Electric energy data real-time monitoring system of factory power distribution cabinet

Abstract

The invention relates to the technical field of electric energy data monitoring, in particular to an electric energy data real-time monitoring system of a factory power distribution cabinet. The system comprises a data preprocessing module which collects the current and voltage of the power distribution cabinet at each moment, and the temperature and humidity in the power distribution cabinet; the current and the voltage at one moment form a data point, and different data segments are obtained; the environment similarity acquisition module is used for aligning the initial moments of the data segments corresponding to the production stage to which the current moment belongs and then acquiring all data points with the same position to form a data point set; acquiring the environment similarity of each data point in the data point set; the clustering weight acquisition module is used for acquiring the normal possibility and the clustering weight of each data point in the data point set; the monitoring module obtains a current normal range and a voltage normal range at the current moment; and monitoring the voltage and the current at the current moment by using the current normal range and the voltage normal range at the current moment. According to the invention, the electric energy data of the factory can be effectively monitored.

Description

Technical Field

[0001] This invention relates to the field of power data monitoring technology, specifically to a real-time power data monitoring system for factory power distribution cabinets. Background Technology

[0002] Factory power distribution cabinets are core devices in a factory's electrical system, used to distribute electrical energy from the upstream power source to various production equipment and to monitor, control, and protect the circuits. Their operational stability directly affects the continuity and safety of the factory's production equipment. Therefore, real-time monitoring of electrical data such as voltage and current in the distribution cabinet and identifying anomalies are crucial means to ensure the safety and stability of factory production.

[0003] Currently, the main method for monitoring power distribution cabinet data is the fixed threshold monitoring method. That is, when the monitored data exceeds the set threshold or the prediction result based on time-series data exceeds the set threshold, an anomaly warning is issued. However, the load of the factory power distribution system changes with the production stage, and the fixed threshold is difficult to adapt to the actual operating conditions. Moreover, the power data is also affected by changes in the factory environment and will drift to a certain extent. The fixed threshold cannot maintain high accuracy when the data distribution changes. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide a real-time monitoring system for power data of factory distribution cabinets. The specific technical solution adopted is as follows: One embodiment of the present invention provides a real-time monitoring system for power data of a factory power distribution cabinet, the system comprising: The data preprocessing module is used to collect the current, voltage, temperature, and humidity inside the power distribution cabinet at various times; the current and voltage at a given moment constitute a data point; the sequence of data points is segmented according to different production stages of the product to obtain different data segments; The environment similarity acquisition module is used to align the initial times of each data segment corresponding to the current production stage and obtain all data points with the same position to form a data point set for the current production stage; it obtains the environmental similarity of each data point in the data point set based on the differences in temperature and humidity between each data point in the data point set and the data point at the current time; it performs cluster analysis on each data point in the data point set of each production stage and obtains the environmental similarity of each data point by combining the environmental similarity of each data point; The clustering weight acquisition module is used to obtain the normal probability of a data point by using the difference in current and voltage between a data point in the data point set of the production stage at the current moment and other data points in the same data segment in the same data segment as the data point; and multiplies the normal probability of a data point by the environmental similarity to obtain the clustering weight of the data point. The monitoring module is used to cluster the current and voltage of each data point in the data point set based on the clustering weights, current and voltage of each data point in the data point set of the production stage at the current moment to obtain the normal range of current and voltage at the current moment; and to monitor the voltage and current at the current moment using the normal range of current and voltage at the current moment.

[0005] Preferably, the environmental similarity of each data point in the data point set is obtained based on the differences in temperature and humidity between each data point in the data point set and the data point at the current time, including: The temperature proximity of a data point is obtained by adding a hyperparameter to the absolute value of the difference between the temperature of a data point and the temperature of the current data point, and then calculating the inverse. Similarly, the humidity proximity of a data point is obtained by adding a hyperparameter to the absolute value of the difference between the humidity of a data point and the humidity of the current data point, and then calculating the inverse. Finally, the environmental proximity of a data point is obtained by multiplying the normalized value of the humidity proximity by the normalized value of the temperature proximity.

