Electric energy meter safe electricity utilization monitoring method and system based on Internet of Things

Abstract

The invention relates to the technical field of data processing. The invention particularly relates to an electric energy meter safe electricity utilization monitoring method and system based on the Internet of Things. The method comprises the following steps: collecting power utilization monitoring data pairs of the electric energy meter; clustering data points of the collected power utilization monitoring data to obtain a cluster; obtaining the consistency of the data points in the clustering cluster and the outlier degree of each data point in the clustering cluster; the outlier degree of the data points is the degree that the index stronghold deviates from the cluster to which the index stronghold belongs, an LOF anomaly detection algorithm is used for obtaining the anomaly score of each collected data point, and when the anomaly scores are calculated, the optimal K value corresponding to the data points is calculated according to the outlier degree of the data points and the consistency of the data points in the cluster to which the data points belong; and according to the abnormal score of each data point, carrying out monitoring and early warning on the power consumption behavior. By adopting the method, the accuracy of the safe power utilization monitoring result of the power distribution network can be improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology. More specifically, the present invention relates to a method and system for monitoring safe electricity consumption of electric energy meters based on the Internet of Things. Background Art

[0002] With the development of science and technology, the use of power systems has become more and more popular, bringing many conveniences to people's daily life and production activities. However, at the user end of the distribution network, there will be users stealing electricity or electrical equipment failures. In order to ensure the safe and stable operation of the power grid, it is necessary to detect the power consumption behavior of the distribution network, so as to issue an early warning or disconnect the knife switch at the user end of the power grid when abnormal power consumption behavior occurs. The way to detect the power consumption behavior of the distribution network is usually to use an energy meter to collect power consumption data from the user end of the distribution network, such as current data, voltage data or power data. After collecting these power consumption data, they are sent to the data center through the Internet of Things. The data center detects whether there are abnormalities in the power consumption data, so as to determine whether there is dangerous power consumption behavior on the distribution network, thereby helping power companies and users to achieve more efficient power use and more accurate fault prediction.

[0003] LOF algorithm is usually used to detect whether there are anomalies in electricity consumption data. LOF algorithm is an algorithm used to detect anomalies in data sets and is suitable for identifying abnormal patterns in electricity consumption behavior. In the safe electricity consumption monitoring of the user end of the distribution network, LOF algorithm can help identify some abnormal electricity consumption patterns such as electricity theft and equipment failure.

[0004] However, the traditional LOF algorithm uses a fixed parameter K to obtain the abnormality of each data point. However, due to the volatility of electricity monitoring data, the fixed parameter K cannot reflect the fluctuation of the data well, which makes the abnormal detection result inaccurate, and further makes the early warning result inaccurate. For example, the real-time anti-electricity theft monitoring method of the electricity information collection system based on outlier detection disclosed in the Chinese patent application document with application publication number CN106373025A calculates the suspected electricity theft coefficient by combining the collected electrical parameters with the LOF algorithm. However, due to the volatility of the collected electrical parameters, the fixed parameter K cannot reflect the fluctuation of the data well, which makes the suspected electricity theft coefficient calculation result inaccurate, resulting in low accuracy of the monitoring results of the distribution network. Summary of the invention

[0005] In order to solve the technical problem of low accuracy of monitoring results of safe electricity consumption in distribution networks in the prior art, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for monitoring safe electricity consumption of an electric energy meter based on the Internet of Things, comprising: Collect electricity consumption monitoring data from electric energy meters; Clustering the data points of the collected electricity consumption monitoring data to obtain cluster clusters; Obtain the consistency of data points within the cluster and the degree of outliers of each data point within the cluster; the degree of outliers of a data point refers to the degree to which the data point deviates from the cluster to which it belongs; The LOF anomaly detection algorithm is used to obtain the anomaly score of each collected data point. When calculating the anomaly score, the optimal K value corresponding to the data point is calculated based on the degree of outlier of the data point and the consistency of the data points in the cluster to which it belongs. The calculation expression of the optimal K value corresponding to the data point is: ; In the formula, Indicates i The first j The best data point K value; Indicates the default K value, Indicates i The first j The degree of outlier of a data point; Indicates i The consistency of data points within a cluster; It represents the upward rounding function, and exp() represents the exponential function with the natural constant e as the base; represents the consistency threshold of the data points; Power usage behavior is monitored and early warning is issued based on the abnormal scores of each data point.

