A transformer overload monitoring method and device based on an RPA robot

Abstract

The application discloses a transformer overload monitoring method and device based on an RPA robot, and is applied to the RPA robot, and the method comprises the following steps: obtaining historical user load current values corresponding to all user terminals in a jurisdiction and classifying the historical user load current values, generating multiple groups of load data and respectively performing normal distribution testing to generate multiple normal groups; searching the normal groups, determining associated normal distributions of each transformer and superimposing the associated normal distributions to generate a normal distribution model of each transformer; when receiving a user load current value of the current month, selecting an abnormal transformer according to a discrimination result of the user load current value of the current month in the associated normal distribution model; if a load index of the abnormal transformer satisfies a preset overload condition, generating a rectification strategy corresponding to the preset overload condition and outputting the rectification strategy; the method can apply the RPA robot to replace manual work to perform tedious information analysis, information processing and system operation, efficiently and scientifically monitors the transformer, and provides a rectification scheme for transformer overload operation.

Description

Technical Field

[0001] This invention relates to the field of transformer overload monitoring technology, and in particular to a transformer overload monitoring method and device based on an RPA robot. Background Technology

[0002] Distribution transformers are the final link in the power grid's transmission of electricity to residents, and their healthy operation determines the quality of electricity supply. Currently, with the continuous increase in electricity demand, the original capacity planning of transformers cannot meet the demand, leading to transformers operating under overload. If transformers operate under overload for extended periods, their output will decrease, losses will increase, and transformer equipment may be damaged, which is detrimental to the safe operation of the transformers.

[0003] Current technology typically involves technicians analyzing peak load periods each month, recording abnormal transformers that are overloaded or experiencing significant load fluctuations, and then combining data from multiple systems to determine the cause of the transformer overload and implement corrective measures. However, this method requires a significant amount of manpower and involves a large volume of data, making it difficult to efficiently and scientifically monitor transformers and generate corrective plans for transformer overload operation. Summary of the Invention

[0004] This invention provides a transformer overload monitoring method and device based on RPA robot, which solves the technical problems in the prior art that transformer overload monitoring requires a lot of manual labor, involves a large amount of data, and is difficult to monitor transformers efficiently and scientifically and generate rectification plans for transformer overload operation.

[0005] This invention provides a transformer overload monitoring method based on an RPA robot, applied to an RPA robot, where each transformer is associated with multiple user terminals. The method includes:

[0006] Obtain and classify the historical user load current values ​​corresponding to all user terminals within the jurisdiction, and generate multiple sets of load data;

[0007] Each group of load data is subjected to a normality test to generate multiple groups of normal groups.

[0008] Retrieve the normal groupings in each group, determine the normal distribution associated with each transformer, and superimpose them to generate a normal distribution model for each transformer;

[0009] When the monthly user load current values ​​corresponding to all the transformers are received, the abnormal transformer is selected according to the anomaly detection result of the monthly user load current values ​​in the associated normal distribution model.

[0010] If the load index of the abnormal transformer meets the preset overload condition, then the rectification strategy corresponding to the preset overload condition is generated and output.

[0011] Optionally, the step of performing a normality test on each group of load data to generate multiple normally distributed groups includes:

[0012] The historical user load current values ​​in each group of load data are sorted from smallest to largest, and the values ​​are taken sequentially and added to the first group for normality test, generating the test results.

[0013] If the test result does not meet the preset normal distribution result, the tail value of the first group is removed, and a normal group is generated;

[0014] Select the tail value of the first group that has been removed as the head value of the new first group, and jump to execute the step of sequentially taking values ​​and adding them to the first group to perform normality test and generate test results, until the number of normality test tests reaches the test threshold.

[0015] Optionally, the step of sorting the historical user load current values ​​in each group of load data from smallest to largest, taking values ​​sequentially and adding them to the first group for normality testing, and generating test results includes:

[0016] Sort the historical user load current values ​​in each group of load data from smallest to largest, and add them to the first group in sequence;

[0017] When the number of times a value is taken is less than or equal to the number of times a value is taken, a KS test is performed on the first group to generate a KS test result.

[0018] When the number of times a value is taken exceeds the threshold, a SW test is performed on the first group to generate the SW test result.

[0019] Optionally, the step of selecting the abnormal transformer based on the anomaly detection result of the monthly user load current value within the associated normal distribution model when all the transformers are received includes:

[0020] When the monthly user load current values ​​corresponding to all the transformers are received, the mean-standard deviation distribution of the monthly user load current values ​​in the associated normal distribution model is plotted.

[0021] If the mean-standard deviation distribution map satisfies any of the preset anomaly criteria, then an anomaly result is generated indicating that the transformer is an abnormal transformer, and the abnormal transformer is selected.

