Power meter over-capacity burning prediction method based on multi-dimensional weighted clustering algorithm

Abstract

The invention discloses a multi-dimensional weighted clustering algorithm-based overcapacity burning prediction method for an electric energy meter. The method comprises the steps of S1, acquiring meter information through an information table of the electric energy meter; s2, acquiring load data of a user in historical cycle time through an electrical load data table; and S3, for the load data stored in the database, taking users with a preset number of continuous points and above in the day as over-capacity users, and performing multi-dimensional weighted clustering calculation on the over-capacity users. The invention discloses a method for predicting over-capacity burnout of an electric power meter based on a multi-dimensional weighted clustering algorithm. Quantized over-capacity burnout risk evaluation is given by considering over-capacity multiple, continuity and heating conditions of the electric power meter and combining weather and power utilization changes; the basic power supply station can change the meter in a targeted manner according to the output data or the received short message, and can change a proper meter according to the actual demand, thereby preventing the meter from being burnt, and improving the utilization rate of equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of prediction of overcapacity and burnout of electric meters, and specifically relates to a method for predicting overcapacity and burnout of electric meters based on a multi-dimensional weighted clustering algorithm of an electric power data center. Background Art

[0002] With the improvement of living standards, the addition of new electrical appliances by residential users has led to an increase in electricity consumption. A large number of electricity meters are over-capacity. Long-term operation can easily cause the electricity meters and accessory equipment to burn out, which not only affects electricity consumption, but can even cause fires in meter boxes. The existing conventional over-capacity meter analysis generally extracts over-capacity data, which requires comparison across multiple systems such as marketing and collection. The amount of data and the workload of comparison are too large. Due to the lack of operation and maintenance manpower, it is not feasible to simply replace all over-capacity meters. The capacity of the replaced meters often depends on experience, and there is a situation where the capacity of the replaced meters does not match the user; on the other hand, during the peak electricity consumption period in winter and summer, the probability of over-capacity meters burning out increases significantly. There is an urgent need for a data-based positioning method for easily burned meters to guide grassroots power supply stations to carry out targeted meter replacement and reduce the occurrence of meter burning incidents. Summary of the invention

[0003] The main purpose of the present invention is to provide an electric meter overcapacity and burnout prediction method based on a multi-dimensional weighted clustering algorithm, which considers the overcapacity multiples, continuity, and heating conditions of the electric meter in combination with weather and electricity consumption changes to give a quantitative overcapacity and burnout risk assessment; the grassroots power supply station can replace the meter in a targeted manner according to the output data or received text messages, and can replace the appropriate meter according to actual needs to prevent the occurrence of meter burnout and improve equipment utilization; the grassroots power supply station is guided by the intuitive meter burnout risk value to give priority to replacing the meters that are easy to burn out.

[0004] To achieve the above objectives, the present invention provides a method for predicting overcapacity and burnout of electric power meters based on a multi-dimensional weighted clustering algorithm, comprising the following methods:

[0005] Step S1: Obtain meter information through the information table (dwd_cst_meter_run) of the electric energy meter, where the meter information includes rated voltage, rated current, rated capacity (Srated), wiring method and installation date;

[0006] Step S2: Obtain the load data of the user's historical period (96 points every day in the past year) through the power load data table (dwd_cst_t_itf_comp_curve). The load data includes voltage, current and apparent power (S). The load data with apparent power greater than the rated capacity is stored in the database (each phase of the three-phase user is calculated separately);

[0007] Step S3: For the stored load data, users with a preset number (preferably 4) of consecutive points (points where the apparent power is greater than the rated capacity) or more in a day are considered as overcapacity users (three-phase users reflect the phase difference of overcapacity), and multi-dimensional weighted clustering calculation is performed on the overcapacity users, which is specifically implemented as the following steps:

[0008] Step S3.1: Calculate the overcapacity ratio, where the overcapacity ratio = average value of the maximum five apparent powers / rated capacity; overcapacity ratio weight R1 = (overcapacity ratio - 1) * 5;

[0009] Step S3.2: Calculate the continuous overcapacity. The weight of the historical continuous overcapacity is R2 = the maximum number of continuous overcapacity measurement points in the past year / 8; the weight of the recent continuous overcapacity is R3 = the maximum number of continuous overcapacity measurement points in the past week / 4;

[0010] Step S3.3: Calculate the daily over-limit rate. The weight of the historical daily over-limit rate R4 = the total over-limit test points of the three largest over-limit days in the past year / 72; the weight of the recent daily over-limit rate R5 = the maximum number of daily over-limit test points in the past week / 12;

[0011] Step S3.4: Calculate the weekly over-limit rate, and the weekly over-limit weight R6 = the number of over-limit days in the past week / 7*4;

[0012] Step S3.5: Calculate the overheat index to reflect the single overheating that exceeds a certain ratio.