[0007] Its beneficial effects are as follows: during the peak period of power consumption by users, when multiple power-consuming devices are turned on at the same time, the acquired power consumption monitoring data will change, resulting in local similarities in the power consumption data, but certain differences in the global data. If the LOF algorithm is directly used for anomaly detection, the anomaly detection result will be inaccurate. Therefore, the method for monitoring safe power consumption of electric energy meters based on the Internet of Things of the present invention first clusters the power consumption monitoring data before using the LOF algorithm for anomaly detection; when using the LOF algorithm to calculate the anomaly score of the data point, the similarity between the features of different data points in the clustering cluster is taken into account. When the consistency of the data point in the clustering cluster to which the data point belongs is high, the optimal K value of the data point is inversely proportional to the degree of outlier of the data point; when the consistency of the data point in the clustering cluster to which the data point belongs is low, the optimal K value of the data point is proportional to the degree of outlier of the data point; thereby the optimal K value of the data point is adaptively adjusted, so that the calculated anomaly score is more accurate, thereby improving the accuracy of the results of the safe power consumption monitoring of the distribution network.

[0008] Preferably, obtaining the consistency of data points within the cluster includes: Obtain the differences in the collection time of data points within the clusters, as well as the uniformity of the distribution of data points within the clusters; The consistency of the data points in the cluster is calculated based on the difference in the acquisition time and the uniformity, and the calculation expression is: ; In the formula, Indicates the consistency of the data points within the i-th cluster, represents the difference in the collection time of data points within the i-th cluster, Indicates the uniformity of the distribution of data points within the i-th cluster.

[0009] The beneficial effects are: the more uniform the distribution of data points within the cluster, the more similar the numerical characteristics of each data point in the cluster, the lower the abnormal performance in the cluster, and the higher the consistency of the data points within the cluster; the greater the difference in the collection time of the data points within the cluster, the greater the deviation in the collection time of different data points within the cluster, and the lower the consistency of the data points within the cluster; therefore, the consistency of the data points within the cluster is proportional to the uniformity of the distribution of the data points within the cluster, and inversely proportional to the difference in the collection time of the data points within the cluster, so as to more accurately calculate the consistency of the data points within the cluster.

[0010] Preferably, the calculation expression for the difference in the collection time of data points within the cluster is: ; In the formula, Indicates i The total number of data points in the clusters; j Indicates the data point number; Indicates i The first j The time when the data points are collected; Indicates i The mean of the collection times of all data points in a cluster.

[0011] The beneficial effect is as follows: when calculating the difference in the collection time of data points within the cluster, the present invention first calculates the deviation between the collection time of each data point in the cluster and the mean collection time of all data points in the cluster, and takes the mean of the deviations of each data point as the difference in the collection time of the data points within the cluster, thereby making the calculated difference in the collection time of the data points within the cluster more accurate.

[0012] Preferably, the method for calculating the uniformity of the distribution of data points within the cluster includes: Calculate the Euclidean distance between each data point in the cluster and other data points; Calculate the variance of the Euclidean distance of each data point in the cluster respectively; The uniformity of the distribution of the data points within the cluster is calculated according to the variance of the Euclidean distance of each data point, and the uniformity is negatively correlated with the variance.

[0013] The beneficial effect is that the larger the variance of the Euclidean distance corresponding to a data point, the higher the difference in the Euclidean distance between the data point and other data points, and the more uneven the distribution of the data points. Therefore, the uniformity of the distribution of data points within the cluster is calculated based on the variance of the Euclidean distance corresponding to each data point, and the uniformity is negatively correlated with the variance, so as to more accurately calculate the uniformity of the distribution of data points within the cluster.

[0014] Preferably, the calculation expression for the uniformity of the distribution of data points within the cluster is: ; In the formula, represents the ordinal number of the data point, ; Indicates and The Euclidean distance between data points; Indicates The average Euclidean distance between a data point and all the data points in its cluster.

[0015] The beneficial effect is that when calculating the uniformity of the distribution of data points within the cluster, the variance of the Euclidean distance corresponding to each data point within the cluster is comprehensively considered, so that the calculated uniformity of the distribution of data points within the cluster is more accurate and objective.

