.

[0070] The mode switching evaluation module includes:

[0071] The deviation calculation unit calculates the cumulative value of the matching deviation between the benchmark monitoring task input set and the input set to be switched, based on the switching link label.

[0072] The threshold judgment unit is used to compare the cumulative deviation value with the preset threshold. If it is less than or equal to the threshold, the switching warning condition is determined to be met; otherwise, it is not met.

[0073] Please see Figure 1In this second embodiment: a smart switching method for electricity meter modes with self-identification of abnormal electricity consumption patterns is provided for use in the first embodiment above, including:

[0074] Step S1: Configure an electricity consumption parameter database and an electricity consumption status database in the memory integrated into the electricity meter. The electricity consumption parameter database stores electricity consumption parameters for different electricity consumption scenarios, and the electricity consumption status database stores early warning sample feature values ​​for different electricity consumption modes.

[0075] For example, the electricity consumption parameter database contains several electricity consumption parameter sample clusters. The electricity consumption parameter sample clusters are configured based on electricity consumption scenarios and are used to store electricity consumption parameters. One electricity consumption scenario corresponds to one electricity consumption parameter sample cluster. The electricity consumption parameters refer to the electrical energy parameters collected under the electricity consumption scenario.

[0076] The power consumption status database contains several power consumption status sample clusters. The power consumption status sample clusters are configured based on power consumption patterns and are used to store power consumption pattern warning sample feature values. One power consumption pattern corresponds to one power consumption status sample cluster. The power consumption pattern warning sample feature values ​​are configured by initialization, and one power consumption pattern warning sample feature value corresponds to one warning item indicator.

[0077] For example, taking the monitoring of electricity consumption in a residential area with multiple households as an example, the area includes 500 households and electrical equipment such as air conditioners, refrigerators, and electric vehicle charging stations. It is necessary to solve the problem of false alarms caused by load fluctuations during peak and off-peak periods, while accurately identifying real anomalies such as leakage and overload.

[0078] The electricity consumption parameter database is configured with sample clusters according to three scenarios: "daily residential use", "peak residential use (18:00-22:00)" and "charging pile activation".

[0079] Resident daily sample cluster (EP1): Voltage 220±5V, Current 0-10A, Power 0-2200W;

[0080] Residential peak hour sample cluster (EP2): Voltage 220±8V, Current 0-20A, Power 0-4400W;

[0081] Charging pile activation sample cluster (EP3): Voltage 220±10V, Current 5-30A, Power 1100-6600W;

[0082] The electricity consumption status database is configured with sample clusters according to three modes: "normal", "overload", and "leakage".

[0083] Normal mode (ES1): The warning characteristics are power fluctuation ≤5% and leakage current ≤10mA;

[0084] Overload mode (ES2): Warning characteristics are power fluctuation > 10% and duration > 30s;

[0085] Leakage mode (ES3): The warning characteristic value is leakage current > 30mA.

[0086] Step S2: Construct a cyclic sampling model, which forms a monitoring task input set by cyclically selecting samples of different electricity consumption parameters, tests the electricity meter and obtains a real-time early warning feature value set, until the sampling of all combinations of electricity consumption parameters is completed;

[0087] For example, a cyclic sampling model is constructed, in which one electricity consumption parameter is selected from each electricity consumption parameter sample cluster to form the monitoring task input set for the m-th cyclic sampling, denoted as . Where I represents the total number of power consumption parameter sample clusters, This represents the x-th electricity parameter selected from the i-th electricity parameter sample cluster; the input set of the monitoring task for the (m+1)-th cyclic sampling is denoted as... ,and ;

[0088] One cycle of sampling is considered as one monitoring task, and the monitoring task input set is used. When testing an electricity meter, the meter outputs real-time characteristic values ​​of the electricity consumption mode corresponding to each warning item, forming a real-time warning characteristic value set, denoted as... ,and ,in, This represents the real-time characteristic value of the power consumption mode warning corresponding to the y-th warning item indicator, where Y represents the total number of warning items indicators;

[0089] like Then the cyclic sampling stops, where, This represents the i-th power consumption parameter sample cluster. This indicates the number of electricity parameters included in the electricity parameter sample cluster;

[0090] For example, the parameters of the cyclic sampling model are: I=3 (3 power consumption parameter sample clusters), |EP1|=3, |EP2|=3, |EP3|=3, and the total number of cycles is m=3×3×3=27.

[0091] The input set for the first monitoring task is IN1={P} 11 (Voltage 220V), P 21 (Current 5A), P 31 (Power 1100W)}, output real-time warning feature set L1={F1 (power fluctuation 2%), F2 (leakage current 5mA)};

[0092] The input set for the second monitoring task is IN2={P} 12(Voltage 215V), P 22 (Current 8A), P 32 (Power 1720W)}, ensure that it does not overlap with IN1, output L2={F1 (power fluctuation 3%), F2 (leakage current 8mA)}.