[0022] Optionally, the load indicators include: fuse fusing current, capacity, three-phase imbalance, and monthly user electricity consumption; the step of generating and outputting a rectification strategy corresponding to the preset overload condition if the load indicators of the abnormal transformer meet the preset overload condition includes:

[0023] Based on the capacity of the abnormal transformer, calculate the rated current on the secondary side; if the fuse current is greater than a preset multiple of the rated current on the secondary side, it is determined that the fuse of the abnormal transformer is mismatched, and a corresponding rectification strategy is generated and output.

[0024] When the three-phase unbalance of the abnormal transformer is greater than the unbalance threshold, the three-phase load of the abnormal transformer is determined to be unbalanced, and a corresponding rectification strategy is generated and output.

[0025] If the month-on-month ratio of the user electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, it is determined that the user terminal's electricity consumption fluctuates greatly, and a corresponding rectification strategy is generated and output.

[0026] Calculate the ratio of the annual electricity consumption of the residential area associated with the abnormal transformer to the annual electricity consumption of the residential area associated with the abnormal transformer. If the ratio continues to rise for three consecutive years, determine whether the residential area needs to be transformed, generate the corresponding transformation prediction strategy and output it.

[0027] Optionally, the step of determining that the three-phase load of the abnormal transformer is unbalanced when the three-phase imbalance of the abnormal transformer is greater than the imbalance threshold, generating a corresponding rectification strategy and outputting it includes:

[0028] When the three-phase imbalance of the abnormal transformer is greater than the imbalance threshold, the user power consumption of all associated user terminals of the largest phase and the user power consumption of all associated user terminals of the smallest phase are obtained and the difference is calculated.

[0029] Sort the user electricity consumption of all user terminals associated with the largest phase from largest to smallest, and accumulate them sequentially until the accumulated value is greater than half of the difference.

[0030] Generate and output a strategy to migrate the user terminal corresponding to the accumulated value to the smallest phase of the abnormal transformer.

[0031] Optionally, the step of determining that the user terminal's electricity consumption fluctuates significantly when the month-on-month ratio of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, generating a corresponding rectification strategy, and outputting it includes:

[0032] If the month-on-month ratio of the user electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, it is determined that the power consumption of the user terminal fluctuates greatly.

[0033] Generate and output a rectification strategy that will switch user terminals with large power fluctuations to those powered by other transformers.

[0034] This invention also provides a transformer overload monitoring device based on an RPA robot, applied to an RPA robot, wherein each transformer is associated with multiple user terminals, including:

[0035] The load data acquisition unit is used to acquire and classify the historical user load current values ​​corresponding to all user terminals in the jurisdiction, and generate multiple sets of load data.

[0036] The normal grouping generation unit is used to perform normal distribution tests on each group of load data to generate multiple groups of normal groups;

[0037] The normal distribution model generation unit is used to retrieve the normal groupings of each group, determine the normal distribution associated with each transformer and superimpose them to generate a normal distribution model for each transformer.

[0038] The abnormal transformer selection unit is used to select abnormal transformers according to the anomaly detection results of the monthly user load current values ​​corresponding to all the transformers when the monthly user load current values ​​are received.

[0039] The output unit is used to generate and output a rectification strategy corresponding to the preset overload condition if the load index of the abnormal transformer meets the preset overload condition.

[0040] Optionally, the normal grouping generation unit includes:

[0041] The verification subunit is used to sort the historical user load current values ​​in each group of load data from smallest to largest, take the values ​​in sequence and add them to the first group for normal distribution verification, and generate verification results.

[0042] A normal grouping generation subunit is used to remove the tail value of the first group and generate a normal group when the test result does not meet the preset normal distribution result;

[0043] The jump rotor unit is used to select the tail value of the first group that has been removed as the head value of the new first group, and jump to execute the step of sequentially taking values ​​and adding them to the first group to perform normal distribution test and generate test results, until the number of normal distribution tests reaches the test threshold.

[0044] Optionally, the abnormal transformer selection unit includes:

[0045] The distribution plotting subunit is used to plot the mean-standard deviation distribution of the monthly user load current values ​​within the associated normal distribution model when all the monthly user load current values ​​corresponding to the transformers are received.

[0046] An abnormal transformer selection sub-unit is used to generate an anomaly result for the transformer as an abnormal transformer if the mean-standard deviation distribution map meets any of the preset anomaly criteria, and then select the abnormal transformer.