[0006] Preferably, cluster analysis is performed on each data point in the data point set of each production stage, and the environmental similarity of each data point is obtained by combining the environmental proximity of each data point, including: Based on the current and voltage corresponding to each data point in a production stage's data point set, clustering is performed to obtain different data point clusters for that production stage. The maximum and minimum environmental proximity values ​​of data points within each data point cluster for each production stage are obtained as suspected environmental similarity level segmentation values. Each suspected environmental similarity level segmentation value is mapped onto a one-dimensional coordinate axis, and k-means clustering is used to cluster the points on the coordinate axis, obtaining different suspected environmental similarity level segmentation value clusters. The mean of the suspected environmental similarity level segmentation values ​​in each cluster is calculated as the actual segmentation value for each cluster. The actual segmentation values ​​of each cluster are mapped onto a one-dimensional coordinate axis, with the interval between every two adjacent actual segmentation values ​​on the one-dimensional coordinate axis forming a level interval. The environmental proximity of each data point is projected into its corresponding level interval, and the lower limit of the level interval to which the environmental proximity of each data point belongs is the environmental similarity of that data point.

[0007] Preferably, the probability of a data point being considered normal is obtained by utilizing the difference in current and voltage between a data point in the data point set of the current production stage and other data points in the same data segment within the same data point set, including: Record any data point in the data point set of the current production stage as the target data point; obtain data points in other data segments of the data point set of the current production stage that are in the same position as the target data point to form a set of data points with the same position as the target data point; obtain the environmental proximity of each data point in the set of data points with the same position to the target data point based on the differences in temperature and humidity between each data point in the set of data points with the same position and the target data point; multiply the absolute value of the difference in current between a data point in the set of data points with the same position and the target data point by the environmental proximity of that data point in the set of data points with the same position to obtain the result of the data points with the same position. The weighted current difference of the data point in the data point set is calculated; the number of data points in the data point set with the same location is used as the numerator, and the sum of the weighted current differences of each data point in the data point set with the sum of the hyperparameters is used as the denominator, and the result is compared to obtain the current characteristic term of the target data point; similarly, the same operation is performed on the voltage of the target data point and the voltage of each data point in the data point set with the same location to obtain the voltage characteristic term of the target data point; the current characteristic term and the voltage characteristic term of the target data point are normalized and then multiplied to obtain the normal probability of the target data point; thus, the normal probability of each data point in the data point set of the production stage to which the current moment belongs is obtained.

[0008] Preferably, based on the clustering weights, current, and voltage of each data point in the data point set belonging to the current production stage, clustering is performed on the current and voltage of each data point in the data point set to obtain the normal range of current and voltage at the current time, including: The clustering weight of each data point in the data point set of the current production stage is multiplied by the initial neighborhood radius of the current of each data point in the data point set during the clustering process to obtain the corrected neighborhood radius of the current of each data point. Using the corrected neighborhood radius of the current of each data point, the current of each data point in the data point set of the current production stage is clustered to obtain different current clusters. The clustering weight of each data point in the data point set of the current production stage is multiplied by the initial neighborhood radius of the voltage of each data point in the data point set during the clustering process to obtain the corrected neighborhood radius of the voltage of each data point. Using the corrected neighborhood radius of the voltage of each data point, the voltage of each data point in the data point set of the current production stage is clustered to obtain different voltage clusters. The normal current range and normal voltage range at the current moment are obtained based on the different current clusters and voltage clusters acquired.

[0009] Preferably, the normal current range and normal voltage range at the current moment are obtained based on the different current clusters and voltage clusters acquired, including: The minimum and maximum values ​​of the current in the current cluster with the most elements in different current clusters are used to form the normal current range at the current moment; the minimum and maximum values ​​of the voltage in the voltage cluster with the most elements in different voltage clusters are used to form the normal voltage range at the current moment.

[0010] Preferably, monitoring the voltage and current at the current moment using the current normal range and voltage normal range includes: If the current at the current moment is not within the normal range or the voltage at the current moment is not within the normal range, an abnormality warning will be issued.