[0016] Preferably, the method for calculating the degree of outlier of a data point includes: Get the Euclidean distance from the data point to the cluster center of the cluster to which it belongs, as well as the mean of the Euclidean distances from other data points in the cluster to the cluster center; The degree of outlier of the data point is calculated based on the mean of the Euclidean distance and the Euclidean distance from the data point to the cluster center of the cluster to which it belongs. The calculation expression is: ; In the formula, Indicates i The first j The degree of outlier of a data point, represents the standard normalization function, Indicates i The first j The Euclidean distance between a data point and the cluster center of the cluster to which it belongs; Indicates i In all clusters except jThe average of the Euclidean distances of all data points except the cluster center.

[0017] Its beneficial effects are: and The larger the ratio of , the larger the Euclidean distance of the data point from the cluster center relative to other data points in the cluster, and the more likely the data point is to deviate from the group composed of other data points in the cluster; on the contrary, the smaller the ratio of the Euclidean distance of the data point from the cluster center to the average value of the Euclidean distance of other data points in the cluster to the cluster center, the smaller the Euclidean distance of the data point from the cluster center relative to other data points in the cluster, and the less likely the data point is to deviate from the group composed of other data points in the cluster. Therefore, the calculation expression of the present invention can more accurately calculate the degree of outlier of the data point.

[0018] Preferably, monitoring and warning electricity usage behavior based on the abnormal scores of each data point includes: Calculate the mean of the anomaly scores of all data points, and compare the mean of the anomaly scores with a preset anomaly threshold; In response to the average being greater than a preset abnormal threshold, it is determined that there is a danger of electricity use, and an early warning is issued to the staff.

[0019] The beneficial effect is as follows: since the current and voltage data of the electric energy meter fluctuate during the electricity use process, the fluctuation may be caused by noise or load changes. If the presence of an electricity use hazard is judged based on the abnormal scores of a single or several data points, there is a risk of misjudgment. Therefore, the present invention calculates the mean of the abnormal scores of all data points, and uses the mean to judge whether there is dangerous electricity use, thereby avoiding interference from noise data and making the electric energy meter's safe electricity use monitoring results more accurate.

[0020] Preferably, an iterative self-organizing clustering algorithm is used when clustering the data points of the collected electricity consumption monitoring data.

[0021] Preferably, the collection frequency when collecting the electricity consumption monitoring data of the electric energy meter is 50HZ.

[0022] In a second aspect, the present invention provides an electric energy meter safety electricity usage monitoring system based on the Internet of Things, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the electric energy meter safety electricity usage monitoring method based on the Internet of Things of the present invention is implemented.

[0023] In summary, the Internet of Things-based electric energy meter safe electricity consumption monitoring system and method of the present invention can effectively improve the accuracy of the distribution network safe electricity consumption monitoring results. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 is a flowchart schematically illustrating a method for monitoring safe electricity consumption of an electric energy meter based on the Internet of Things according to an embodiment of the present invention; Figure 2 The diagram schematically shows the distribution of data points in a cluster according to an embodiment of the present invention. Figure 1 ; Figure 3 The diagram schematically shows the distribution of data points in a cluster according to an embodiment of the present invention. Figure 2 ; Figure 4 The figure schematically shows the structure of an electric energy meter safety electricity consumption monitoring system based on the Internet of Things according to an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0026] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0027] Embodiment of the method for monitoring safe electricity consumption of electric energy meters based on the Internet of Things: like Figure 1 As shown, the method for monitoring safe electricity consumption of electric energy meters based on the Internet of Things of the present invention comprises: S101, collecting electricity consumption monitoring data of the electric energy meter; In this embodiment, the collected power consumption monitoring data of the electric energy meter includes current data and voltage data, and the collection frequency is 50HZ. The current data and the voltage data can be collected by a current sensor and a voltage sensor respectively.

[0028] S102, clustering the data points of the collected electricity consumption monitoring data to obtain cluster clusters; When clustering the data points of the power consumption monitoring data, a K-means clustering algorithm, an iterative self-organizing clustering algorithm or other clustering algorithms may be used. Preferably, in this embodiment, when clustering the data points of the collected power consumption monitoring data, an iterative self-organizing clustering algorithm is used.

[0029] The clustering results include: data points in all clusters and the cluster center of each cluster.