[0047] As can be seen from the above technical solutions, the present invention has the following advantages:

[0048] This method acquires and categorizes historical user load current values ​​corresponding to all user terminals within the jurisdiction, generating multiple sets of load data. Each set of load data undergoes a normal distribution test, generating multiple sets of normal distribution groups. These normal distribution groups are retrieved, and the associated normal distribution for each transformer is determined and superimposed to generate a normal distribution model for each transformer. When the monthly user load current values ​​corresponding to all transformers are received, abnormal transformers are selected based on the anomaly detection results of the monthly user load current values ​​within the associated normal distribution models. If the load indicators of the abnormal transformer meet preset overload conditions, a rectification strategy corresponding to the preset overload conditions is generated and output. This method solves the technical problems of existing technologies that require significant manual labor, involve large amounts of data, and are difficult to efficiently and scientifically monitor transformers and generate rectification plans for transformer overload operation. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A flowchart illustrating the steps of a transformer overload monitoring method based on an RPA robot, as provided in this embodiment of the invention;

[0051] Figure 2 A flowchart illustrating the steps of an RPA robot-based transformer overload monitoring method, provided as an optional embodiment of the present invention;

[0052] Figure 3 This invention provides a structural block diagram of a transformer overload monitoring device based on an RPA robot. Detailed Implementation

[0053] This invention provides a transformer overload monitoring method based on RPA robots, which solves the technical problem that existing technologies require a large amount of manual labor and involve a large amount of data, making it difficult to efficiently and scientifically monitor transformers and generate rectification plans for transformer overload operation.

[0054] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0055] RPA Robotics: A Python-based process design software that helps to easily automate processes.

[0056] Three-phase unbalance: This is a coefficient that measures the degree of three-phase imbalance in a three-phase power system. It can be expressed as a percentage of the root mean square value of the negative-sequence component of voltage or current to the root mean square value of the positive-sequence component.

[0057] The permissible voltage imbalance values ​​specified in GB / T 15543-2008 are as follows: the permissible normal voltage imbalance value at the point of common coupling of the power system is 2%, and shall not exceed 4% for short periods; for each customer at the common node, the permissible value for normal voltage imbalance at that point is generally 1.3%.

[0058] Please see Figure 1 , Figure 1 The flowchart illustrates the steps of a transformer overload monitoring method based on an RPA robot, as provided in this embodiment of the invention.

[0059] This invention provides a transformer overload monitoring method based on an RPA robot, applied to an RPA robot, where each transformer is associated with multiple user terminals. The method includes:

[0060] Step 101: Obtain and classify the historical user load current values ​​corresponding to all user terminals within the jurisdiction, and generate multiple sets of load data.

[0061] In this embodiment of the application, the historical user load current value corresponding to all user terminals in the jurisdiction is first obtained. Then, the historical user load current value is classified according to industry type and month to generate multiple sets of load data. This is to take into account that the electricity consumption of different industries such as residential, commercial and agricultural will vary significantly with the seasons.

[0062] Step 102: Perform normality tests on each group of load data to generate multiple groups of normal distributions.

[0063] It should be noted that after obtaining multiple sets of load data, in order to ensure that each set of load data follows a normal distribution, it is necessary to further subdivide the load data into smaller subdivisions to improve the accuracy of the normal distribution of each set of load data. Therefore, each set of load data is subjected to a normal distribution test to generate multiple sets of normal groups.

[0064] Step 103: Retrieve each group of normal distributions, determine the normal distribution associated with each transformer, and superimpose them to generate a normal distribution model for each transformer.

[0065] In this embodiment of the application, in order to obtain the normal distribution model of each transformer, according to the principle of superposition of normal distributions, each group of normal distributions is retrieved by the historical user load current value corresponding to the user terminal associated with each transformer, thereby determining the normal distribution associated with each transformer and superimposing it, and finally obtaining the normal distribution model of each transformer.

[0066] Step 104: When the monthly user load current values ​​corresponding to all transformers are received, select the abnormal transformer according to the anomaly detection results of the monthly user load current values ​​in the associated normal distribution model.

[0067] It should be noted that when the monthly user load current values ​​corresponding to all transformers are received, the abnormal transformers are selected according to the anomaly detection results of the monthly user load current values ​​within the associated normal distribution model.

[0068] Step 105: If the load index of the abnormal transformer meets the preset overload condition, generate and output the rectification strategy corresponding to the preset overload condition.

[0069] It should be noted that if the load index of the abnormal transformer meets the preset overload conditions, a list of rectification strategies corresponding to the preset overload conditions will be generated and output.

[0070] In this embodiment, the historical user load current values ​​corresponding to all user terminals within the jurisdiction are obtained and classified to generate multiple sets of load data. Each set of load data is then subjected to a normal distribution test to generate multiple sets of normal distribution groups. Each set of normal distribution groups is retrieved, and the normal distribution associated with each transformer is determined and superimposed to generate a normal distribution model for each transformer. When the monthly user load current values ​​corresponding to all transformers are received, abnormal transformers are selected based on the anomaly detection results of the monthly user load current values ​​within the associated normal distribution model. If the load index of the abnormal transformer meets the preset overload conditions, a rectification strategy corresponding to the preset overload conditions is generated and output. This method can use RPA robots to replace manual labor in performing tedious information analysis, information processing, and system operations, enabling efficient and scientific monitoring of transformers and providing rectification plans for transformer overload operation.