[0011] The embodiments of the present invention have at least the following beneficial effects: This application collects the current and voltage of the distribution cabinet at various times to form data points for each time moment, and obtains the temperature and humidity inside the distribution cabinet. Then, according to different production stages of the product, the sequence of data points is segmented to obtain different data segments, and data analysis is performed within the same production stage, reducing the impact of different production stages on the accuracy of the analysis. Furthermore, the environmental similarity of each data point in the data point set is obtained based on the differences in temperature and humidity between each data point in the data point set and the data point at the current time moment. Cluster analysis is performed on each data point in the data point set of each production stage, and the environmental similarity of each data point is considered in conjunction with the analysis. The system obtains the environmental similarity of each data point; then it obtains the normal probability of each data point in the data point set, and multiplies the normal probability of a data point by the environmental similarity to obtain the clustering weight of that data point. Combining environmental information and the information of the power data itself, the system obtains the clustering weight when clustering a data point, making it easier for historical data that better matches the normal fluctuation characteristics of the current power data to be clustered into the same cluster. Based on the clustering results, the normal range of current and voltage is determined. By using the normal range of current and voltage at the current moment to monitor the voltage and current at the current moment, the system can effectively improve the accuracy of anomaly monitoring in different production stages and environmental conditions of the power data monitoring system. Attached Figure Description

[0012] To more clearly illustrate the technical solutions and advantages 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.

[0013] Figure 1 This is a system block diagram of a real-time monitoring system for power data of a factory power distribution cabinet, provided as an embodiment of the present invention. Detailed Implementation

[0014] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a real-time power data monitoring system for a factory distribution cabinet according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0015] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0016] The following description, in conjunction with the accompanying drawings, details the specific scheme of a real-time monitoring system for power data of a factory distribution cabinet provided by the present invention.

[0017] Example: The main application scenario of this invention is as follows: This application analyzes the characteristics of changes in historical power data of enterprises, combines the environmental conditions of historical monitoring data, analyzes the degree to which each historical data reflects the current normal power data range, and uses this as a weight to analyze the current normal range of power data, thereby realizing dynamic adjustment of abnormal thresholds and monitoring the power data of enterprises.

[0018] Please see Figure 1 The diagram illustrates a system block diagram of a real-time power data monitoring system for a factory power distribution cabinet according to an embodiment of the present invention. The system includes the following modules: The data preprocessing module is used to collect the current, voltage, temperature, and humidity inside the power distribution cabinet at various times. The current and voltage at a given moment constitute a data point. The sequence of data points is segmented according to different production stages of the product to obtain different data segments.

[0019] This application requires the installation of power detection instruments inside the distribution cabinet to monitor the current and voltage data of the distribution cabinet. Since temperature and humidity are the main environmental factors affecting the monitoring data, temperature and humidity sensors are installed on the top of the distribution cabinet to monitor the temperature and humidity inside the distribution cabinet, respectively.

[0020] This allows for the collection of current, voltage, temperature, and humidity data within the distribution cabinet at various times, with the data collection frequency set to once per second. To ensure sufficient historical data for analyzing normal power fluctuations, the voltage, current, temperature, and humidity data collected over the previous 30 days are used as historical data for analysis. The implementer can determine the appropriate historical data period for subsequent analysis based on actual circumstances. Because the collected voltage, current, temperature, and humidity values ​​vary considerably and have dimensions, all types of data must be normalized for ease of analysis and comparison. The data used in the following analysis are all normalized data.

[0021] Furthermore, by combining voltage and current at a given moment to form a data point for that moment, multiple data points can be obtained. Simultaneously, product production is divided into different stages; for example, food processing includes stages such as ingredient cleaning, sorting, raw material cooking, and sterilization. These stages have specific operating cycles. To avoid the impact of differences in the variation patterns of electrical energy data across different production stages on normal data fluctuation analysis, the sequence of data points is segmented according to the different production stages of the product, obtaining different data segments. One data segment represents one production stage, and one production stage corresponds to multiple data segments. Additionally, when obtaining data segments corresponding to production stages, incomplete data segments corresponding to those stages need to be discarded to prevent them from affecting subsequent analysis.

