A method and apparatus for electroanalysis

Abstract

The application relates to the electric power technical field and provides a power consumption analysis method and device, which comprises the following steps: determining the first similarity between the energy consumption data of an object and the energy consumption data of each sub-object corresponding to the object in any power supply mode of an energy storage system used by the object; selecting a sub-object with a first similarity less than a first threshold value from all the sub-objects corresponding to the object; determining the second similarity between the energy consumption data of a preset key energy consumption equipment corresponding to the selected sub-object and the energy consumption data of the object, selecting a preset key energy consumption equipment with a second similarity less than a second threshold value from the preset key energy consumption equipment, and determining the average minimum power of the selected preset key energy consumption equipment in a minimum valley period, so that the average minimum power provides an effective basis for the new distributed energy of the object, and the power consumption strategy corresponding to the power supply mode is determined according to the average minimum power, thereby reducing the energy cost consumed by the object in production.

Description

Technical Field

[0001] The present application relates to the field of electric power technology, and in particular to a method and device for analyzing electricity consumption. Background Art

[0002] In recent years, with the rapid development of big data technology and the increasing popularity of the Industrial Internet of Things (IIoT), data mining technology has gradually been applied to the traditional energy industry. The rise of the Energy Internet has also promoted the intelligent development of various elements in the energy network, enabling enterprises to gradually shift their energy management from the previous extensive management to refined management. To achieve refined management, enterprises must have an accurate and in-depth understanding of their current energy usage behaviors.

[0003] Currently, many high-energy-consuming enterprises have implemented energy consumption data collection and accumulated a wealth of data during production and operation, laying a solid foundation for energy usage analysis. In the smart grid environment, the development and construction of smart terminals, power communication technologies, and advanced measurement technologies have enabled more comprehensive electricity usage data for many enterprises. Furthermore, electricity consumption is the primary energy consumption mode for most enterprises and accounts for the largest proportion of their energy expenses. Therefore, analyzing enterprise electricity usage is crucial. Understanding the characteristics of enterprise electricity usage can help regulate their electricity usage behavior, develop appropriate electricity usage strategies, enhance demand-supply interaction with the grid, and reduce energy costs. However, most studies analyzing enterprise electricity usage have focused on the grid's perspective, analyzing the electricity usage characteristics of large enterprise customers. However, the electricity usage characteristics of large enterprise customers are not representative of those of smaller enterprises. Consequently, electricity usage strategies based on these characteristics are not suitable for the electricity needs of smaller enterprises. Summary of the Invention

[0004] The embodiments of the present application provide a power consumption analysis method and apparatus for formulating a corresponding power consumption strategy based on the power consumption characteristics of a certain object obtained through analysis.

[0005] In a first aspect, an embodiment of the present application provides a method for analyzing electricity consumption, comprising:

[0006] For any power supply mode, determining a first similarity between energy usage data of an object under the power supply mode and energy usage data of each sub-object corresponding to the object, wherein the power supply mode is a power supply mode of an energy storage system used by the object;

[0007] Selecting a sub-object whose first similarity is less than a first threshold from all sub-objects corresponding to the object;

[0008] Determining a second similarity between energy usage data of a preset key energy-consuming device corresponding to the selected sub-object and the energy usage data of the object, and selecting a preset key energy-consuming device whose second similarity is less than a second threshold from the preset key energy-consuming devices;

[0009] The power consumption strategy corresponding to the power supply mode is determined according to the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period.

[0010] Optionally, the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period is determined by:

[0011] Determining the sum of the power of the selected preset key energy-consuming equipment during the minimum valley period of each time period within the clustering period, wherein the minimum valley period is the minimum valley period obtained after clustering the energy consumption data of the object in each time period, and the clustering period includes at least one time period;

[0012] Determine the average power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods according to the sum of the powers of the selected preset key energy-consuming equipment during the minimum valley period of all time periods;

[0013] The determined average power is used as the average minimum power of the selected preset key energy-consuming equipment within the clustering period.

[0014] Optionally, the first similarity is determined by:

[0015] determining, for any sub-object among all sub-objects corresponding to the object, multiple Euclidean distances between multiple energy usage data of the object and multiple energy usage data of the sub-object under the power supply mode, and determining the shortest Euclidean distance as a first similarity between the energy usage data of the object and the energy usage data of the sub-object;

[0016] The second similarity is determined by:

[0017] For any preset key energy-consuming device among the preset key energy-consuming devices corresponding to any selected sub-object, determine multiple Euclidean distances between multiple energy consumption data of the preset key energy-consuming device and multiple energy consumption data of the object, and determine the shortest Euclidean distance as the second similarity between the energy consumption data of the preset key energy-consuming device and the energy consumption data of the object.

[0018] Optionally, the method further includes:

[0019] Eliminating abnormal data from the energy usage data of the object and the energy usage data of sub-objects corresponding to the object collected during the clustering period;

[0020] According to the set summary period, the collected energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are summarized respectively, and each summary period includes multiple collection periods;

[0021] Normalizing the aggregated energy usage data of the object and the energy usage data of the sub-objects corresponding to the object;

[0022] Based on the trained clustering model, the normalized energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are clustered to obtain the clustered energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object.

[0023] Optionally, the clustering model is trained in the following manner:

[0024] Traverse the preset minimum number of clusters to the maximum number of clusters, and obtain the intra-cluster error variance and silhouette coefficient corresponding to each number of clusters;

[0025] The number of clusters with the inflection point of all intra-cluster error variances and the largest silhouette coefficient is selected as the number of clusters of the clustering model.

[0026] In a second aspect, the present application provides a power consumption analysis device, comprising:

[0027] a similarity determination module configured to determine, for any power supply mode, a first similarity between energy usage data of an object under the power supply mode and energy usage data of each sub-object corresponding to the object; and a second similarity between energy usage data of a preset key energy-consuming device corresponding to a selected sub-object and the energy usage data of the object, wherein the power supply mode is a power supply mode of an energy storage system used by the object;

[0028] A selection module is configured to select, from all sub-objects corresponding to the object, sub-objects whose first similarity is less than a first threshold; and select, from preset key energy-consuming devices, preset key energy-consuming devices whose second similarity is less than a second threshold;

[0029] The power usage strategy determination module is used to determine the power usage strategy corresponding to the power supply mode according to the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period.

[0030] Optionally, the device further includes a power determination module, configured to:

[0031] Determining the sum of the power of the selected preset key energy-consuming equipment during the minimum valley period of each time period within the clustering period, wherein the minimum valley period is the minimum valley period obtained after clustering the energy consumption data of the object in each time period, and the clustering period includes at least one time period;

[0032] Determine the average power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods according to the sum of the powers of the selected preset key energy-consuming equipment during the minimum valley period of all time periods;

[0033] The determined average power is used as the average minimum power of the selected preset key energy-consuming equipment within the clustering period.