[0071] Please see Figure 2 , Figure 2 The flowchart illustrates the steps of an RPA robot-based transformer overload monitoring method, which is an optional embodiment of the present invention.

[0072] This invention provides a transformer overload monitoring method based on RPA robots, applicable to multiple terminals, each with a corresponding load. The method includes:

[0073] Step 201: Obtain and classify the historical user load current values ​​corresponding to all user terminals within the jurisdiction, and generate multiple sets of load data.

[0074] In this embodiment of the application, the specific implementation process of step 201 is similar to that of step 101 above, and will not be repeated here.

[0075] Step 202: Perform normality tests on each group of load data to generate multiple groups of normal distributions.

[0076] It should be noted that step 202 includes the following sub-steps:

[0077] Step S101: Sort the historical user load current values ​​in each group of load data from smallest to largest, take the values ​​in sequence and add them to the first group for normal distribution test, and generate test results.

[0078] Optionally, step S101 includes: sorting the historical user load current values ​​in each group of load data from smallest to largest, and taking values ​​in sequence to add them to the first group; when the number of times the value is taken is less than or equal to the number of times threshold, performing a KS test on the first group and generating a KS test result; when the number of times the value is taken is greater than the number of times threshold, performing a SW test on the first group and generating a SW test result; wherein, the number of times threshold can be set to 50.

[0079] It is worth mentioning that the KS test or SW test can be used with the help of SPSS tools.

[0080] Step S102: When the test results do not meet the preset normal distribution results, remove the tail value of the first group and generate a normal group.

[0081] Step S103: Select the tail value of the first group that was removed as the head value of the new first group, and jump to execute the step of taking values ​​in sequence and adding them to the first group to perform normality test and generate test results, until the number of normality test tests reaches the test threshold.

[0082] In this embodiment of the application, since the historical user load current values ​​in each month of each industry are sorted from smallest to largest, the mean values ​​of each group of normal groups follow different normal distributions.

[0083] Step 203: Retrieve each group of normal distributions, determine the normal distribution associated with each transformer, and superimpose them to generate a normal distribution model for each transformer.

[0084] It should be noted that, in order to generate the normal distribution model of each transformer, based on the principle of superposition of normal distributions, the normal distribution groups are retrieved by searching the historical user load current values ​​corresponding to the user terminals associated with each transformer, thereby determining the normal distribution associated with each transformer and superimposing them to finally obtain the normal distribution model of each transformer; where the horizontal axis of the normal distribution model is the user load current value and the vertical axis is the probability density.

[0085] Step 204: When the monthly user load current values ​​corresponding to all transformers are received, draw the mean-standard deviation distribution map of the monthly user load current values ​​in the associated normal distribution model.

[0086] It should be noted that when the monthly user load current values ​​corresponding to all transformers are received, it is necessary to plot the mean-standard deviation distribution of the monthly user load current values ​​within the associated normal distribution model for subsequent anomaly detection analysis.

[0087] Step 205: If the mean-standard deviation distribution plot satisfies any of the preset anomaly criteria, then an anomaly result is generated indicating that the transformer is an abnormal transformer, and the abnormal transformer is selected.

[0088] Table 1. Normal Distribution Probability Table for a Certain Transformer

[0089]

[0090] In the specific implementation, Table 1 provides a normal distribution probability table for a certain transformer, which follows a distribution μ = 0.7I. max σ=0.05I max The normal distribution is given, where I is the user load current value; since the study focuses on transformer overload operation, it is only necessary to consider the transformer operating at a load greater than 0.7I. max The user load current value is considered, i.e., only one-sided probability is taken; among which... Figure 3 The probabilities provided are all less than 50%, which are one-sided probabilities. In actual use, you only need to add 50% directly.

[0091] The probability z in the value of a transformer exceeding the mean by z times the standard deviation is composed of two parts: the first two digits of z are formed by 'a' in the normal distribution probability table, and the last digit of z is formed by 'b' in the normal distribution probability table. The corresponding z is found by summing a and b, thus determining the probability of z. For example, to find the probability of a user load current value exceeding the mean by 0.11 times the standard deviation, the corresponding probability is 0.0438, which is found by summing 0.1 and 0.01. To find the probability of a user load current value exceeding the mean by 1.53 times the standard deviation, the corresponding probability is 0.4370, which is found by summing 1.5 and 0.03.

[0092] Assuming the probability of detection is 1%, the corresponding preset detection criteria specifically include:

[0093] Rule 1: A point is far from the center line and exceeds 2.33 times the standard deviation of the center line;

[0094] Among them, the probability found in the normal distribution probability table is P(μ+2.33σ,∞)=1-(0.4901+0.5)=0.99%, which is equivalent to a probability of 0.99% that a user's load current value exceeds 2.33 times the standard deviation of the mean.