[0013] Single overheat index = the integral value corresponding to the overcapacity ratio of the overcapacity continuous point greater than a preset multiple (preferably 1.12 times) minus the threshold ratio;

[0014]

[0015] In the formula, R7 is the overheating index; It is the overcapacity ratio; Imax is the threshold ratio;

[0016] Overheat index weight = the largest single overheat index in the past week * the first value (the first value is preferably 2);

[0017] Step S3.6: Calculate the aging index, aging base = (current date - installation date) / 365, aging index weight R8 = aging base * historical daily over-capacity rate;

[0018] Step S3.7: Perform load surge processing, obtain historical weather information of the data center (taking into account the impact of air conditioning load), and use load conversion coefficient = user's average summer and winter load / spring and autumn average load for judgment. If the user's spring and autumn average surge (average load increases by more than 30% year-on-year) and the current spring and autumn average load * load conversion coefficient is greater than the rated capacity, then during the next season change period, increase the season change weight R9 and prompt the risk of over-capacity burnout during the season change, R9 = (current spring and autumn average load * load conversion coefficient - rated capacity) / rated capacity * 20;

[0019] Step S3.8: Perform special case processing and obtain weather forecast information from the data center. If the maximum temperature of the next day is above the preset temperature (30 degrees Celsius), increase the overheating index risk value R7 ’ = R7*(maximum temperature - 29) to increase the priority of replacing the overheating meter;

[0020] Step S3.9: Update the above dimension data (R1-R8) daily and perform weighted summation;

[0021]

[0022] By sorting the total R of each power meter in reverse order, the over-capacity meters with risk values ​​greater than the total threshold or at the front are the ones that need to be replaced first.

[0023] As a further preferred technical solution of the above technical solution, step S3.9 further includes:

[0024] Step S3.10: Calculate the minimum capacity of the meter recommended for replacement, minimum capacity = average value of the apparent power of the 10 largest collection points throughout the year * 1.2;

[0025] If there is a sudden load increase in step S7, a year-on-year conversion prediction process is performed.

[0026] As a further preferred technical solution of the above technical solution, the user information table (dim_cst_elec_cons_cust) is associated with the substation information table (dim_cst_connection) and the electric energy meter information table (dwd_cst_meter_run), and a list of high-risk overcapacity meters is collected by substation. Through the information of the meter reader (substation maintainer) in the substation information table, a text message is sent to notify the meter in time to replace the meter, and the minimum capacity information is provided for reference to select the appropriate meter.

[0027] To achieve the above objectives, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for predicting overcapacity and burnout of an electric meter based on a multi-dimensional weighted clustering algorithm are implemented.

[0028] To achieve the above objectives, the present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for predicting overcapacity and burnout of electric meters based on a multi-dimensional weighted clustering algorithm are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0030] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles of the present invention defined in the following description can be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not deviate from the spirit and scope of the present invention.

[0031] In the preferred embodiments of the present invention, those skilled in the art should note that the electric power meter and the like involved in the present invention may be regarded as prior art.

[0032] Preferred embodiments.

[0033] The present invention discloses a method for predicting overcapacity and burnout of electric power meters based on a multi-dimensional weighted clustering algorithm, comprising the following methods:

[0034] Step S1: Obtain meter information through the information table (dwd_cst_meter_run) of the electric energy meter, where the meter information includes rated voltage, rated current, rated capacity (Srated), wiring method and installation date;

[0035] Step S2: Obtain the load data of the user's historical period (96 points every day in the past year) through the power load data table (dwd_cst_t_itf_comp_curve). The load data includes voltage, current and apparent power (S). The load data with apparent power greater than the rated capacity is stored in the database (each phase of the three-phase user is calculated separately);

[0036] Step S3: For the stored load data, users with a preset number (preferably 4) of consecutive points (points where the apparent power is greater than the rated capacity) or more in a day are considered as overcapacity users (three-phase users reflect the phase difference of overcapacity), and multi-dimensional weighted clustering calculation is performed on the overcapacity users, which is specifically implemented as the following steps:

[0037] Step S3.1: Calculate the overcapacity ratio, where the overcapacity ratio = average value of the maximum five apparent powers / rated capacity; overcapacity ratio weight R1 = (overcapacity ratio - 1) * 5;

[0038] Step S3.2: Calculate the continuous overcapacity. The weight of the historical continuous overcapacity is R2 = the maximum number of continuous overcapacity measurement points in the past year / 8; the weight of the recent continuous overcapacity is R3 = the maximum number of continuous overcapacity measurement points in the past week / 4;

[0039] Step S3.3: Calculate the daily over-limit rate. The weight of the historical daily over-limit rate R4 = the total over-limit test points of the three largest over-limit days in the past year / 72; the weight of the recent daily over-limit rate R5 = the maximum number of daily over-limit test points in the past week / 12;

[0040] Step S3.4: Calculate the weekly over-limit rate, and the weekly over-limit weight R6 = the number of over-limit days in the past week / 7*4;

[0041] Step S3.5: Calculate the overheat index to reflect the single overheating that exceeds a certain ratio.

[0042] Single overheat index = the integral value corresponding to the overcapacity ratio of the overcapacity continuous point greater than a preset multiple (preferably 1.12 times) minus the threshold ratio;

[0043]

[0044] In the formula, R7 is the overheating index; It is the overcapacity ratio; Imax is the threshold ratio;

[0045] Overheat index weight = the largest single overheat index in the past week * the first value (the first value is preferably 2);

[0046] Step S3.6: Calculate the aging index, aging base = (current date - installation date) / 365, aging index weight R8 = aging base * historical daily over-capacity rate;

[0047] Step S3.7: Perform load surge processing, obtain historical weather information from the data center (taking into account the impact of air conditioning load), and use the load conversion coefficient = user's average summer and winter load / spring and autumn average load for judgment. If the user's spring and autumn average surges (average load increases by more than 30% year-on-year) and the current spring and autumn average load * load conversion coefficient is greater than the rated capacity, then during the next season change period, increase the season change weight R9 and prompt the risk of over-capacity burnout during the season change, R9 = (current spring and autumn average load * load conversion coefficient - rated capacity) / rated capacity * 20;

[0048] Step S3.8: Perform special case processing and obtain weather forecast information from the data center. If the maximum temperature of the next day is above the preset temperature (30 degrees Celsius), increase the overheating index risk value R7 ’ = R7*(maximum temperature - 29) to increase the priority of replacing the overheating meter;

[0049] Step S3.9: Update the above dimension data (R1-R9) daily and perform weighted summation;

[0050]

[0051] By sorting the total R of each power meter in reverse order, the over-capacity meters with risk values ​​greater than the total threshold or at the front are the ones that need to be replaced first.

[0052] Specifically, after step S3.9, the following steps are also included:

[0053] Step S3.10: Calculate the minimum capacity of the meter recommended for replacement, minimum capacity = average value of the apparent power of the 10 largest collection points throughout the year * 1.2;

[0054] If there is a sudden load increase in step S7, a year-on-year conversion prediction process is performed.

[0055] If it is a three-phase meter, it is the maximum value of the recommended minimum capacity among the three phases ABC (max(recommended minimum capacity A, recommended minimum capacity B, recommended minimum capacity C)).

[0056] More specifically, the user information table (dim_cst_elec_cons_cust) is associated with the substation information table (dim_cst_connection) and the electricity meter information table (dwd_cst_meter_run), and a list of high-risk overcapacity meters is collected by substation. Through the meter reader (substation maintainer) information in the substation information table, a text message notification is sent to promptly replace the meter, and the minimum capacity information is provided for reference to select the appropriate meter.

[0057] It is worth mentioning that the various values ​​in the present invention can be changed according to actual needs, rather than being fixed.

[0058] The present invention also discloses an electronic device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the method for predicting overcapacity and burnout of electric power meters based on a multi-dimensional weighted clustering algorithm are implemented.

[0059] The present invention also discloses a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for predicting overcapacity and burnout of electric power meters based on a multi-dimensional weighted clustering algorithm are implemented.

[0060] It is worth mentioning that the technical features such as the electric power meter involved in the patent application of this invention should be regarded as the prior art. The specific structure, working principle and possible control method and spatial layout method of these technical features can be selected by conventional methods in the field, and should not be regarded as the inventive point of the patent of this invention. The patent of this invention will not be further elaborated.