[0034] Optionally, the similarity determination module is specifically configured to:

[0035] determining, for any sub-object among all sub-objects corresponding to the object, multiple Euclidean distances between multiple energy usage data of the object and multiple energy usage data of the sub-object under the power supply mode, and determining the shortest Euclidean distance as a first similarity between the energy usage data of the object and the energy usage data of the sub-object;

[0036] For any preset key energy-consuming device among the preset key energy-consuming devices corresponding to any selected sub-object, determine multiple Euclidean distances between multiple energy consumption data of the preset key energy-consuming device and multiple energy consumption data of the object, and determine the shortest Euclidean distance as the second similarity between the energy consumption data of the preset key energy-consuming device and the energy consumption data of the object.

[0037] Optionally, the device further includes a clustering module, configured to:

[0038] Eliminating abnormal data from the energy usage data of the object and the energy usage data of sub-objects corresponding to the object collected during the clustering period;

[0039] According to the set summary period, the collected energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are summarized respectively, and each summary period includes multiple collection periods;

[0040] Normalizing the aggregated energy usage data of the object and the energy usage data of the sub-objects corresponding to the object;

[0041] Based on the trained clustering model, the normalized energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are clustered to obtain the clustered energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object.

[0042] Optionally, the clustering model is trained in the following manner:

[0043] Traverse the preset minimum number of clusters to the maximum number of clusters, and obtain the intra-cluster error variance and silhouette coefficient corresponding to each number of clusters;

[0044] The number of clusters with the inflection point of all intra-cluster error variances and the largest silhouette coefficient is selected as the number of clusters of the clustering model.

[0045] In a third aspect, the present application provides an electricity consumption analysis device, comprising a memory and a processor;

[0046] The memory is used to store program instructions;

[0047] The processor is configured to execute the method as described in any one of the first aspects according to the program instructions stored in the memory.

[0048] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the method as described in any one of the first aspects.

[0049] In the above-mentioned embodiment of the present application, for any power supply mode of the energy storage system used by the object, the sub-objects whose first similarity under the power supply mode is less than the first threshold are first screened out, and the second similarity between the energy consumption data of the preset key energy-consuming equipment corresponding to the selected sub-object and the energy consumption data of the object is determined. Then, the preset key energy-consuming equipment whose second similarity is less than the second threshold is screened out, and the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period is determined. The average minimum power provides an effective basis for the object to newly create distributed energy, so that the power consumption strategy under the power supply mode determined according to the average minimum power can accurately adapt to the power consumption demand of the object. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0051] Figure 1 The following is a schematic diagram of an application scenario provided by an embodiment of the present application;

[0052] Figure 2 The flowchart of the method for cluster analysis of the electricity consumption characteristics of objects provided by the embodiment of the present application is exemplified;

[0053] Figure 3 The flowchart of the method for cluster analysis of the electricity consumption characteristics of an enterprise provided in an embodiment of the present application is exemplarily shown;

[0054] Figure 4 An example of an aggregated curve of electricity consumption data for a certain enterprise in April provided in an embodiment of the present application is shown;

[0055] Figure 5aAn example of an aggregated curve of energy consumption data of a certain enterprise in power supply mode in April provided by an embodiment of the present application is shown;

[0056] Figure 5b An example of an aggregated curve of energy consumption data of a certain enterprise in April under another power supply mode provided in an embodiment of the present application is shown;

[0057] Figure 6 An example of an aggregated curve of energy consumption data of a certain department in April provided in an embodiment of the present application is shown;

[0058] Figure 7 The aggregated curve of energy consumption data of various departments in April provided in the embodiment of the present application is exemplarily shown;

[0059] Figure 8 An example of an aggregate curve of energy consumption data of a preset key energy-consuming device in April provided in an embodiment of the present application is shown;

[0060] Figure 9 The flowchart of the method for formulating a power consumption strategy based on the clustered power load curve provided in an embodiment of the present application is exemplified;

[0061] Figure 10a The following is an exemplary diagram of determining the shortest Euclidean distance provided by an embodiment of the present application;

[0062] Figure 10b The following is an exemplary diagram of determining the shortest Euclidean distance provided by an embodiment of the present application;

[0063] Figure 11 The flowchart of the method for formulating a power consumption strategy for an enterprise based on the clustered power load curve provided in an embodiment of the present application is exemplified;

[0064] Figure 12 The functional structure diagram of a power consumption analysis method and device provided in an embodiment of the present application is exemplarily shown;

[0065] Figure 13 The hardware structure diagram of an electricity consumption analysis method device provided in an embodiment of the present application is exemplified. DETAILED DESCRIPTION

[0066] To make the objectives, technical solutions, and advantages of this application more clear, this application will be further described in detail below with reference to the accompanying drawings. It should be understood that the embodiments described herein are only some of the embodiments of this application, and not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of this application without inventive effort are intended to fall within the scope of protection of this application.

[0067] Based on the exemplary embodiments shown in this application, all other embodiments obtained by persons of ordinary skill in the art without inventive effort are within the scope of protection of this application. In addition, although the disclosure in this application is presented based on one or several exemplary examples, it should be understood that each aspect of the disclosure can independently constitute a complete technical solution.

[0068] In addition, the terms "comprises" and "comprising" and any variations thereof are intended to cover but not exclude inclusion, for example, a product or device comprising a list of components is not necessarily limited to those components expressly listed but may include other components not expressly listed or inherent to such product or device.

[0069] The term "module" as used in this application refers to any known or later developed hardware, software, firmware, artificial intelligence, fuzzy logic, or combination of hardware and / or software code that is capable of performing the functions associated with that element.

[0070] The term "electricity consumption data" used in this application refers to the grid-side electric energy data consumed by electrical equipment, which can be obtained from the incoming line meter.

[0071] The term "energy consumption data" used in this application refers to the energy consumption data of the actual energy consumed by electrical equipment, which can be obtained from energy consumption meters.

[0072] The term "object" used in this application can be a target transaction such as an enterprise, a cell, a city, or a community that needs to analyze electricity usage characteristics.

[0073] The term "sub-object" used in this application refers to a branch under an object. For example, when the object is an enterprise, the sub-object may be a department of the enterprise; when the object is a community, the sub-object may be a building in the community; when the object is a city, the sub-object may be the districts or counties of the city; when the object is a community, the sub-object may be the streets of the community, and so on.