[0093] Step S3: Compare the real-time early warning feature value set with the power consumption status sample cluster in the power consumption status database, calculate the membership degree to determine the power consumption status sample cluster with the highest matching degree, and establish a switching link label between the power consumption status sample cluster and the real-time data.

[0094] For example, the real-time early warning feature set is compared with sample clusters of different power consumption states to analyze the real-time early warning feature set. The power consumption status sample cluster with the highest membership degree:

[0095] Let the j-th electricity consumption status sample cluster be denoted as ,and ,in, This represents the feature value of the electricity consumption pattern warning sample corresponding to the y-th warning item indicator;

[0096] Real-time early warning feature set For power consumption status sample clusters membership degree In the formula, Indicates the characteristic value of the power consumption pattern early warning sample. Real-time characteristic values ​​of power consumption pattern early warning The absolute value of the difference between them, max{} and min{} represent the maximum and minimum value functions, respectively;

[0097] By filtering function Return the power consumption state sample cluster with the highest membership degree. In the formula, J represents the total number of power consumption status sample clusters;

[0098] And in the returned power consumption status sample cluster With real-time early warning feature set Monitoring task input set The transition links are formed between the tags, denoted as .

[0099] Step S4: Based on the switching link label, evaluate the cumulative value of the matching deviation between different monitoring task input sets, and determine whether the switching warning condition of the power consumption status sample cluster is met according to the preset threshold.

[0100] For example, monitoring task input sets is based on switching link labels. As the baseline task, the nth monitoring task input set For the matching task to be switched, evaluate the monitoring task input set. With monitoring task input set Cumulative value of matching deviation between ;

[0101] A preset cumulative deviation threshold is set; if the cumulative deviation value matches... If the value is less than or equal to the cumulative deviation threshold, then the monitoring task input set is determined. With monitoring task input set If the switching warning for the power consumption status sample cluster is met, the monitoring task input set is determined to be invalid. With monitoring task input set Early warning for switching between sample clusters that do not meet the power consumption status requirements.

[0102] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0103] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent switching of electricity meter modes based on self-identification of abnormal electricity consumption patterns, characterized in that, The method includes the following steps: Step S1: Configure an electricity consumption parameter database and an electricity consumption status database in the memory integrated into the electricity meter. The electricity consumption parameter database stores electricity consumption parameters for different electricity consumption scenarios, and the electricity consumption status database stores early warning sample feature values ​​for different electricity consumption modes. Step S2: Construct a cyclic sampling model, which forms a monitoring task input set by cyclically selecting samples of different electricity consumption parameters, tests the electricity meter and obtains a real-time early warning feature value set, until the sampling of all combinations of electricity consumption parameters is completed; Step S3: Compare the real-time early warning feature value set with the power consumption status sample cluster in the power consumption status database, calculate the membership degree to determine the power consumption status sample cluster with the highest matching degree, and establish a switching link label between the power consumption status sample cluster and the real-time data. Step S4: Based on the switching link label, evaluate the cumulative value of the matching deviation between different monitoring task input sets, and determine whether the switching warning condition of the power consumption status sample cluster is met according to the preset threshold. The specific implementation process of step S1 includes: The electricity consumption parameter database contains several electricity consumption parameter sample clusters. The electricity consumption parameter sample clusters are configured based on electricity consumption scenarios and are used to store electricity consumption parameters. Each electricity consumption scenario corresponds to one electricity consumption parameter sample cluster. The electricity consumption parameters refer to the electrical energy parameters collected under the electricity consumption scenario. The electricity consumption status database contains several electricity consumption status sample clusters. The electricity consumption status sample clusters are configured based on electricity consumption patterns and are used to store the characteristic values ​​of electricity consumption pattern warning samples. One electricity consumption pattern corresponds to one electricity consumption status sample cluster. The characteristic values ​​of the electricity consumption pattern warning samples are configured by initialization, and one characteristic value of the electricity consumption pattern warning samples corresponds to one warning item indicator.

2. The method for intelligent switching of electricity meter modes based on self-identification of abnormal electricity consumption modes according to claim 1, characterized in that, The specific implementation process of step S2 includes: Construct a cyclic sampling model, selecting one electricity consumption parameter from each electricity consumption parameter sample cluster to form the input set for the monitoring task in the m-th cyclic sampling, denoted as . Where I represents the total number of power consumption parameter sample clusters, This represents the x-th electricity parameter selected from the i-th electricity parameter sample cluster; the input set of the monitoring task for the (m+1)-th cyclic sampling is denoted as... ,and ; One cycle of sampling is considered as one monitoring task, and the monitoring task input set is used. When testing an electricity meter, the meter outputs real-time characteristic values ​​of the electricity consumption mode corresponding to each warning item, forming a real-time warning characteristic value set, denoted as... ,and ,in, This represents the real-time characteristic value of the power consumption mode warning corresponding to the y-th warning item indicator, where Y represents the total number of warning items indicators; like Then the cyclic sampling stops, where, This represents the i-th power consumption parameter sample cluster. This indicates the number of electricity parameters included in the electricity parameter sample cluster.