[0030] S103, obtaining the consistency of the data points in the clusters and the degree of outliers of each data point in the clusters; the degree of outliers of a data point refers to the degree to which the data point deviates from the cluster to which it belongs; The consistency of data points within a cluster is used to characterize the deviation between different data points within the cluster, including the deviation of the collection time and the deviation of the value. Generally speaking, the farther the data point is from the cluster center of the cluster to which it belongs, the greater the degree of deviation of the data point from the cluster to which it belongs; therefore, the degree of outlier of the data point can be calculated based on the distance between the data point and the cluster center of the cluster to which it belongs. After calculating the degree of outlier of the data point and the degree of deviation from the cluster to which it belongs, it is convenient to calculate the optimal K value of the LOF anomaly detection algorithm in the subsequent step.

[0031] S104, obtaining the anomaly score of each collected data point; specifically: using the LOF anomaly detection algorithm to obtain the anomaly score of each collected data point, wherein, when calculating the anomaly score, the optimal K value corresponding to the data point is calculated based on the degree of outlier of the data point and the consistency of the data points in the cluster to which it belongs. The calculation expression of the optimal K value corresponding to the data point is: ; In the formula, Indicates i The first j The best data point K value; Indicates the default K value, Indicates i The first j The degree of outlier of a data point; Indicates i The consistency of data points within a cluster; It represents the upward rounding function, and exp() represents the exponential function with the natural constant e as the base; represents the consistency threshold of the data points; like , it means that the consistency of the data points in the cluster is high, that is, their characteristics are very similar. In this case, if a data point is significantly different from other data points in the cluster, this difference is unlikely to be caused by noise or error, but it is more likely that the data point itself is a real outlier, that is, the more credible the outlier degree of the data point is; because the high consistency of the cluster indicates that other data points are highly similar, the data point that is more different from other points in the cluster is more likely to be a real outlier, rather than accidental noise. Therefore, in this case, if the outlier degree of the data point is greater, the optimal K value of the data point is made smaller, and if the outlier degree of the data point is smaller, the optimal K value of the data point is made larger, so as to avoid the interference of noise data.

[0032] On the contrary, if , it means that the consistency of the data points in the cluster is low, the differences between the data points in the cluster are large, and the feature distribution within the cluster is relatively scattered. If a data point shows a large outlier feature at this time (for example, it is very different from other data points), this difference is likely not a true anomaly, but is simply due to the large overall difference in the data points in the cluster, causing the data point to be incorrectly classified as an outlier. Therefore, in this case, if the degree of outlier of the data point is greater, the optimal K value of the data point is made larger, and if the degree of outlier of the data point is smaller, the optimal K value of the data point is made smaller to avoid interference from noise data.

[0033] S105. Monitor and issue early warnings on electricity usage behavior based on the abnormal scores of each data point.

[0034] Monitoring and warning electricity usage behavior based on the abnormal scores of each data point includes determining whether there is a danger of electricity usage or abnormal electricity usage based on the abnormal scores of the data points. There are many ways to determine whether there is a danger of electricity usage or abnormal electricity usage based on the abnormal scores of the data points. For example, you can calculate whether the abnormal scores of each data point are greater than the abnormal score threshold. If the abnormal scores of all data points or some data points are greater than the abnormal score threshold, it can be determined that there is a danger of electricity usage or abnormal electricity usage. You can also calculate the average of the abnormal scores of all data points and determine whether there is a danger of electricity usage or abnormal electricity usage based on the average of the abnormal scores.

[0035] During the peak period of power consumption by users, when multiple power-consuming devices are turned on at the same time, the acquired power consumption monitoring data will change, resulting in local similarities in the power consumption data, but certain differences in the global data. If the LOF algorithm is directly used for anomaly detection, the anomaly detection result will be inaccurate. Therefore, the method for monitoring safe power consumption of electric energy meters based on the Internet of Things of the present invention first clusters the power consumption monitoring data before using the LOF algorithm for anomaly detection; when using the LOF algorithm to calculate the anomaly score of the data point, the similarity between the features of different data points in the clustering cluster is taken into account. When the consistency of the data point in the clustering cluster to which the data point belongs is high, the optimal K value of the data point is inversely proportional to the degree of outlier of the data point; when the consistency of the data point in the clustering cluster to which the data point belongs is low, the optimal K value of the data point is proportional to the degree of outlier of the data point; thereby the optimal K value of the data point is adaptively adjusted, so that the calculated anomaly score is more accurate, thereby improving the accuracy of the results of the safe power consumption monitoring of the distribution network.