[0074] The term "preset key energy-consuming equipment" used in this application can be user-defined or determined based on the energy consumed by the energy-consuming equipment, such as setting energy-consuming equipment with energy consumption greater than a set threshold as the preset key energy-consuming equipment.

[0075] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0076] Figure 1 The following is a schematic diagram of an application scenario provided by an embodiment of the present application. Figure 1As shown, the object 100 has a distributed energy system 101, which includes a photovoltaic system 101_1, an energy storage system 101_2, etc. The distributed energy system 101 supplies energy to the energy-consuming devices in the object. At the same time, the object can also be powered by the power grid 102; multiple sub-objects 103 are provided in the object, and sub-objects with the same power consumption characteristics can be combined into a sub-object class. For example, sub-object 1 and sub-object 2 belong to the same sub-object class; each sub-object is provided with one or more key energy-consuming devices 104. Different energy-consuming devices have the same or different working hours and consume the same or different amounts of energy. The key energy-consuming devices provided under different sub-objects 1 can be the same or different.

[0077] At present, most studies on corporate electricity consumption are conducted from the perspective of the power grid. However, the electricity consumption characteristics of large corporate customer groups cannot represent the electricity consumption characteristics of small-category enterprises. As a result, the electricity consumption strategies formulated based on the electricity consumption characteristics of large corporate customer groups are not suitable for the electricity demand of small-category enterprises. For example, the minimum electricity load required by large corporate customer groups in a day is greater than the maximum electricity load required by a certain small-category enterprise in a day. With the rapid development of the energy Internet, the energy consumption methods of various enterprises have gradually changed from extensive management to refined management. Figure 1 In the scenario shown, cluster analysis can be used to analyze the energy consumption of energy-consuming equipment within a company. Cluster analysis is a mathematical method that classifies objects according to certain requirements and patterns. It is an exploratory analysis method that does not require a predefined classification standard. Cluster analysis automatically classifies data based on sample data. Therefore, its characteristics are suitable for analyzing the characteristics of corporate electricity consumption.

[0078] Figure 2 The flowchart of the method for cluster analysis of the electricity consumption characteristics of objects provided in an embodiment of the present application is exemplified. The process can be implemented by software and mainly includes the following steps:

[0079] S201: Collecting energy usage data of an object and energy usage data of sub-objects corresponding to the object.

[0080] In this step, the energy consumption data of the object and its sub-objects can be collected from the historical energy consumption power data of the energy consumption meter of the energy-consuming equipment according to the set collection period. Optionally, the collection period is 5 minutes.

[0081] S202: Eliminate abnormal data from the energy usage data of the object and the energy usage data of the sub-objects corresponding to the object collected within the clustering period.

[0082] In this step, the collected energy consumption data is cleaned to remove null values, abnormal values ​​and other energy consumption data.

[0083] S203: Summarize the collected energy usage data of the object and the energy usage data of the sub-objects corresponding to the object according to the set summary period.

[0084] In this step, each aggregation cycle includes multiple collection cycles, and the aggregated energy usage data is the average of the energy usage data collected in each collection cycle. For example, if the energy usage data of an object is collected over a 15-minute aggregation cycle and includes three collection cycles, that is, there are three energy usage data in one aggregation cycle, denoted as m1, m2, and m3. The aggregated energy usage data for the object, m, is then m = (m1, m2, m3) / 3. Aggregating energy usage data reduces both the data processing load and the collection latency.

[0085] S204: Normalizing the aggregated energy usage data of the object and the energy usage data of the sub-objects corresponding to the object.

[0086] In this step, the normalization formula is:

[0087]

[0088] Among them, X std represents the normalized energy consumption data, ranging from 0 to 1, and X represents the original energy consumption data; X min represents the minimum value of the energy consumption data in each time period within the clustering period; X max Represents the maximum value of energy usage data for each time period within the clustering period. Optionally, the clustering period is 1 month and the time period is 1 day.

[0089] S205: Based on the trained clustering model, cluster the normalized energy usage data of the object and the energy usage data of the sub-objects corresponding to the object respectively to obtain the clustered energy usage data of the object and the energy usage data of the sub-objects corresponding to the object.

[0090] In an optional embodiment, the clustering model can use the K-mean algorithm, set the maximum number of clusters, and use the intra-cluster variance (Sum of Squares due to Error, SSE) and silhouette coefficient as the objective function to verify and evaluate the number of clusters to obtain the optimal number of clusters. Specifically, the preset minimum number of clusters to the maximum number of clusters are traversed, and the intra-cluster error variance and silhouette coefficient corresponding to each number of clusters are obtained respectively; the cluster number with the inflection point of all intra-cluster error variances and the largest silhouette coefficient is selected as the cluster number of the clustering model. The minimum number of clusters can be 1.

[0091] The following example uses an enterprise as an example. The analysis includes electricity and energy usage data for the enterprise, by department, and by key energy-consuming equipment on a monthly basis. The clustering period is one month, the time period is one day, the aggregation period is 15 minutes, and the collection period is 5 minutes.

[0092] Figure 3 The flowchart of the method for cluster analysis of electricity consumption characteristics within an enterprise provided by an embodiment of the present application is exemplified. The process mainly includes the following steps:

[0093] S301: According to a set collection cycle, the enterprise's electricity consumption data is obtained from the enterprise's incoming line meter, and the enterprise's energy consumption data is obtained from the energy consumption meters of all energy-consuming equipment.

[0094] In this step, the electricity consumption data collected from the enterprise's incoming line meter is the load data on the grid side, and the energy consumption data collected from the energy consumption meters of all energy-consuming equipment is the load data provided by the energy storage system.

[0095] S302: Eliminate null values ​​and abnormal values ​​in the electricity consumption data and energy consumption data of the enterprise.

[0096] Eliminating null values ​​and outliers in electricity and energy consumption data is conducive to improving the accuracy of cluster analysis.

[0097] S303: According to the set summary cycle, the electricity consumption data and energy consumption data of the enterprise are summarized respectively, and the energy consumption data of each department is summarized according to the department to which the energy-consuming equipment belongs, and the energy consumption data of the preset key energy-consuming equipment in each department is summarized.

[0098] In this step, the energy consumption data of an enterprise is the sum of the energy consumption data of all energy-consuming equipment in the enterprise, and the energy consumption data of a department is the sum of the energy consumption data of all energy-consuming equipment in the department.

[0099] S304: Normalizing the enterprise's electricity consumption data, the enterprise's energy consumption data, the energy consumption data of each department, and the energy consumption data of preset key energy-consuming equipment according to the set summary period.

[0100] In this step, the normalization processing method is as described in S204 and will not be repeated here.