[0022] The environment similarity acquisition module is used to align the initial times of each data segment corresponding to the current production stage and obtain all data points with the same position to form a data point set for the current production stage; to obtain the environmental similarity of each data point in the data point set based on the differences in temperature and humidity between each data point in the data point set and the data point at the current time; and to obtain the environmental similarity of each data point by performing cluster analysis on each data point in the data point set of each production stage and combining the environmental similarity of each data point.

[0023] Because changes in ambient temperature and humidity may cause a certain degree of drift in electrical energy data (voltage and current data), there may be differences in electrical energy data at the same production stage. In order to obtain the fluctuation pattern of electrical energy data under the current environmental conditions more accurately, it is necessary to improve the influence of data with higher environmental similarity on the analysis results. Therefore, it is necessary to calculate the environmental similarity between each historical data and the current moment.

[0024] First, determine the current production stage, which can be directly obtained based on the product's current production stage. Then, align the initial times of each data segment corresponding to the current production stage to obtain all data points at the same position within each data segment, forming a data point set. Since the lengths of different historical data segments corresponding to the same production stage may vary, to avoid randomness due to limited data volume, this data point set only analyzes data points that have corresponding positions in all data segments. For example, if there are two data segments corresponding to the current production stage, one with data points a1, b1, c1, d1, and the other with data points a2, b2, c2, then aligning a1 and a2 will result in data points at the same positions: a1 and a2, b1 and b2, c1 and c2. Therefore, the resulting data point set contains data points a1, a2, b1, b2, c1, and c2. The same operation can be performed on historical data to obtain the data point set for each production stage.

[0025] Since electrical energy data is relatively insensitive to temperature and humidity, meaning that within a certain range of temperature and humidity changes, environmental fluctuations have a generally consistent impact on electrical energy data, in order to improve the correlation between the calculated environmental similarity and the impact of the environment on electrical energy data, it is necessary to divide temperature and humidity into multiple levels and use this to calculate the similarity between historical data points and the current environment.

[0026] The degree of environmental similarity between each data point in the data point set and the data point at the current moment is obtained by comparing the temperature and humidity of each data point in the data point set.

[0027] Specifically, the temperature proximity of a data point is obtained by adding a hyperparameter to the absolute value of the difference between the temperature of a data point and the temperature of the data point at the current moment, and then calculating the inverse. Similarly, the humidity proximity of a data point is obtained by adding a hyperparameter to the absolute value of the difference between the humidity of a data point and the humidity of the data point at the current moment, and then calculating the inverse. Finally, the environmental proximity of a data point is obtained by multiplying the normalized value of the humidity proximity by the normalized value of the temperature proximity.

[0028] The specific calculation model for the environmental proximity of a data point in a data point set is as follows: , in, Let represent the environmental proximity of the i-th data point in the data point set, indicating how closely it resembles the environmental data corresponding to the data point at the current time. This represents the temperature corresponding to the i-th data point in the data set. This indicates the temperature corresponding to the data point at the current moment. This represents the absolute value of the difference between the temperature corresponding to the i-th data point in the data set and the temperature corresponding to the data point at the current moment. The smaller the absolute value, the closer the temperatures are. This represents the humidity corresponding to the i-th data point in the data point set. This indicates the humidity corresponding to the data point at the current moment. This represents the absolute value of the difference between the humidity corresponding to the i-th data point in the data point set and the humidity corresponding to the data point at the current moment; ε is a hyperparameter used to ensure that the fraction is meaningful, and here ε=0.01 is specified; norm represents the normalization function. This allows us to obtain the environmental proximity of each data point in the data point set for each production stage.

[0029] Next, cluster analysis was performed on each data point in the data point set of each production stage, and the environmental similarity of each data point was obtained by combining the environmental proximity of each data point.