[0095] Criterion 2: Seven consecutive points are to the right of the center line and are less than 2.33 times the standard deviation of the center line;

[0096] Where, P(n) = 0.4901 n When n is 7, the probability of 7 consecutive points being greater than the mean is 0.4901 to the power of 7, or 0.679%.

[0097] Rule 3: Six consecutive points of continuous increase or decrease;

[0098] in, When n is 6, the probability of 6 consecutive points rising or falling is 0.262%.

[0099] Criterion 4: Two out of three consecutive points are greater than 1.5 times the standard deviation of the center line and less than 2.33 times the standard deviation of the center line;

[0100] Among these, the probability that a point on the right side of the center line falls between 1.5 times the standard deviation of the mean and 2.33 times the standard deviation of the mean is 0.9901 - 0.9332 = 0.0569. Since two out of three consecutive points have three possible outcomes, the probability of the event occurring is P = 3 × 0.0569. 2 ×(0.9901-0.0569)=0.906%.

[0101] Criterion 5: Four out of five consecutive points are greater than 0.75 times the standard deviation of the center line and less than 2.33 times the standard deviation of the center line.

[0102] The probability that 4 out of 5 consecutive points are greater than 0.75 standard deviations above the center line and less than 2.33 standard deviations above the center line is:

[0103] Rule 6: Five consecutive points must be to the right of the center line and less than one standard deviation from the center line.

[0104] The probability that 5 consecutive points fall between the mean and one standard deviation from the mean is P = (0.3413). 5 =0.463%.

[0105] Step 206: If the load index of the abnormal transformer meets the preset overload condition, generate and output the rectification strategy corresponding to the preset overload condition.

[0106] It should be noted that the load indicators include: fuse fusing current, capacity, three-phase imbalance, and monthly electricity consumption, etc., among which step 206 includes:

[0107] Step S201: Calculate the rated current on the secondary side based on the capacity of the abnormal transformer; if the fuse blowing current is greater than the preset multiple of the rated current on the secondary side, it is determined that the fuse of the abnormal transformer is mismatched, and a corresponding rectification strategy is generated and output.

[0108] In the specific implementation, assuming the capacity of the abnormal transformer is 100kVA, the rated current of the secondary side of the abnormal transformer should be I = 100 ÷ (1.732 × 0.4) ≈ 144A, which means that the fuse current of the abnormal transformer should be within 216A. If the obtained fuse current of the abnormal transformer is 250A, it is determined that the fuse of the abnormal transformer is mismatched, and a corresponding rectification strategy is generated. The format of the rectification strategy is "Transformer A has a capacity of 100kVA, the currently installed 250A fuse is mismatched, and a fuse of 216A or less should be installed".

[0109] Step S202: When the three-phase unbalance of the abnormal transformer is greater than the unbalance threshold, the three-phase load of the abnormal transformer is determined to be unbalanced, and a corresponding rectification strategy is generated and output.

[0110] It should be noted that when the three-phase imbalance of the abnormal transformer is greater than the imbalance threshold, the user electricity consumption of all associated user terminals of the largest phase and the user electricity consumption of all associated user terminals of the smallest phase are obtained and the difference is calculated; the user electricity consumption of all associated user terminals of the largest phase is sorted from largest to smallest and accumulated sequentially until the accumulated value is greater than half of the difference; a strategy for migrating the user corresponding to the accumulated value to the smallest phase of the abnormal transformer is generated and output.

[0111] The unbalance threshold is specifically the allowable unbalance value specified in GB / T 15543-2008. Meanwhile, the allowable voltage unbalance at the point of common coupling of the power system is 2%. Therefore, the unbalance threshold can be set to 2%.

[0112] In specific implementation, when the three-phase imbalance of the abnormal transformer is greater than the imbalance threshold, the format of the rectification strategy is as follows: "The three-phase imbalance of transformer A is 2.5%, which does not meet the requirements. The current phase with the largest load is phase A, and the daily load difference between phase A and phase C is 200 kWh. The users with larger loads in phase A are A (50 kWh), B (40 kWh), and C (30 kWh). It is advisable to consider migrating the above user terminals to phase C."

[0113] Step S203: When the month-on-month ratio of the user electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, it is determined that the user terminal's electricity consumption fluctuates greatly, and a corresponding rectification strategy is generated and output.

[0114] It should be noted that when the month-on-month ratio of the user electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, it is determined that the user terminal's electricity consumption fluctuates greatly; a rectification strategy is generated and output to transfer the user terminal associated with the large electricity consumption fluctuation to be powered by other transformers.