[0101] S305: Based on the trained clustering model, cluster the normalized enterprise electricity consumption data, enterprise energy consumption data, energy consumption data of each department, and energy consumption data of preset key energy-consuming equipment.

[0102] In this step, the maximum number of clusters set in the clustering models for clustering different energy usage data may be the same or different, but the training method of the clustering models is the same, that is, the method of determining the optimal number of clusters is the same.

[0103] S306: Based on the clustering results, draw the enterprise's electricity load curve, the enterprise's energy load curve, the energy load data curves of each department, and the energy load curves of preset key energy-consuming equipment, and save the clustering results in the database.

[0104] The relevant descriptions of steps S301 to S306 refer to the above embodiments and will not be repeated here.

[0105] In an alternative implementation, an enterprise consists of multiple departments. The energy usage data for each department over a month can be clustered. The central curve with the most days in the clustering results is used as the department's energy load curve. This central curve can represent the department's most likely energy usage pattern for the month. After obtaining the energy load curves for each department, the energy usage data for each department is clustered to obtain inter-departmental energy load curves.

[0106] In an optional implementation, cluster analysis is performed on the electricity consumption characteristics of the enterprise at regular intervals based on the clustering results stored in the database. For example, cluster analysis is performed on the electricity consumption data and energy consumption data of the previous month on the 1st of each month.

[0107] In the above embodiment of the present application, a cluster analysis is performed on the monthly electricity consumption characteristics of the enterprise, and the collected electricity consumption data and energy consumption data of the enterprise are summarized according to the energy consumption data of the enterprise, each department, and the preset key energy-consuming equipment. After normalization, cluster analysis is performed again to obtain the monthly grid-side electricity load curve, energy-consuming side energy load curve, energy load curve of each department, electricity load curve between departments, and electricity characteristic curve of preset key energy-consuming equipment. This method makes up for the deficiency of existing enterprise electricity consumption characteristic analysis research that only analyzes large customer groups, is conducive to a detailed understanding of the electricity consumption characteristics within the enterprise, thereby providing strong support for the enterprise to participate in demand-side response, and providing an effective basis for the enterprise to build new distributed energy. At the same time, it can enable the enterprise to further optimize production scheduling from the perspective of reducing energy costs.

[0108] Taking the cluster analysis of a company's electricity consumption data in April as an example, the company's electricity consumption data is obtained from the grid side (i.e., 35kV incoming line), the null values ​​and abnormal values ​​in the electricity consumption data are removed, and normalization is performed. The normalized daily electricity consumption data is clustered and analyzed. The clustering results are as follows: Figure 4 As shown by Figure 4 It can be seen that the company's daily electricity load reaches its lowest value (minimum trough value) around 11:00, and another obvious electricity load trough occurs between 18:00 and 19:30. Figure 4It also shows the time periods of different electricity prices in a day, 0:00-6:00 is the low electricity price period, 6:00-8:00 is the flat electricity price period, 8:00-11:00 is the peak electricity price period, 11:00-18:00 is the flat electricity price period, 18:00-21:00 is the peak electricity price period, 21:00-22:00 is the flat electricity price period, and 22:00-24:00 is the low electricity price period.

[0109] Taking the energy consumption data of all electrical equipment of a certain enterprise in April as an example, we obtain the energy consumption data of the enterprise, remove the null values ​​and outliers in the energy consumption data, and perform normalization. We then perform cluster analysis on the normalized daily energy consumption data. The clustering results of the energy consumption data of all electrical equipment of the enterprise in April show two patterns. Figure 5a 、 Figure 5b , where the 13-day daily energy consumption data curves have similar trends and can be classified as mode 1, such as Figure 5a As shown in Figure 2, the trends of the other 13 days’ daily energy consumption data curves are similar and can be classified as Mode 2. Figure 5b As shown. Figure 5a As shown in Figure 1, mode 1 maintains a high load rate throughout the day and has multiple load troughs. The trough periods of mode 1 include: 4:15-5:30; 6:00-7:00; 10:30-12:00; 17:30-19:30; 22:00-0:00; Figure 5b As shown, mode 2 also has multiple troughs throughout the day, but compared with mode 1, the load rate of the trough values ​​is lower. The trough periods of mode 2 include: 5:00-7:30; 10:00-12:30; 17:15-20:00; 22:30-24:00. Figure 5a as well as Figure 5b Different electricity prices during the day Figure 4 The same, no repetition here.

[0110] Taking the cluster analysis of the energy consumption data of the machining department 1 of a certain enterprise in April as an example, the energy consumption data of the machining department 1 in April on the energy consumption side is obtained, the null values ​​and outliers in the energy consumption data are removed, and normalization is performed. The normalized daily energy consumption data is clustered and analyzed. Among them, the energy consumption data of the machining department 1 for 20 days in April can be aggregated into one category, and the energy consumption data of the machining department 1 for another 8 days can be aggregated into other categories. Therefore, the central curve of the aggregated energy consumption data of the machining department 1 for 20 days is used as the daily load curve of the department in April, as shown in Figure 1. Figure 6 As shown. Figure 6 It can be seen that the energy load of the processing department 1 reaches its lowest value around 18:30. Figure 6 Different electricity prices during the day Figure 4 The same, no repetition here.

[0111] Assuming that the enterprise consists of 11 departments, after obtaining the daily energy load curve of each department of the enterprise, the daily energy load curves are clustered again after clustering each department. The clustering results are as follows: Figure 7 As shown in the figure, 11 departments can be divided into four categories, among which assembly, assembly, core, sheet metal, machining 1, machining 2 and East Public Department can be classified as the first category. After aggregation, there are four obvious troughs in this category throughout the day. The trough periods are mainly concentrated in the morning, noon and evening shift change time, such as Figure 7 As shown in the upper left figure; East AMTC, winding, and East product certificate departments can be classified as the second type of departments. After aggregation, these departments reach the energy load peak from 11:00 to 15:00 throughout the day. Figure 7 As shown in the upper right figure; the East identified department as the third type of department, the energy load of this type of department does not fluctuate much throughout the day, and the energy load rate is low, such as Figure 7 As shown in the lower left figure; the East Living Department is the fourth type of department. The trough periods of energy load of this department throughout the day are 0:00~3:30 and 12:00~16:00, and 6:00~11:00 and 17:00~23:00 are more obvious peak periods. Figure 7 As shown in the lower right figure.