[0030] Specifically, since the electrical energy data corresponding to data points with similar environmental influences should be highly consistent, the k-means clustering method is used to cluster the data points in the data point set of each production stage according to the current and voltage data at the corresponding time, thereby obtaining different data point clusters corresponding to each production stage. In other words, the data points in the data point set of each production stage are clustered separately. Among them, the elbow method is used to determine the k value, and the distance between data points is calculated based on the two-dimensional data (voltage and current) corresponding to two data points.

[0031] Furthermore, the maximum and minimum values ​​of environmental proximity of data points in each data point cluster corresponding to each production stage are obtained as suspected environmental similarity level segmentation values. Each suspected environmental similarity level segmentation value is mapped onto a one-dimensional coordinate axis, and k-means clustering is used to cluster the points on the coordinate axis to obtain different suspected environmental similarity level segmentation value clusters. The mean of the suspected environmental similarity level segmentation values ​​in each suspected environmental similarity level segmentation value cluster is calculated as the actual segmentation value of each cluster. The actual segmentation values ​​of each cluster are mapped onto a one-dimensional coordinate axis, and the interval between every two adjacent actual segmentation values ​​on the one-dimensional coordinate axis is a level interval. The environmental proximity of each data point is projected into its corresponding level interval, and the lower limit of the level interval to which the environmental proximity of each data point belongs is the environmental similarity of that data point.

[0032] The influence of environmental data on power data is basically the same within each level interval. Therefore, the minimum value of environmental similarity within each level interval (the lower limit of each level interval) is used as the similarity between environmental data within each level interval and the current environment.

[0033] The clustering weight acquisition module is used to obtain the normal probability of a data point by using the difference in current and voltage between a data point in the data point set of the production stage at the current moment and other data points in the same data segment in the same data segment as the data point; the normal probability of a data point is multiplied by the environmental similarity to obtain the clustering weight of the data point.

[0034] Since historical data may contain abnormal data such as noise caused by factors such as unstable circuits, electromagnetic interference, and poor sensor contact, in order to reduce the impact of these abnormal data on the analysis results of the fluctuation range of normal power data, it is also necessary to analyze the probability of abnormality for each historical data and reduce the influence weight of historical data with a higher probability of abnormality.

[0035] Because data fluctuation patterns may differ across production stages, the analysis of the current normal power data range is based solely on historical data from the same production stage. Therefore, a set of data points from the same production stage as the current moment is obtained. The probability of a data point being considered normal is calculated based on the degree of difference between this data point and other data points in the same data set that occupy the same position within each data segment. This allows for the calculation of the normal probability of all data points in the same production stage as the current moment. To reduce interference from environmental factors, the method for calculating environmental proximity is used to calculate the environmental proximity between a data point in the same production stage at the current moment and another data point in the same data set that occupy the same position within each data segment. This proximity is then used as a weight to calculate the normal probability.

[0036] Therefore, the normal probability of a data point can be obtained by using the difference in current and voltage between a data point in the data point set of the current production stage and other data points in the same data segment in the same data segment.

[0037] Specifically, any data point in the data point set of the current production stage is designated as the target data point; data points in other data segments that occupy the same position as the target data point in the data point set of the current production stage are collected to form a set of data points with the same position as the target data point; the environmental proximity of each data point in the set of data points with the same position is obtained based on the differences in temperature and humidity between the data points in the set of data points with the same position and the target data point; the absolute value of the difference in current between a data point in the set of data points with the same position and the target data point is multiplied by the environmental proximity of that data point in the set of data points with the same position to obtain the position. The weighted current difference of the data point in the same data point set is used as the numerator, and the sum of the weighted current differences of each data point in the same data point set is used as the denominator, and the result is compared to obtain the current characteristic term of the target data point. Similarly, the voltage of the target data point and the voltage of each data point in the same data point set are processed in the same way to obtain the voltage characteristic term of the target data point. The current characteristic term and the voltage characteristic term of the target data point are normalized and then multiplied to obtain the normal probability of the target data point. Thus, the normal probability of each data point in the data point set of the production stage to which the current moment belongs is obtained.