[0036] In one embodiment, obtaining the consistency of data points in the cluster includes: S201, obtaining the difference in the collection time of data points within the clusters, and the uniformity of the distribution of the data points within the clusters; After the power consumption monitoring data is clustered, the data distribution in different clusters is different. Different distributions represent different power consumption performances of the power consumption data in the cluster. If the power consumption data in the cluster is distributed more evenly, it means that the abnormal performance in the cluster is lower and the consistency of the data points in the cluster is higher. On the contrary, if the power consumption data in the cluster is distributed more unevenly, it means that the abnormal performance in the cluster is higher and the consistency of the data points in the cluster is lower. Therefore, when calculating the consistency of the data points in the cluster, it is necessary to calculate the uniformity of the distribution of the data points in the cluster.

[0037] When clustering, since the clustering is performed based on the size of the electricity consumption monitoring data, the influence of the time factor is ignored. Therefore, it is also necessary to calculate the consistency of the data points in the cluster based on the difference in the collection time of the data points in the cluster.

[0038] The difference in the collection time of data points within a cluster refers to the deviation between the collection times of different data points. There are many ways to calculate the deviation, for example: calculating the variance or standard deviation corresponding to the collection time of different data points, and using the calculated variance or standard deviation as the deviation between the collection times of different data points, or calculating the mean of the collection time of different data points, and measuring the deviation between the collection times of different data points based on the deviation between each data point and the mean.

[0039] S202, calculating the consistency of the data points in the cluster according to the difference in the acquisition time and the uniformity, and the calculation expression is: ; In the formula, Indicates the consistency of the data points within the i-th cluster, represents the difference in the collection time of data points within the i-th cluster, Indicates the uniformity of the distribution of data points within the i-th cluster.

[0040] The more uniform the distribution of data points within the cluster, the more similar the numerical characteristics of each data point in the cluster, the lower the abnormal performance in the cluster, and the higher the consistency of the data points within the cluster; the greater the difference in the collection time of the data points within the cluster, the greater the deviation in the collection time of different data points within the cluster, and the lower the consistency of the data points within the cluster; therefore, the consistency of the data points within the cluster is proportional to the uniformity of the distribution of the data points within the cluster, and inversely proportional to the difference in the collection time of the data points within the cluster, so as to more accurately calculate the consistency of the data points within the cluster.

[0041] It can be seen from the above embodiments that when calculating the difference in the collection time of data points within a cluster, the mean of the collection time of different data points can be calculated, and the deviation between the collection time of different data points can be measured according to the deviation between each data point and the mean. In one embodiment, the calculation expression for the difference in the collection time of data points within a cluster is: ; In the formula, Indicates i The total number of data points in the clusters; j Indicates the data point number; Indicates i The first j The time when the data points are collected; Indicates i The mean of the collection times of all data points in a cluster.

[0042] In this embodiment, when calculating the differences in the collection times of data points within the cluster, first, for each data point in the cluster, the deviation between its collection time and the mean collection time of all data points in the cluster is calculated, and the mean of the deviations of each data point is used as the difference in the collection time of the data points within the cluster, so that the calculated differences in the collection time of the data points within the cluster are more accurate.

[0043] In one embodiment, a method for calculating the uniformity of data point distribution within a cluster includes: S301, calculating the Euclidean distance between each data point in the cluster and each other data point; Since a data point includes voltage data and current data, the voltage data value and the current data value can be set as the horizontal coordinate and vertical coordinate of the data point respectively, and then the Euclidean distance between the two data points can be calculated based on the coordinates of the two data points. For example: assuming that the voltage data value of data point A is x1 and the current data value is y1; the voltage data value of data point B is x2 and the current data value is y2; then the calculation expression of the Euclidean distance L between data point A and data point B is: .

[0044] S302, respectively calculating the variance of the Euclidean distance of each data point in the cluster; In a cluster, there is a Euclidean distance between a data point and all other data points in the cluster, that is, one data point corresponds to multiple Euclidean distances. Therefore, calculating the variance of these Euclidean distance clusters will obtain the variance of the Euclidean distance between the data point and other data points in the cluster.

[0045] S303: Calculate the uniformity of distribution of data points within the cluster according to the variance of the Euclidean distance of each data point, wherein the uniformity is negatively correlated with the variance.

[0046] The larger the variance of the Euclidean distance corresponding to a data point, the higher the difference in the Euclidean distance between the data point and other data points, and the more uneven the distribution of the data points. Therefore, the uniformity of the distribution of data points within the cluster is calculated based on the variance of the Euclidean distance corresponding to each data point, and the uniformity is negatively correlated with the variance, so as to more accurately calculate the uniformity of the distribution of data points within the cluster.