[0112] Taking the energy consumption data of electrocoated swimsuits in April, a preset key energy-consuming equipment of a certain enterprise, as an example, the energy consumption data of electrocoated swimsuits in April was obtained, the null values ​​and outliers in the energy consumption data were removed, and normalization was performed. The normalized daily energy consumption data was clustered and analyzed. Among them, the energy consumption data of electrocoated swimsuits for 17 days in April can be aggregated into one category, and the energy consumption data of electrocoated swimsuits for another 11 days can be aggregated into other categories. Therefore, the central curve of the aggregated energy consumption data of electrocoated swimsuits for 17 days is used as the daily load curve of electrocoated swimsuits in April, as shown in Figure 1. Figure 8 As shown. Figure 8 It can be seen that the energy load of electrocoated swimsuit reaches its lowest value around 11:30. Figure 8 Different electricity prices during the day Figure 4 The same, no repetition here.

[0113] In some embodiments, when an enterprise is equipped with an energy storage system, an electricity consumption strategy that participates in demand-side response can be formulated based on the electricity consumption characteristic curve obtained after clustering.

[0114] Figure 9 The flowchart of the method for formulating a power consumption strategy based on the clustered power load curve provided in an embodiment of the present application is exemplified. The process can be implemented by software and mainly includes the following steps:

[0115] S901: For any power supply mode, determine a first similarity between energy usage data of an object under the power supply mode and energy usage data of each sub-object corresponding to the object, wherein the power supply mode is a power supply mode of an energy storage system used by the object.

[0116] In this step, the energy storage system can be a solar energy storage system, a wind energy storage system, a hydropower storage system, a geothermal energy storage system, and the like. Different energy storage systems have different power supply modes. Assume that the set of energy usage data of objects under different power supply modes is Q = {q1, q2, ..., qn}, where n is an integer greater than or equal to 1, where q1 represents the energy usage data of the object under the first energy storage system power supply mode, q2 represents the energy usage data of the object under the second energy storage system power supply mode, and so on. The set of energy usage data of all sub-objects corresponding to the object is M = {m1, m2, ..., mm}, where m is an integer greater than or equal to 1, and m can be equal to n, where m1 represents the energy usage data of sub-object 1, m2 represents the energy usage data of sub-object 2, and so on. It should be noted that each element in the set Q is a plurality of aggregated energy consumption data of the object under a certain power supply mode, that is, each element in the set Q is an aggregated energy load curve of the object under a certain power supply mode; each element in the set M is a plurality of aggregated energy consumption data of each sub-object, that is, each element in the set M is an aggregated energy load curve of each sub-object.

[0117] In S901, a first similarity between the energy usage data of the object and the energy usage data of each sub-object can be determined based on a dynamic time warping (DTW) algorithm. Specifically, for any power supply mode, multiple Euclidean distances between multiple energy usage data of the object and multiple energy usage data of each sub-object under the power supply mode are calculated, and the shortest Euclidean distance is determined as the first similarity between the energy usage data of the object and the energy usage data of the sub-object. The specific calculation process is as follows:

[0118] (1) Construct an n*m matrix D.

[0119] Element d in matrix D ij =dist(q i ,m j ), i=1,2,…,n, j=1,2,…,m, dist is q i and m j The Euclidean distance between multiple energy consumption data.

[0120] (2) Determine d in matrix D 11 to d nm The shortest Euclidean distance.

[0121] Search d sequentially in matrix D11 to d nm The shortest Euclidean distance, for example, in d ij The search path for the Euclidean distance of a position is upward, right, and diagonally to the upper right, such as Figure 10a As shown, or, in d ij The search path for the Euclidean distance of a position is downward, right, and diagonally downward and right, such as Figure 10b The embodiment of the present application does not impose any restrictions on the search method.

[0122] (3) Substitute d in matrix D 11 to d nm The shortest Euclidean distance is used as the first similarity between the energy usage data of the object and the energy usage data of the sub-object.

[0123] S902: Selecting sub-objects whose first similarity is less than a first threshold from all sub-objects corresponding to the object.

[0124] In this step, the smaller the first similarity is, the more the power supply mode of the energy storage system can meet the energy demand of the sub-object.

[0125] S903: Determine a second similarity between energy usage data of a preset key energy-consuming device corresponding to the selected sub-object and the energy usage data of the object, and select preset key energy-consuming devices whose second similarity is less than a second threshold from the preset key energy-consuming devices.

[0126] In this step, the energy consumption data of the preset key energy-consuming devices under the sub-objects whose first similarity is less than the first threshold value is obtained, and the second similarity between the energy consumption data of the object and the energy consumption data of each preset key energy-consuming device corresponding to each selected sub-object is determined. Specifically, for any preset key energy-consuming device among the preset key energy-consuming devices corresponding to any selected sub-object, multiple Euclidean distances between the multiple energy consumption data of the preset key energy-consuming device and the multiple energy consumption data of the object are determined, and the shortest Euclidean distance is determined as the second similarity between the energy consumption data of the preset key energy-consuming device and the energy consumption data of the object. The calculation method of the second similarity is the same as the first similarity type and will not be repeated here. The second threshold value can be equal to the first threshold value.

[0127] After obtaining the second similarity between the energy usage data of each preset key energy-consuming device and the energy usage data of the object, a preset key energy-consuming device having a second similarity less than a second threshold is selected from all the preset key energy-consuming devices.

[0128] S904: Determine a power consumption strategy corresponding to the power supply mode according to the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period.

[0129] In this step, the minimum valley period is determined based on the valley moment corresponding to the minimum energy consumption data of the aggregated object. The minimum valley moment is the minimum inflection point of the energy load curve of the aggregated object. The left endpoint of the (minimum) valley period is the valley moment minus the first preset time, and the right endpoint of the valley period is the valley moment plus the second preset time. The first preset time and the second preset time can be the same or different. Among them, the energy load of the minimum valley period is the smallest, that is, the energy power of the minimum valley period is the smallest. By Figure 5a and Figure 5b It can be seen that the energy load curve of the object under different power supply modes has multiple valley moments, that is, there are multiple valley periods.

[0130] In S904, when calculating the average minimum power, the minimum valley period of the energy load curve of the object in the corresponding mode is used. An optional implementation method is to determine the sum of the power of the selected preset key energy-consuming equipment during the minimum valley period in each time period within the clustering period, where the minimum valley period is the minimum valley period obtained after clustering the energy consumption data of the objects in each time period, and the clustering period includes at least one time period.