[0115] In practice, the month-on-month threshold can be set to 10%. When the month-on-month ratio of the electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, the generated rectification strategy is in the format of: "Users with a month-on-month electricity consumption ratio of transformer A exceeding 10% include A (37% increase) and B (18% increase). It is advisable to consider transferring the power supply of this user terminal to other transformers."

[0116] Step S204: Calculate the ratio of annual electricity consumption of residents associated with abnormal transformers to annual electricity consumption of residents associated with abnormal transformers. If the ratio continues to rise for three consecutive years, determine whether the residential area needs to be transformed, generate the corresponding transformation prediction strategy and output it.

[0117] In practice, the transformation forecast strategy is formatted as follows: "The proportion of electricity consumption in residential areas of transformer A is increasing year by year. Please analyze whether to transform it into a non-residential area in conjunction with the local government's development plan."

[0118] In this embodiment, the historical user load current values ​​corresponding to all user terminals within the jurisdiction are obtained and classified to generate multiple sets of load data. Each set of load data is then subjected to a normal distribution test to generate multiple sets of normal distribution groups. Each set of normal distribution groups is retrieved, and the normal distribution associated with each transformer is determined and superimposed to generate a normal distribution model for each transformer. When the monthly user load current values ​​corresponding to all transformers are received, a mean-standard deviation distribution map of the monthly user load current values ​​within the associated normal distribution model is plotted. If the mean-standard deviation distribution map satisfies any of the preset anomaly criteria, an anomaly result is generated, and the abnormal transformer is selected. If the load index of the abnormal transformer meets the preset overload conditions, a rectification strategy corresponding to the preset overload conditions is generated and output. This method can use RPA robots to replace manual labor in performing tedious information analysis, information processing, and system operations, enabling efficient and scientific monitoring of transformers and providing rectification solutions for transformer overload operation.

[0119] Please see Figure 3 , Figure 3 This is a structural block diagram of a transformer overload monitoring device based on an RPA robot, provided in an embodiment of the present invention.

[0120] This invention also provides a transformer overload monitoring device based on an RPA robot, applied to an RPA robot, wherein each transformer is associated with multiple user terminals, and the device includes:

[0121] The load data acquisition unit 301 is used to acquire and classify the historical user load current values ​​corresponding to all user terminals in the jurisdiction, and generate multiple sets of load data.

[0122] The normal grouping generation unit 302 is used to perform normality tests on each group of load data and generate multiple groups of normal groups.

[0123] The normal distribution model generation unit 303 is used to retrieve each group of normal groups, determine the normal distribution associated with each transformer and superimpose them to generate the normal distribution model of each transformer;

[0124] The abnormal transformer selection unit 304 is used to select abnormal transformers according to the anomaly judgment results of the monthly user load current values ​​in the associated normal distribution model when all transformers are received respectively.

[0125] The output unit 305 is used to generate and output the rectification strategy corresponding to the preset overload condition if the load index of the abnormal transformer meets the preset overload condition.

[0126] Optionally, the normal grouping generation unit 302 includes:

[0127] The test subunit is used to sort the historical user load current values ​​in each group of load data from smallest to largest, take the values ​​in sequence and add them to the first group for normal distribution test, and generate test results;

[0128] The normal grouping generation sub-unit is used to remove the tail value of the first group and generate a normal group when the test result does not meet the preset normal distribution result;

[0129] The jump rotor unit is used to select the tail value of the first group to be removed as the head value of the new first group, jump to execute the step of taking values ​​in sequence and adding them to the first group to perform normal distribution test and generate test results, until the number of normal distribution tests reaches the test threshold.

[0130] Optionally, the verification subunit is also used for:

[0131] Sort the historical user load current values ​​in each load data group from smallest to largest, and add them to the first group in sequence;

[0132] When the number of times a value is taken is less than or equal to the number of times a value is taken, a KS test is performed on the first group to generate the KS test results.

[0133] When the number of times a value is taken exceeds the threshold, a SW test is performed on the first group to generate the SW test results.

[0134] Optionally, the abnormal transformer selection unit 304 includes:

[0135] The distribution plotting sub-unit is used to plot the mean-standard deviation distribution of the monthly user load current value within the associated normal distribution model when all the user load current values ​​corresponding to each transformer for the month are received.

[0136] The abnormal transformer selection sub-unit is used to generate an abnormal transformer result when the mean-standard deviation distribution map meets any of the preset anomaly criteria, and then select the abnormal transformer.

[0137] Optionally, the output unit 305 includes:

[0138] The fuse rectification strategy output subunit is used to calculate the secondary side rated current based on the capacity of the abnormal transformer. If the fuse breaking current is greater than the secondary side rated current by a preset multiple, it is determined that the fuse of the abnormal transformer is mismatched, and the corresponding rectification strategy is generated and output.

[0139] The three-phase imbalance rectification strategy output subunit is used to determine the three-phase load imbalance of the abnormal transformer when the three-phase imbalance degree of the abnormal transformer is greater than the imbalance degree threshold, generate the corresponding rectification strategy and output it.