[0047] In one embodiment, the calculation expression for the uniformity of the distribution of data points within the cluster is: ; In the formula, represents the ordinal number of the data point, ; Indicates and The Euclidean distance between data points; Indicates The average Euclidean distance between a data point and all the data points in its cluster. Indicates The smaller the value of the distance difference between a data point and other data points, the smaller the distance difference between the data point and other data points. The mean of the distance difference between all data points and other data points in the cluster is obtained. The smaller its value is, the more evenly the data points in the cluster are distributed.

[0048] When calculating the uniformity of the distribution of data points within the cluster, the variance of the Euclidean distance corresponding to each data point in the cluster is comprehensively considered, so that the calculated uniformity of the distribution of data points in the cluster is more accurate and objective.

[0049] In one embodiment, the method for calculating the outlier degree of a data point includes: S401, obtaining the Euclidean distance from the data point to the cluster center of the cluster to which it belongs, and the average of the Euclidean distances from other data points in the cluster to which the corresponding data point belongs to the cluster center; S402, calculating the degree of outlier of the data point according to the mean of the Euclidean distance and the Euclidean distance from the data point to the cluster center of the cluster to which it belongs, and the calculation expression is: ; In the formula, Indicates i The first j The degree of outlier of a data point, represents the standard normalization function, Indicates i The first j The Euclidean distance between a data point and the cluster center of the cluster to which it belongs; Indicates i In all clusters except j The average of the Euclidean distances of all data points except the cluster center.

[0050] and The larger the ratio of , the larger the Euclidean distance of the data point from the cluster center relative to other data points in the cluster, and the more likely the data point is to deviate from the group composed of other data points in the cluster; on the contrary, the smaller the ratio of the Euclidean distance of the data point from the cluster center to the average value of the Euclidean distance of other data points in the cluster to the cluster center, the smaller the Euclidean distance of the data point from the cluster center relative to other data points in the cluster, and the less likely the data point is to deviate from the group composed of other data points in the cluster. Therefore, the calculation expression of this embodiment can more accurately calculate the degree of outlier of the data point.

[0051] like Figure 2 As shown in , when the ratio of the Euclidean distance between data point A and the cluster center (data point B) of the cluster it belongs to is larger than the average value of the Euclidean distance between other data points in the cluster and the cluster center, data point A deviates from the group composed of other data points in the cluster. Figure 3As shown in the figure, when the ratio of the Euclidean distance between data point A and the cluster center (data point B) of the cluster to which it belongs is smaller than the average value of the Euclidean distance between other data points in the cluster and the cluster center, data point A is closer to the group composed of other data points in the cluster.

[0052] It can be seen from the above embodiments that whether there is a danger of electricity use can be determined based on the average of the abnormal scores of all data points. In one embodiment, monitoring and warning electricity use behavior based on the abnormal scores of each data point includes: S501, calculating the mean of the abnormality scores of all data points, and comparing the mean of the abnormality scores with a preset abnormality threshold; S502: In response to the average being greater than a preset abnormal threshold, it is determined that there is a danger of electricity use, and an early warning is issued to the staff.

[0053] The preset abnormality threshold may be set to 0.6. In other embodiments, the preset abnormality threshold may also be set to other appropriate values.

[0054] During the electricity usage process, the current and voltage data of the electric energy meter fluctuate, and the fluctuation may be caused by noise or load changes. If the judgment on whether there is a danger in electricity usage is based on the abnormal scores of a single or several data points, there is a risk of misjudgment. Therefore, this embodiment calculates the mean of the abnormal scores of all data points, and uses the mean to judge whether there is dangerous electricity usage, thereby avoiding interference from noise data and making the safe electricity usage monitoring results of the electric energy meter more accurate.

[0055] Example of an energy meter safety electricity consumption monitoring system based on the Internet of Things: The present invention also provides an electric energy meter safety electricity consumption monitoring system based on the Internet of Things. Figure 4 As shown, the electric energy meter safety electricity usage monitoring system based on the Internet of Things includes a processor and a memory, and the memory stores computer program instructions. When the computer program instructions are executed by the processor, the electric energy meter safety electricity usage monitoring method based on the Internet of Things described in the above embodiments is implemented.

[0056] The Internet of Things-based electric energy meter safe electricity consumption monitoring system also includes other components familiar to technical personnel in this field, such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be repeated here.