[0131] For example, the clustering period is 1 month, the time period is 1 day, and in the first energy storage system power supply mode, the valley periods of the energy consumption data of the clustered objects are [a1, a2], [a3, a4], [a5, a6], and the minimum valley period is [a3, a4]. [a3, a4] includes multiple summary moments. The preset key energy-consuming devices selected are device 1, device 2, and device 3. Device 1 is at time t1, time t2, and time t3 in time period 1 [a3, a4]. The powers of device 2 at t1, t2, t3, and t4 in time period [a3, a4] of No. 1 are x1, x2, x3, and x4 respectively. The powers of device 2 at t1, t2, t3, and t4 in time period [a3, a4] of No. 1 are y1, y2, y3, and y4 respectively. The powers of device 3 at t1, t2, t3, and t4 in time period [a3, a4] of No. 1 are z1, z2, z3, and z4 respectively. The sum of the powers of device 1, device 2, and device 3 at each moment in time period [a3, a4] of No. 1 is They are W1=(x1+y1+z1), W2=(x2+y2+z2), W3=(x3+y3+z3), W4=(x4+y4+z4), among which W2 is the smallest; the power of device 1 at time t1, time t2, time t3, and time t4 in time period 2 [a3, a4] are x5, x6, x7, and x8 respectively, and the power of device 2 at time t1, time t2, time t3, and time t4 in time period 2 [a3, a4] are y5, y6 respectively , y7, y8, the powers of device 3 at t1, t2, t3, and t4 in time period 2 [a3, a4] are z5, z6, z7, and z8 respectively. Then the sum of the powers of device 1, device 2, and device 3 at each moment in time period 2 [a3, a4] are W5 = (x5 + y5 + z5), W6 = (x6 + y6 + z6), W7 = (x7 + y7 + z7), and W8 = (x8 + y8 + z8), respectively. W7 is the smallest, and so on.

[0132] Determine the average power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods according to the sum of the powers of the selected preset key energy-consuming equipment during the minimum valley period of all time periods;

[0133] Assuming there are 30 days in a month, the sum of the minimum power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods is (W2+W7+…), and the average power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods is (W2+W7+…) / 30.

[0134] The determined average power is used as the average minimum power of the selected preset key energy-consuming equipment within the clustering period, and the power consumption strategy corresponding to the corresponding power supply mode is determined according to the average minimum power;

[0135] For example, the object can participate in the peak-shaving response during the minimum valley period [a3, a4], and the first energy storage system can supply power to the object. At the same time, it is ensured that the power stored by the first energy storage system during the minimum valley period [a3, a4] before the response meets the average minimum power requirement of the object.

[0136] In the above-mentioned embodiment of the present application, for any power supply mode of the energy storage system used by the object, the sub-objects whose first similarity under the power supply mode is less than the first threshold are first screened out, and the second similarity between the energy consumption data of the preset key energy-consuming equipment corresponding to the selected sub-object and the energy consumption data of the object is determined. Then, the preset key energy-consuming equipment whose second similarity is less than the second threshold is screened out, and the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period is determined. The average minimum power provides an effective basis for the object to newly create distributed energy, thereby determining the power consumption strategy corresponding to each power supply mode according to the average minimum power, thereby reducing the energy cost of the object production.

[0137] In some embodiments, all sub-objects corresponding to the object can be classified first, and then the first similarity between the energy consumption data of the object and the energy consumption data of the sub-object class under each power supply mode can be determined, and the sub-object class whose first similarity is less than the first threshold is selected, and the second similarity between the energy consumption data of the preset key energy-consuming devices corresponding to each sub-object in the selected sub-object class and the energy consumption data of the object is determined, and the preset key energy-consuming devices whose second similarity is less than the second threshold are selected from the preset key energy-consuming devices. Furthermore, the power consumption strategy corresponding to each power supply mode is determined based on the average minimum power of the selected preset key energy-consuming devices during the minimum valley period.

[0138] In the above embodiment, all sub-objects included in the object are classified, and the first similarity between the energy consumption data of the object and the energy consumption data of each sub-object is determined. Compared with directly calculating the first similarity between the energy consumption data of the object and the energy consumption data of each sub-object, this saves the settlement process and improves the analysis efficiency.

[0139] For example, if the object is an enterprise and the sub-objects are departments within the enterprise, Figure 11 The flowchart of the method for formulating an electricity usage strategy for an enterprise provided by an embodiment of the present application is shown as an example. As shown in the figure, the process mainly includes the following steps:

[0140] S1101: For any power supply mode, obtain aggregated energy consumption data of the enterprise, and obtain the valley value of the aggregated energy consumption data of the enterprise and the valley time corresponding to the valley value.

[0141] In this step, the trough value of the enterprise's energy consumption data is the minimum inflection point of the aggregated enterprise's energy load curve. The left endpoint of the trough time is the trough time corresponding to the minimum inflection point minus 30 minutes, and the right endpoint of the trough time is the trough time corresponding to the minimum inflection point plus 30 minutes. An enterprise's energy consumption data can have multiple trough values, that is, multiple trough times.

[0142] Assume that the company has a 500kW / 2MWh energy storage system. The 500kW energy storage system corresponds to power supply mode 1, and the 2MWh energy storage system corresponds to power supply mode 2. Based on the company's energy consumption data in April, all the trough values ​​and corresponding trough times of the energy consumption load curve of the company under the two power supply modes are obtained as shown in Table 1.

[0143] Table 1. All trough values ​​and corresponding trough moments of the energy load curve on the energy consumption side of the enterprise

[0144]

[0145] S1102: Obtain energy consumption data of all departments in the enterprise.

[0146] In this step, a department class includes at least one department. The energy consumption data for a department class is the aggregated energy consumption data for all departments included in that department class. For example, if the first department class includes Assembly, Assembly, Iron Core, Sheet Metal, Machining 1, Machining 2, and East Public Utilities, the energy consumption data for the first department class is the aggregated energy consumption data for Assembly, Assembly, Iron Core, Sheet Metal, Machining 1, Machining 2, and East Public Utilities.

[0147] S1103: Determine a first similarity between the acquired energy consumption data of the enterprise and the energy consumption data of each department category.

[0148] In this step, the DTW algorithm may be used to calculate the first similarity. The specific calculation method is described in the above embodiment and will not be repeated here.

[0149] For example, the DTW algorithm is used to calculate the first similarity between the energy consumption data of the enterprise and the energy consumption data of the four department categories under each power supply mode on the energy consumption side. As shown in Table 2, assuming the first threshold is 10, there is no department category that meets the requirements under power supply mode 1. The first similarity of the first department category under power supply mode 2 (including seven departments: assembly, assembly, iron core, sheet metal, machining 1, machining 2, and East Public) is the smallest, and its matching degree with the power supply mode of the enterprise's second energy storage system is the highest.

[0150] Table 2. First similarity between the energy consumption data of enterprises and the energy consumption data of four departments

[0151] First sector category Second sector The third sector The fourth sector Power supply mode 1 18.7 28.8 45.7 20.8 Power supply mode 2 7.2 17.4 26.5 13.6

[0152] S1104: Select a department class whose first similarity is less than a first threshold, and obtain energy consumption data of preset key energy-consuming equipment corresponding to each department included in the selected department class.