3. The method for intelligent switching of electricity meter modes based on self-identification of abnormal electricity consumption modes according to claim 2, characterized in that, The specific implementation process of step S3 includes: The real-time early warning feature set is compared with sample clusters of different power consumption states to analyze the real-time early warning feature set. The power consumption status sample cluster with the highest membership degree: Let the j-th electricity consumption status sample cluster be denoted as ,and ,in, This represents the feature value of the electricity consumption pattern warning sample corresponding to the y-th warning item indicator; Real-time early warning feature set For power consumption status sample clusters membership degree In the formula, Indicates the characteristic value of the power consumption pattern early warning sample. Real-time characteristic values ​​of power consumption pattern early warning The absolute value of the difference between them, max{} and min{} represent the maximum and minimum value functions, respectively; By filtering function Return the power consumption state sample cluster with the highest membership degree. In the formula, J represents the total number of power consumption status sample clusters; And in the returned power consumption status sample cluster With real-time early warning feature set Monitoring task input set The transition link is formed between the two elements, denoted as . .

4. The method for intelligent switching of electricity meter modes based on self-identification of abnormal electricity consumption modes according to claim 3, characterized in that, The specific implementation process of step S4 includes: Based on switching link tags, to monitor the task input set As the baseline task, the nth monitoring task input set For the matching task to be switched, evaluate the monitoring task input set. With monitoring task input set Cumulative value of matching deviation between ; A preset cumulative deviation threshold is set; if the cumulative deviation value matches... If the value is less than or equal to the cumulative deviation threshold, then the monitoring task input set is determined. With monitoring task input set If the switching warning for the power consumption status sample cluster is met, the monitoring task input set is determined to be invalid. With monitoring task input set Early warning for switching between sample clusters that do not meet the power consumption status requirements.

5. An energy meter with self-identification of abnormal electricity consumption patterns, executing the intelligent switching method for energy meter modes as described in any one of claims 1-4, characterized in that, The electricity meter includes: The database configuration module is used to configure the power consumption parameter database and the power consumption status database in the memory, and to store the power consumption parameter sample cluster and the power consumption status sample cluster respectively. The cyclic sampling and monitoring module is used to build a cyclic sampling model, generate a monitoring task input set and execute tests, and output a real-time early warning feature value set. The membership analysis module is used to compare the real-time early warning feature value set with the power consumption status sample cluster, calculate the membership degree and determine the sample cluster with the highest matching degree, and establish switching link labels; The mode switching evaluation module calculates the cumulative matching deviation value of different monitoring task input sets based on the switching link label, and determines whether to trigger a state switching warning based on the threshold.

6. An energy meter with self-identification of abnormal power consumption patterns according to claim 5, characterized in that, The database configuration module includes: The power consumption parameter configuration unit creates a power consumption parameter sample cluster based on the power consumption scenario and stores the power consumption parameters under the corresponding scenario. The power consumption status configuration unit creates a power consumption status sample cluster based on the power consumption mode, and stores the early warning sample feature values ​​and associated early warning item indicators for the corresponding mode.

7. An energy meter with self-identification of abnormal power consumption patterns according to claim 5, characterized in that, The cyclic sampling and monitoring module includes: The sampling model construction unit is used to build a cyclic sampling model, which cyclically selects electricity parameters from the electricity parameter sample cluster to form the monitoring task input set, ensuring that the input set is not repeated each time. The real-time monitoring unit is used to perform tests with the monitoring task input set, collect and output the real-time warning feature values ​​corresponding to each warning item indicator, and form a real-time warning feature value set.

8. An energy meter with self-identification of abnormal power consumption patterns according to claim 5, characterized in that, The membership analysis module includes: The feature comparison unit is used to calculate the difference between the real-time warning feature value and the warning sample feature value in the power consumption status sample cluster, and to determine the maximum and minimum difference. The membership calculation unit calculates the membership degree of the real-time early warning feature value set to each power consumption status sample cluster based on the difference, and selects the power consumption status sample cluster with the largest membership degree. The link tag creation unit is used to establish switching link tags between the power consumption status sample cluster and the real-time early warning feature value set and the monitoring task input set.

9. An energy meter with self-identification of abnormal power consumption patterns according to claim 5, characterized in that, The mode switching evaluation module includes: The deviation calculation unit calculates the cumulative value of the matching deviation between the benchmark monitoring task input set and the input set to be switched, based on the switching link label. The threshold judgment unit is used to compare the cumulative deviation value with the preset threshold. If it is less than or equal to the threshold, the switching warning condition is determined to be met; otherwise, it is not met.

Images