[0057] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0058] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0059] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for monitoring safe electricity consumption of electric energy meters based on the Internet of Things, characterized in that: include: Collect electricity consumption monitoring data from electric energy meters; Clustering the data points of the collected electricity consumption monitoring data to obtain cluster clusters; Obtain the consistency of data points within the cluster and the degree of outliers of each data point within the cluster; The outlier degree of a data point refers to the degree to which the data point deviates from the cluster to which it belongs; The LOF anomaly detection algorithm is used to obtain the anomaly score of each collected data point. When calculating the anomaly score, the optimal K value corresponding to the data point is calculated based on the degree of outlier of the data point and the consistency of the data points in the cluster to which it belongs. The calculation expression of the optimal K value corresponding to the data point is: ； In the formula, Indicates i The first j The best data point K value; Indicates the default K value, Indicates i The first j The degree of outlier of a data point; Indicates i The consistency of data points within a cluster; It represents the upward rounding function, and exp() represents the exponential function with the natural constant e as the base; represents the consistency threshold of the data points; Power usage behavior is monitored and early warning is issued based on the abnormal scores of each data point.

2. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things according to claim 1, characterized in that: The obtaining of consistency of data points within the cluster includes: Obtain the differences in the collection time of data points within the clusters, as well as the uniformity of the distribution of data points within the clusters; The consistency of the data points in the cluster is calculated based on the difference in the acquisition time and the uniformity, and the calculation expression is: ； In the formula, Indicates the consistency of the data points within the i-th cluster, represents the difference in the collection time of the data points in the i-th cluster, Indicates the uniformity of the distribution of data points within the i-th cluster.

3. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things as claimed in claim 2, characterized in that: The calculation expression of the difference in the collection time of data points within the cluster is: ； In the formula, Indicates i The total number of data points in the clusters; j Indicates the data point number; Indicates i The first j The time when the data points are collected; Indicates i The mean value of the collection time of all data points in a cluster.

4. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things according to claim 2, characterized in that: The calculation methods for the uniformity of the distribution of data points within the cluster include: Calculate the Euclidean distance between each data point in the cluster and each other data point; Calculate the variance of the Euclidean distance of each data point in the cluster respectively; The uniformity of the distribution of the data points within the cluster is calculated according to the variance of the Euclidean distance of each data point, and the uniformity is negatively correlated with the variance.

5. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things as claimed in claim 4, characterized in that: The calculation expression for the uniformity of the distribution of data points within the cluster is: ； In the formula, represents the ordinal number of the data point, ; Indicates and The Euclidean distance between data points; Indicates The average Euclidean distance between a data point and all the data points in its cluster.

6. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things according to claim 1, characterized in that: The calculation method of the outlier degree of the data point includes: Get the Euclidean distance from the data point to the cluster center of the cluster to which it belongs, as well as the mean of the Euclidean distances from other data points in the cluster to the cluster center; The degree of outlier of the data point is calculated based on the mean of the Euclidean distance and the Euclidean distance from the data point to the cluster center of the cluster to which it belongs. The calculation expression is: ； In the formula, Indicates i The first j The degree of outlier of a data point, represents the standard normalization function, Indicates i The first j The Euclidean distance between a data point and the cluster center of the cluster to which it belongs; Indicates i In all clusters except j The average of the Euclidean distances of all data points except the cluster center.

7. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things according to claim 1, characterized in that: The monitoring and early warning of electricity usage behavior based on the abnormal scores of each data point includes: Calculate the mean of the anomaly scores of all data points, and compare the mean of the anomaly scores with a preset anomaly threshold; In response to the average being greater than a preset abnormal threshold, it is determined that there is a danger of electricity use, and an early warning is issued to the staff.

8. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things according to any one of claims 1 to 7, characterized in that: An iterative self-organizing clustering algorithm is used to cluster the data points of the collected electricity consumption monitoring data.

9. The method for monitoring safe electricity use of electric energy meters based on the Internet of Things according to any one of claims 1 to 7, characterized in that: The collection frequency for collecting electricity consumption monitoring data from the electric energy meter is 50HZ.

10. An electric energy meter safety electricity consumption monitoring system based on the Internet of Things, comprising a processor and a memory, wherein the memory stores computer program instructions, characterized in that: When the computer program instructions are executed by the processor, the method for monitoring safe electricity use of an electric energy meter based on the Internet of Things as described in any one of claims 1 to 9 is implemented.

Images