[0038] The specific calculation model for the normal probability of a data point in the data point set of the current production stage is as follows: , in, This represents the normal probability of the i-th data point (target data point) in the data point set belonging to the current production stage. This represents the total number of data points in the set where the i-th data point is at the same position. This represents the current at the i-th historical data point. Let r be the current of the r-th data point in the set of data points with the same position as the i-th data point. Let be the absolute value of the difference in current between the i-th data point and the r-th data point in the set of data points at the same position as the i-th data point. This represents the absolute difference in current between data points at the same position in different data segments. The smaller this absolute value, the more likely the current at the i-th data point is to be normal. Let represent the environmental proximity of the i-th data point to the r-th data point in the set of data points with the same location as the i-th data point. This indicates the similarity of the environment between the i-th data point and the r-th data point in the set of data points with the same location as the i-th data point. The method for obtaining this value is the same as the method for obtaining the environmental proximity described above. The weighted current difference of the r-th data point in the set of data points with the same position as the i-th data point. Let be the current characteristic term of the i-th data point, representing the probability that the current at the i-th data point is normal; This represents the voltage of the i-th data point in the data point set belonging to the current production stage. This represents the voltage of the i-th data point in the data point set corresponding to the current production stage, and the voltage is analyzed in the same way as the current. The weighted voltage difference of the r-th data point in a set of data points with the same position as the i-th data point. Let be the voltage characteristic term for the i-th data point. Since voltage and current changes are correlated, the data can only be considered normal at a given moment when both voltage and current are normal. ε is a hyperparameter used to prevent the denominator from being zero and to ensure that the fraction is meaningful; here, ε is specified as 0.01. norm represents the normalization function.

[0039] Data points from the current production stage that exhibit similar environmental conditions and normal values ​​are more likely to reflect the normal fluctuation range of electricity data at the current moment. Therefore, by combining the environmental similarity and normal probability of each data point in the current production stage data set, the influence weight of each data point on the determination of the normal threshold is calculated. Thus, in the current production stage data set, the cluster weight of a data point is obtained by multiplying its normal probability by its environmental similarity. The specific calculation model is as follows: , in, Let be the cluster weight of the i-th data point in the data point set belonging to the current production stage, representing its influence weight on the determination of the normal threshold. Let represent the environmental similarity between the i-th data point in the data point set belonging to the current production stage, indicating the similarity between this data point and the environment at the current moment. This represents the normal probability of the i-th data point in the data point set belonging to the current production stage.

[0040] This allows us to obtain the cluster weights of each data point in the data point set belonging to the current production stage.

[0041] The monitoring module is used to cluster the current and voltage of each data point in the data point set based on the clustering weights, current and voltage of each data point in the data point set of the production stage at the current moment to obtain the normal range of current and voltage at the current moment; and to monitor the voltage and current at the current moment using the normal range of current and voltage at the current moment.

[0042] Since each data point in the data set of the current production stage includes two features—current and voltage—it is necessary to process current and voltage separately. Specifically, taking current as an example, to make the current of data points with smaller weights in the data set of the current production stage less likely to be clustered, the initial neighborhood radius corresponding to the current of each data point is adjusted during the clustering process according to the clustering weight of each data point, thereby achieving clustering.

[0043] Specifically, the clustering weight of each data point in the data set belonging to the current production stage at the current moment is multiplied by the initial neighborhood radius of the current of each data point in the clustering process to obtain the corrected neighborhood radius of the current of each data point. Using the corrected neighborhood radius of the current of each data point, the current of each data point in the data set belonging to the current production stage at the current moment is clustered to obtain different current clusters. The clustering weight of each data point in the data set belonging to the current production stage at the current moment is multiplied by the initial neighborhood radius of the voltage of each data point in the clustering process to obtain the corrected neighborhood radius of the voltage of each data point. Using the corrected neighborhood radius of the voltage of each data point, the voltage of each data point in the data set belonging to the current production stage at the current moment is clustered to obtain different voltage clusters. The initial neighborhood radius can be obtained using a k-distance curve.

[0044] Then, based on the different current clusters and voltage clusters obtained, the normal current range and normal voltage range at the current moment are obtained.