[0061] For those skilled in the art, it is still possible to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting overcapacity and burnout of electric power meters based on a multi-dimensional weighted clustering algorithm, characterized in that: The following methods are included: Step S1: obtaining meter information through the information table of the electric energy meter, the meter information including rated voltage, rated current, rated capacity, wiring method and installation date; Step S2: Obtain the user's load data of historical period time through the power load data table, the load data includes voltage, current and apparent power, and store the load data with apparent power greater than the rated capacity; Step S3: For the stored load data, users with a preset number of consecutive points or more in a day are considered as over-capacity users, and multi-dimensional weighted clustering calculation is performed on the over-capacity users, which is specifically implemented as follows: Step S3.1: Calculate the overcapacity ratio, where the overcapacity ratio = average value of the maximum five apparent powers / rated capacity; overcapacity ratio weight R1 = (overcapacity ratio - 1) * 5; Step S3.2: Calculate the continuous overcapacity. The weight of the historical continuous overcapacity is R2 = the maximum number of continuous overcapacity measurement points in the past year / 8; the weight of the recent continuous overcapacity is R3 = the maximum number of continuous overcapacity measurement points in the past week / 4; Step S3.3: Calculate the daily over-limit rate. The weight of the historical daily over-limit rate R4 = the total over-limit test points of the three largest over-limit days in the past year / 72; the weight of the recent daily over-limit rate R5 = the maximum number of daily over-limit test points in the past week / 12; Step S3.4: Calculate the weekly over-limit rate, and the weekly over-limit weight R6 = the number of over-limit days in the past week / 7*4; Step S3.5: Calculate the overheat index to reflect the single overheating that exceeds a certain ratio. Single overheat index = the integral value corresponding to the overcapacity ratio of the overcapacity continuous point greater than the preset multiple minus the threshold ratio; In the formula, R7 is the overheating index; It is the overcapacity ratio; Imax is the threshold ratio; Overheating index weight = the largest single overheating index in the past week * the first value; Step S3.6: Calculate the aging index, aging base = (current date - installation date) / 365, aging index weight R8 = aging base * historical daily over-capacity rate; Step S3.7: Perform load surge processing, obtain historical weather information from the data center, and use the load conversion coefficient = user's average load in summer and winter / average load in spring and autumn to make a judgment. If the user's average load in spring and autumn surges and the current average load in spring and autumn * load conversion coefficient is greater than the rated capacity, then increase the seasonal weight R9 during the next seasonal change period and prompt the risk of seasonal over-capacity burnout, R9 = (current average load in spring and autumn * load conversion coefficient - rated capacity) / rated capacity * 20; Step S3.8: Perform special case processing and obtain weather forecast information from the data center. If the maximum temperature of the next day is above the preset temperature, increase the overheating index risk value R7 ’ =R7*(maximum temperature - 29), in order to increase the priority of replacing the overheating meter; Step S3.9: Update the above dimension data daily and perform weighted summation; By sorting the total R of each power meter in reverse order, the over-capacity meters with risk values ​​greater than the total threshold or at the front are the ones that need to be replaced first.

2. The method for predicting overcapacity and burnout of electric power meters based on a multi-dimensional weighted clustering algorithm according to claim 1 is characterized in that: Step S3.9 and later also include: Step S3.10: Calculate the minimum capacity of the meter recommended for replacement, minimum capacity = average value of the apparent power of the 10 largest collection points throughout the year * 1.2; If there is a sudden load increase in step S7, a year-on-year conversion prediction process is performed.

3. The method for predicting overcapacity and burnout of electric power meters based on a multi-dimensional weighted clustering algorithm according to claim 2 is characterized in that: The user information table is associated with the substation information table and the electric energy meter information table, and a list of high-risk over-capacity meters is compiled by substation. Through the meter reader information in the substation information table, SMS notifications are sent to ensure timely meter replacement, and minimum capacity information is provided for reference in selecting appropriate meters.

4. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method for predicting overcapacity and burnout of an electric power meter based on a multi-dimensional weighted clustering algorithm as described in any one of claims 1 to 3 are implemented.

5. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for predicting overcapacity and burnout of an electric power meter based on a multi-dimensional weighted clustering algorithm as described in any one of claims 1 to 3 are implemented.

Images