[0153] In this step, a department class includes at least one department. Some departments may have at least one preset key energy-consuming equipment, while some departments may not have preset key energy-consuming equipment. The preset key energy-consuming equipment of different departments may be the same or different.

[0154] For example, there are 7 preset key energy-consuming equipment in the 7 departments under the first department category of the enterprise, including assembly leak detection, electro-coating swimsuit, die-casting machine, punching machine, molybdenum disulfide, hydraulic press, and cleaning machine. The energy consumption data of the 7 preset key energy-consuming equipment under power supply mode 2 are obtained.

[0155] S1105: Determine a second similarity between the energy consumption data of each preset key energy-consuming device and the energy consumption data of the enterprise.

[0156] In this step, the DTW algorithm may be used to calculate the second similarity. The specific calculation method is described in the above embodiment and will not be repeated here.

[0157] For example, there are 7 preset key energy-consuming equipment in the 7 departments under the first department category. Under power supply mode 2, the second similarity between the energy consumption data of each preset key energy-consuming equipment and the energy consumption data of the enterprise is calculated respectively, as shown in Table 3. Assuming the second threshold is 10, the energy consumption data of the 7 preset key energy-consuming equipment meet the similarity threshold requirements. The preset key energy-consuming equipment includes: assembly leak detection, electrocoating swimwear, die-casting machine, molybdenum disulfide, and cleaning machine.

[0158] Table 3. Second similarity between the energy consumption data of the preset key energy-consuming equipment and the enterprise under power supply mode 2

[0159]

[0160] S1106: Select preset key energy-consuming devices whose second similarity is less than a second threshold from the preset key energy-consuming devices, and determine the average minimum power of the selected preset key energy-consuming devices during the minimum valley period.

[0161] S1107: Determine a power usage strategy corresponding to each power supply mode according to the average minimum power.

[0162] The detailed description of S1106 and S1107 can be found in S904 and will not be repeated here.

[0163] For example, the first department category includes five pre-set key energy-consuming equipment: assembly leak detection, electrocoating swimwear, die-casting machines, molybdenum disulfide, and cleaning machines. These equipment meet the second similarity threshold requirement. The sum of the power consumed by these five pre-set key energy-consuming equipment during the minimum trough period (04:15, 05:15) is calculated daily, and the average minimum power consumed by these five pre-set key energy-consuming equipment in April is calculated to be 260.3 kW. Based on this average minimum power, a demand-side response strategy can be developed: When the enterprise's power consumption characteristics are predicted to meet power supply mode 2, the enterprise can participate in grid-side peak-shaving demand-side response during the minimum trough period, with a load capacity of 260.3 kW. Simultaneously, the energy storage system changes its charging and discharging strategy to ensure a power supply of 260 kWh before responding, meeting the power needs of the enterprise's pre-set key energy-consuming equipment.

[0164] Based on the same technical concept, the present application provides an electric power analysis device, which can realize the above-mentioned embodiment. Figure 2 、 Figure 9 The specific implementation process and technical effects are the same as those in the above embodiment and will not be repeated here.

[0165] See also Figure 12 The device includes a similarity determination module 1201, a selection module 1202, and a power usage strategy determination module 1203.

[0166] A similarity determination module 1201 is configured to determine, for any power supply mode, a first similarity between the energy usage data of the object under the power supply mode and the energy usage data of each sub-object corresponding to the object; and a second similarity between the energy usage data of a preset key energy-consuming device corresponding to a selected sub-object and the energy usage data of the object, wherein the power supply mode is a power supply mode of an energy storage system used by the object;

[0167] The selection module 1202 is configured to select sub-objects whose first similarity is less than a first threshold value from all sub-objects corresponding to the object; and select preset key energy-consuming devices whose second similarity is less than a second threshold value from the preset key energy-consuming devices;

[0168] The power usage strategy determination module 1203 is configured to determine a power usage strategy corresponding to the power supply mode according to the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period.

[0169] Optionally, the apparatus further includes a power determination module 1204, configured to:

[0170] Determine the sum of the power of the selected preset key energy-consuming equipment during the minimum valley period of each time period within the clustering period, wherein the minimum valley period is the minimum valley period obtained after clustering the energy consumption data of the objects within each time period, and the clustering period includes at least one time period;

[0171] Determine the average power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods according to the sum of the powers of the selected preset key energy-consuming equipment during the minimum valley period of all time periods;

[0172] The determined average power is used as the average minimum power of the selected preset key energy-consuming equipment within the clustering period.

[0173] Optionally, the similarity determination module 1201 is specifically configured to:

[0174] For any sub-object among all sub-objects corresponding to the object, determining multiple Euclidean distances between multiple energy usage data of the object and multiple energy usage data of the sub-object under the power supply mode, and determining the shortest Euclidean distance as the first similarity between the energy usage data of the object and the energy usage data of the sub-object;

[0175] For any preset key energy-consuming device among the preset key energy-consuming devices corresponding to any selected sub-object, determine multiple Euclidean distances between multiple energy consumption data of the preset key energy-consuming device and multiple energy consumption data of the object, and determine the shortest Euclidean distance as the second similarity between the energy consumption data of the preset key energy-consuming device and the energy consumption data of the object.

[0176] Optionally, the apparatus further includes a clustering module 1205, configured to:

[0177] Eliminate abnormal data from the energy usage data of the object collected during the clustering period and the energy usage data of the sub-objects corresponding to the object;

[0178] According to the set summary cycle, the energy consumption data of the collected object and the energy consumption data of the sub-objects corresponding to the object are summarized respectively. Each summary cycle includes multiple collection cycles;

[0179] Normalize the aggregated energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object;

[0180] Based on the trained clustering model, the energy consumption data of the normalized object and the energy consumption data of the sub-objects corresponding to the object are clustered to obtain the clustered energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object.

[0181] Optionally, the clustering model is trained by:

[0182] Traverse the preset minimum number of clusters to the maximum number of clusters, and obtain the intra-cluster error variance and silhouette coefficient corresponding to each number of clusters;

[0183] The number of clusters with the inflection point of all intra-cluster error variances and the largest silhouette coefficient is selected as the number of clusters in the clustering model.

[0184] Based on the same technical concept, an embodiment of the present application also provides an electricity consumption analysis device, which can implement the electricity consumption analysis method in the aforementioned embodiment.