[0045] Specifically, the minimum and maximum values ​​of the currents in the current cluster with the most elements in different current clusters are obtained to form the normal current range at the current moment, denoted as . , and These are the minimum and maximum current values ​​in the current cluster with the most elements, respectively; the minimum and maximum voltage values ​​in the voltage cluster with the most elements in different voltage clusters are used to form the normal voltage range at the current moment, denoted as . , and These are the minimum and maximum voltage values ​​in the voltage cluster with the most elements, respectively.

[0046] After obtaining the current and voltage normal ranges at the current moment, the voltage and current at the current moment are monitored using these normal ranges. Specifically, if the current or voltage at the current moment is outside the normal range, an abnormality warning is issued.

[0047] It should also be noted that since normal fluctuations within the same production stage should all fall within this range, threshold adjustments (normal current and voltage ranges) are only performed once at the beginning of each stage. In other words, a threshold adjustment is performed only at the start of each production stage. Furthermore, if environmental conditions change during a production stage, the applicability of the threshold may decrease. Therefore, it is necessary to compare the current environmental conditions with those at the time of the most recent threshold adjustment to determine whether a threshold adjustment is needed at the current moment. Specifically, if the environmental similarity between the current moment and the time of the most recent threshold adjustment is less than the lower limit of the largest level range, a threshold adjustment is required at the current moment. The largest level range is the rightmost range on the coordinate axis.

[0048] For example, time t is the initial moment of a production stage. After adjusting the thresholds (normal current range and normal voltage range) at time t, the environmental proximity between time t and time t+1 is obtained. If the environmental proximity is less than the lower limit of the maximum level range, the threshold needs to be updated at time t+1. The update method is the same as the method described above for obtaining the normal current and normal voltage ranges. If the environmental proximity is within the maximum level range, no threshold update is needed. Furthermore, if the threshold was updated at time t+1, determining whether an update is needed at time t+2 requires obtaining the environmental proximity between time t+1 and time t+2 to determine whether a threshold update is needed at time t+2. If the threshold is not updated at time t+1, then the environmental proximity between time t and time t+2 is needed to determine whether a threshold update is needed at time t+2. In other words, the current environmental conditions need to be compared with the environmental conditions at the time of the most recent threshold adjustment to determine whether a threshold adjustment is needed at the current moment. This eliminates environmental interference, obtains dynamically adaptive normal current and normal voltage ranges, and achieves accurate monitoring of power data.

[0049] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0050] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0051] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A real-time monitoring system for electrical energy data of a factory switchboard, characterized by, The system comprises: A data preprocessing module is configured to collect currents, voltages, temperatures and humidities of the power distribution cabinet at each time point; a data point is composed of the current and the voltage at one time point; different data segments are obtained by segmenting a sequence of data points according to different production stages of the product; An environment similarity obtaining module is configured to align initial time points of each data segment corresponding to the production stage to which the current time point belongs, to obtain a data point set of the production stage to which the current time point belongs, and to obtain environment proximities of data points in the data point set according to differences between temperatures and humidities corresponding to the data points in the data point set and a data point of the current time point; to perform clustering analysis on each data point in the data point set of each production stage, and to obtain environment similarities of the data points in combination with the environment proximities of the data points; A clustering weight obtaining module is configured to obtain a normal possibility of a data point in the data point set of the production stage to which the current time point belongs, by using differences between currents and voltages of the data point and data points in other data segments having the same position as the data point in the data segment; and to obtain a clustering weight of the data point by multiplying the normal possibility of the data point and the environment similarity of the data point; A monitoring module is configured to perform clustering on currents and voltages of each data point in the data point set of the production stage to which the current time point belongs, to obtain a normal range of the current and a normal range of the voltage of the current time point, based on the clustering weights, the currents and the voltages of the data points in the data point set; and to monitor the voltage and the current of the current time point by using the normal range of the current and the normal range of the voltage of the current time point.