[0140] The power fluctuation rectification strategy output subunit is used to determine that the power fluctuation of the user terminal is large when the month-on-month ratio of the user power consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, generate the corresponding rectification strategy and output it.

[0141] The transformation prediction strategy output sub-unit is used to calculate the ratio of the annual electricity consumption of residents associated with abnormal transformers to the annual electricity consumption of residents associated with abnormal transformers. If the ratio continues to rise for three consecutive years, it is determined whether the residential area should undergo transformation, and the corresponding transformation prediction strategy is generated and output.

[0142] Optionally, the three-phase imbalance rectification strategy output sub-unit is also used for:

[0143] When the three-phase imbalance of the abnormal transformer is greater than the imbalance threshold, obtain the user electricity consumption of all associated user terminals of the largest phase and the user electricity consumption of all associated user terminals of the smallest phase and calculate the difference.

[0144] Sort the user electricity consumption of all user terminals associated with the largest phase from largest to smallest, and accumulate them sequentially until the accumulated value is greater than half of the difference.

[0145] Generate and output a strategy to migrate the user terminal corresponding to the accumulated value to the smallest phase of the abnormal transformer.

[0146] Optionally, the power fluctuation rectification strategy output subunit is also used for:

[0147] If the month-on-month ratio of the user electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, it is determined that the user terminal's electricity consumption fluctuates greatly.

[0148] Generate and output a rectification strategy that switches user terminals with large power fluctuations to power supplied by other transformers.

[0149] In this embodiment, the load data acquisition unit acquires and classifies the historical user load current values ​​corresponding to all user terminals within the jurisdiction, generating multiple sets of load data. The normal distribution generation unit performs normal distribution tests on each set of load data, generating multiple sets of normal distributions. The normal distribution model generation unit retrieves each set of normal distributions, determines the normal distribution associated with each transformer, and superimposes them to generate a normal distribution model for each transformer. When the abnormal transformer selection unit receives the monthly user load current values ​​corresponding to all transformers, it selects abnormal transformers according to the anomaly results of the monthly user load current values ​​within the associated normal distribution model. The output unit determines if the load index of the abnormal transformer meets the preset overload conditions, generates and outputs the rectification strategy corresponding to the preset overload conditions. This method can use RPA robots to replace manual labor for tedious information analysis, information processing, and system operation, and can efficiently and scientifically monitor transformers and provide rectification solutions for transformer overload operation.

[0150] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0151] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and other division methods may be used in actual implementation.

[0152] Furthermore, the functional units in the various embodiments of 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. The integrated unit can be implemented in hardware or as a software functional unit.

[0153] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications 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.

Claims

1. A method for monitoring transformer overload based on RPA robot, characterized in that, Applied to RPA robots, where each transformer is associated with multiple user terminals, the method includes: Obtain and classify the historical user load current values ​​corresponding to all user terminals within the jurisdiction, and generate multiple sets of load data; Each group of load data is subjected to a normality test to generate multiple groups of normal groups. Retrieve the normal groupings in each group, determine the normal distribution associated with each transformer, and superimpose them to generate a normal distribution model for each transformer; When the monthly user load current values ​​corresponding to all the transformers are received, the abnormal transformer is selected according to the anomaly detection result of the monthly user load current values ​​in the associated normal distribution model. If the load index of the abnormal transformer meets the preset overload condition, then the rectification strategy corresponding to the preset overload condition is generated and output. The step of performing a normality test on each group of load data to generate multiple normally distributed groups includes: The historical user load current values ​​in each group of load data are sorted from smallest to largest, and the values ​​are taken sequentially and added to the first group for normality test, generating the test results. If the test result does not meet the preset normal distribution result, the tail value of the first group is removed, and a normal group is generated; Select the tail value of the first group that has been removed as the head value of the new first group, and jump to execute the step of sequentially taking values ​​and adding them to the first group to perform normality test and generate test results, until the number of normality test tests reaches the test threshold.

2. The transformer overload monitoring method based on RPA robot according to claim 1, characterized in that, The step of sorting the historical user load current values ​​in each group of load data from smallest to largest, taking values ​​sequentially and adding them to the first group for normality testing, and generating test results includes: Sort the historical user load current values ​​in each group of load data from smallest to largest, and add them to the first group in sequence; When the number of times a value is taken is less than or equal to the number of times a value is taken, a KS test is performed on the first group to generate a KS test result. When the number of times a value is taken exceeds the threshold, a SW test is performed on the first group to generate the SW test result.