[0185] See also Figure 13 The device includes a memory 1301 and a processor 1302:

[0186] Memory 501, used for storing program instructions;

[0187] The processor 502 is used to call the program instructions stored in the memory 501 to implement the above-mentioned power consumption analysis method in the embodiment of the present application.

[0188] The specific implementation methods and technical effects will not be repeated again.

[0189] An embodiment of the present application further provides a computer-readable storage medium, which stores computer-executable instructions. The computer-executable instructions are used to enable a computer to execute the method for analyzing electricity consumption in the above embodiment.

[0190] An embodiment of the present application further provides a computer program product for storing a computer program, wherein the computer program is used to execute the method for analyzing electricity consumption in the aforementioned embodiment.

[0191] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0192] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0193] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0194] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0195] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A method for analyzing electricity consumption, characterized in that: include: For any power supply mode, determining a first similarity between energy usage data of an object under the power supply mode and energy usage data of each sub-object corresponding to the object, wherein the power supply mode is a power supply mode of an energy storage system used by the object; Selecting a sub-object whose first similarity is less than a first threshold from all sub-objects corresponding to the object; Determining a second similarity between energy usage data of a preset key energy-consuming device corresponding to the selected sub-object and the energy usage data of the object, and selecting a preset key energy-consuming device whose second similarity is less than a second threshold from the preset key energy-consuming devices; The power consumption strategy corresponding to the power supply mode is determined according to the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period.

2. The method according to claim 1, wherein The average minimum power of the selected preset key energy-consuming equipment during the minimum trough period is determined by the following method: Determining the sum of the power of the selected preset key energy-consuming equipment during the minimum valley period of each time period within the clustering period, wherein the minimum valley period is the minimum valley period obtained after clustering the energy consumption data of the object in each time period, and the clustering period includes at least one time period; Determine the average power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods according to the sum of the powers of the selected preset key energy-consuming equipment during the minimum valley period of all time periods; The determined average power is used as the average minimum power of the selected preset key energy-consuming equipment within the clustering period.

3. The method according to claim 1, wherein The first similarity is determined by: determining, for any sub-object among all sub-objects corresponding to the object, multiple Euclidean distances between multiple energy usage data of the object and multiple energy usage data of the sub-object under the power supply mode, and determining the shortest Euclidean distance as a first similarity between the energy usage data of the object and the energy usage data of the sub-object; The second similarity is determined by: For any preset key energy-consuming device among the preset key energy-consuming devices corresponding to any selected sub-object, determine multiple Euclidean distances between multiple energy consumption data of the preset key energy-consuming device and multiple energy consumption data of the object, and determine the shortest Euclidean distance as the second similarity between the energy consumption data of the preset key energy-consuming device and the energy consumption data of the object.

4. The method according to claim 1, wherein The method further comprises: Eliminating abnormal data from the energy usage data of the object and the energy usage data of sub-objects corresponding to the object collected during the clustering period; According to the set summary period, the collected energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are summarized respectively, and each summary period includes multiple collection periods; Normalizing the aggregated energy usage data of the object and the energy usage data of the sub-objects corresponding to the object; Based on the trained clustering model, the normalized energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are clustered to obtain the clustered energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object.

5. The method according to claim 4, wherein The clustering model is trained in the following way: Traverse the preset minimum number of clusters to the maximum number of clusters, and obtain the intra-cluster error variance and silhouette coefficient corresponding to each number of clusters; The number of clusters with the inflection point of all intra-cluster error variances and the largest silhouette coefficient is selected as the number of clusters of the clustering model.

6. An electricity consumption analysis device, characterized in that: include: A similarity determination module, configured to determine, for any power supply mode, a first similarity between energy usage data of an object under the power supply mode and energy usage data of each sub-object corresponding to the object; Determining a second similarity between energy usage data of a preset key energy-consuming device corresponding to the selected sub-object and the energy usage data of the object, wherein the power supply mode is a power supply mode of an energy storage system used by the object; A selection module, configured to select a sub-object whose first similarity is less than a first threshold from all sub-objects corresponding to the object; Selecting preset key energy-consuming devices whose second similarity is less than a second threshold from preset key energy-consuming devices; The power usage strategy determination module is used to determine the power usage strategy corresponding to the power supply mode according to the average minimum power of the selected preset key energy-consuming equipment during the minimum valley period.

7. The device according to claim 6, characterized in that The device further includes a power determination module, configured to: Determining the sum of the power of the selected preset key energy-consuming equipment during the minimum valley period of each time period within the clustering period, wherein the minimum valley period is the minimum valley period obtained after clustering the energy consumption data of the object in each time period, and the clustering period includes at least one time period; Determine the average power of the selected preset key energy-consuming equipment during the minimum valley period of all time periods according to the sum of the powers of the selected preset key energy-consuming equipment during the minimum valley period of all time periods; The determined average power is used as the average minimum power of the selected preset key energy-consuming equipment within the clustering period.

8. The device according to claim 6, wherein The similarity determination module is specifically used to: determining, for any sub-object among all sub-objects corresponding to the object, multiple Euclidean distances between multiple energy usage data of the object and multiple energy usage data of the sub-object under the power supply mode, and determining the shortest Euclidean distance as a first similarity between the energy usage data of the object and the energy usage data of the sub-object; For any preset key energy-consuming device among the preset key energy-consuming devices corresponding to any selected sub-object, determine multiple Euclidean distances between multiple energy consumption data of the preset key energy-consuming device and multiple energy consumption data of the object, and determine the shortest Euclidean distance as the second similarity between the energy consumption data of the preset key energy-consuming device and the energy consumption data of the object.

9. The device according to claim 6, wherein The apparatus further comprises a clustering module, configured to: Eliminating abnormal data from the energy usage data of the object and the energy usage data of sub-objects corresponding to the object collected during the clustering period; According to the set summary period, the collected energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are summarized respectively, and each summary period includes multiple collection periods; Normalizing the aggregated energy usage data of the object and the energy usage data of the sub-objects corresponding to the object; Based on the trained clustering model, the normalized energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object are clustered to obtain the clustered energy consumption data of the object and the energy consumption data of the sub-objects corresponding to the object.

10. The device according to claim 9, wherein The clustering model is trained in the following way: Traverse the preset minimum number of clusters to the maximum number of clusters, and obtain the intra-cluster error variance and silhouette coefficient corresponding to each number of clusters; The number of clusters with the inflection point of all intra-cluster error variances and the largest silhouette coefficient is selected as the number of clusters of the clustering model.

11. An electricity consumption analysis device, characterized in that: including memory and processor; The memory is used to store program instructions; The processor is configured to execute the method according to any one of claims 1 to 5 according to the program instructions stored in the memory.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the method according to any one of claims 1 to 5.

Images