2. A real time monitoring system for electrical energy data of a factory switchboard as claimed in claim 1 wherein, The environment proximities of the data points in the data point set are obtained according to differences between temperatures and humidities corresponding to the data points in the data point set and a data point of the current time point, and the environment proximities of the data points in the data point set are obtained by: Adding a hyperparameter to an absolute value of a difference between the temperature corresponding to a data point in the data point set and the temperature corresponding to the data point of the current time point, and taking an inverse to obtain a temperature proximity of the data point; adding a hyperparameter to an absolute value of a difference between the humidity corresponding to the data point and the humidity corresponding to the data point of the current time point, and taking an inverse to obtain a humidity proximity of the data point; and multiplying a normalized value of the humidity proximity of the data point and a normalized value of the temperature proximity of the data point to obtain the environment proximity of the data point.

3. A real time monitoring system for electrical energy data of a factory switchboard according to claim 1 characterized in that, The environment similarities of the data points in the data point set of each production stage are obtained by performing clustering analysis on each data point in the data point set of each production stage and combining the environment proximities of the data points, and the environment similarities of the data points are obtained by: The current and voltage corresponding to each data point in a data point set of a production stage are used to cluster the data points in the data point set of the production stage to obtain different data point clusters corresponding to the production stage; the maximum and minimum of the environmental proximity of the data points in each data point cluster in all data point clusters corresponding to each production stage are obtained as suspected environmental similarity level segmentation values; the suspected environmental similarity level segmentation values are respectively mapped to a one-dimensional coordinate axis, and the points on the coordinate axis are clustered by using k-means clustering to obtain different suspected environmental similarity level segmentation value clusters; the mean of the suspected environmental similarity level segmentation values in each suspected environmental similarity level segmentation value cluster is obtained as the actual segmentation value of each suspected environmental similarity level segmentation value cluster; the actual segmentation values of the suspected environmental similarity level segmentation value clusters are mapped to a one-dimensional coordinate axis, and the interval between each two adjacent actual segmentation values on the one-dimensional coordinate axis is a level interval; the environmental proximity of each data point is projected into the level interval to which the data point belongs, and the lower limit value of the level interval to which the environmental proximity of each data point belongs is the environmental similarity of the data point.

4. A real time monitoring system for electrical energy data of a factory switchboard according to claim 1 characterized in that, The normal possibility of a data point in a data point set of a production stage at a current time is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of the data point and the data point in the same position in the data segment in the data point set, and the normal possibility of the data point is obtained by using the difference between the current and voltage of ​ 5. A real time monitoring system for electrical energy data of a factory switchboard as claimed in claim 1 wherein, ​ The cluster weight of each data point in the data point set of the production stage to which the current time belongs is multiplied by the initial neighborhood radius of the current of each data point in the data point set in the clustering process to obtain the corrected neighborhood radius of the current of each data point; the currents of each data point in the data point set of the production stage to which the current time belongs are clustered by using the corrected neighborhood radius of the current of each data point to obtain different current clusters; the cluster weight of each data point in the data point set of the production stage to which the current time belongs is multiplied by the initial neighborhood radius of the voltage of each data point in the data point set in the clustering process to obtain the corrected neighborhood radius of the voltage of each data point; the voltages of each data point in the data point set of the production stage to which the current time belongs are clustered by using the corrected neighborhood radius of the voltage of each data point to obtain different voltage clusters; The current normal range and the voltage normal range of the current time are obtained according to the obtained different current clusters and voltage clusters.

6. A real time monitoring system for electrical energy data of a factory switchboard according to claim 5, characterized in that, The current normal range and the voltage normal range of the current time are obtained according to the obtained different current clusters and voltage clusters, including: The minimum value and the maximum value of the current in the current cluster with the most elements in the different current clusters are obtained to constitute the current normal range of the current time; the minimum value and the maximum value of the voltage in the voltage cluster with the most elements in the different voltage clusters are obtained to constitute the voltage normal range of the current time.

7. A real time monitoring system for electrical energy data of a factory switchboard as claimed in claim 1 wherein, The current and the voltage of the current time are monitored by using the current normal range and the voltage normal range of the current time, including: If the current of the current time is not in the current normal range of the current time or the voltage of the current time is not in the voltage normal range of the current time, an abnormal warning is performed.

Images