3. The transformer overload monitoring method based on RPA robot according to claim 1, characterized in that, The step of selecting abnormal transformers based on the anomaly detection results of the monthly user load current values ​​corresponding to all the transformers when the monthly user load current values ​​are received includes: When the monthly user load current values ​​corresponding to all the transformers are received, the mean-standard deviation distribution of the monthly user load current values ​​in the associated normal distribution model is plotted. If the mean-standard deviation distribution map satisfies any of the preset anomaly criteria, then an anomaly result is generated indicating that the transformer is an abnormal transformer, and the abnormal transformer is selected.

4. The transformer overload monitoring method based on RPA robot according to claim 1, characterized in that, The load indicators include: fuse fusing current, capacity, three-phase imbalance, and monthly user electricity consumption; the step of generating and outputting a rectification strategy corresponding to the preset overload condition if the load indicators of the abnormal transformer meet the preset overload condition includes: Based on the capacity of the abnormal transformer, calculate the rated current on the secondary side; if the fuse current is greater than a preset multiple of the rated current on the secondary side, it is determined that the fuse of the abnormal transformer is mismatched, and a corresponding rectification strategy is generated and output. When the three-phase unbalance of the abnormal transformer is greater than the unbalance threshold, the three-phase load of the abnormal transformer is determined to be unbalanced, and a corresponding rectification strategy is generated and output. If the month-on-month ratio of the user electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, it is determined that the user terminal's electricity consumption fluctuates greatly, and a corresponding rectification strategy is generated and output. Calculate the ratio of the annual electricity consumption of the residential area associated with the abnormal transformer to the annual electricity consumption of the residential area associated with the abnormal transformer. If the ratio continues to rise for three consecutive years, determine whether the residential area needs to be transformed, generate the corresponding transformation prediction strategy and output it.

5. The transformer overload monitoring method based on RPA robot according to claim 4, characterized in that, The step of determining that the three-phase load of the abnormal transformer is unbalanced when the three-phase imbalance of the abnormal transformer is greater than the imbalance threshold, generating a corresponding rectification strategy and outputting it includes: When the three-phase imbalance of the abnormal transformer is greater than the imbalance threshold, the power consumption of all user terminals associated with the largest phase and the power consumption of all user terminals associated with the smallest phase are obtained and the difference is calculated. Sort the user electricity consumption of all user terminals associated with the largest phase from largest to smallest, and accumulate them sequentially until the accumulated value is greater than half of the difference. Generate and output a strategy to migrate the user terminal corresponding to the accumulated value to the smallest phase of the abnormal transformer.

6. The transformer overload monitoring method based on RPA robot according to claim 4, characterized in that, The step of determining that the user terminal's electricity consumption fluctuates significantly when the month-on-month ratio of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, generating a corresponding rectification strategy, and outputting it includes: If the month-on-month ratio of the user electricity consumption of the user terminal associated with the abnormal transformer is greater than the month-on-month threshold, it is determined that the power consumption of the user terminal fluctuates greatly. Generate and output a rectification strategy that will switch user terminals that are highly correlated with the power fluctuations to be powered by other transformers.

7. A transformer overload monitoring device based on RPA robot, characterized in that, In RPA robots, each transformer is associated with multiple user terminals, including: The load data acquisition unit is used to acquire and classify the historical user load current values ​​corresponding to all user terminals in the jurisdiction, and generate multiple sets of load data. The normal grouping generation unit is used to perform normal distribution tests on each group of load data to generate multiple groups of normal groups; The normal distribution model generation unit is used to retrieve the normal groupings of each group, determine the normal distribution associated with each transformer and superimpose them to generate a normal distribution model for each transformer. The abnormal transformer selection unit is used to select abnormal transformers according to the anomaly detection results of the monthly user load current values ​​corresponding to all the transformers when the monthly user load current values ​​are received. The output unit is used to generate and output a rectification strategy corresponding to the preset overload condition if the load index of the abnormal transformer meets the preset overload condition. The normal grouping generation unit includes: The verification subunit is used to sort the historical user load current values ​​in each group of load data from smallest to largest, take the values ​​in sequence and add them to the first group for normal distribution verification, and generate verification results. A normal grouping generation subunit is used to remove the tail value of the first group and generate a normal group when the test result does not meet the preset normal distribution result; The jump rotor unit is used to select the tail value of the first group that has been removed as the head value of the new first group, and jump to execute the step of sequentially taking values ​​and adding them to the first group to perform normal distribution test and generate test results, until the number of normal distribution tests reaches the test threshold.

8. The transformer overload monitoring device based on RPA robot according to claim 7, characterized in that, The abnormal transformer selection unit includes: The distribution plotting subunit is used to plot the mean-standard deviation distribution of the monthly user load current values ​​within the associated normal distribution model when all the monthly user load current values ​​corresponding to the transformers are received. An abnormal transformer selection sub-unit is used to generate an anomaly result for the transformer as an abnormal transformer if the mean-standard deviation distribution map meets any of the preset anomaly criteria, and then select the abnormal transformer.

Images