Energy data management apparatus

Abstract

To provide an energy data management apparatus that utilizes energy data managed for each product to accurately predict energy demand.SOLUTION: An embodiment comprises a data extraction unit and a demand prediction calculation unit. The data extraction unit extracts energy data from a plurality of pieces of energy data based on extraction conditions, and uses a statistical function to exclude energy data with outliers from the extracted energy data. The demand prediction calculation unit calculates an energy demand prediction for the product whose production is planned based on the average or median value of the remaining energy data excluding energy data with outliers and production information data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] An embodiment of the present invention relates to an energy data management device that predicts energy demand in a plant. [Background technology]

[0002] For the purpose of managing energy data in plants, an energy data management device has been proposed, for example, in Patent Document 1. This energy data management device associates time-series energy data with product information and event information, and manages the energy consumed by each process for each product.

[0003] Furthermore, for example, Patent Document 2 proposes a method of utilizing operational performance data for each rolled material rolled on a rolling line.

[0004] When the type of energy consumed at a plant is electrical energy, an hourly power supply contract may be concluded with an electric power company for each plant. Having an electrical power contract with just the right amount of supply can contribute to the smooth operation of the plant and cost reduction. In addition to electrical energy, various energy sources are required depending on the products, and procuring just the right amount of required energy and supplying it to the site can contribute to cost reduction and improved productivity. In recent years, this has not stopped there; there has been a strong demand for appropriate energy management and use due to the need to reduce energy consumption as part of efforts toward carbon neutrality.

[0005] To make energy demand forecasts more accurately, more accurate energy data should be used, but the energy data that can actually be obtained fluctuates due to various factors on the production line of a plant and may show abnormal values. If energy demand forecasts are made based on data that includes abnormal values, significant errors may occur in the forecast values, resulting in inappropriate results and the risk of problems such as excessive ordering of energy sources. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Publication No. 2022-92898 [Patent Document 2] Japanese Patent Application Publication No. 10-216813 Summary of the Invention [Problem to be solved by the invention]

[0007] An object of an embodiment of the present invention is to provide an energy data management device that utilizes energy data managed for each product to perform energy demand prediction with high accuracy. [Means for solving the problem]

[0008] The energy data management device according to the present invention includes a data extraction unit that extracts two or more energy data from a plurality of energy data each representing the actual energy consumption of a plurality of products based on one or more extraction conditions, detects energy data having outliers from the two or more extracted energy data based on one or more statistical processing functions and judgment criteria for the one or more statistical processing functions, and excludes the energy data having the outliers, and a demand prediction calculation unit that calculates an energy demand prediction for a product planned to be manufactured based on an average value or a median value of the remaining energy data after excluding the energy data having the outliers from the plurality of energy data, and production information data including attributes and a manufacturing schedule of the product planned to be manufactured. [Effects of the Invention]

[0009] According to an embodiment of the present invention, it is possible to provide an energy data management device that utilizes energy data managed for each product to perform energy demand prediction with high accuracy. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a schematic block diagram illustrating an energy data management device according to an embodiment. [Figure 2] 10 is a display example of a schematic image showing the operation of the energy data management device according to the embodiment. [Figure 3] 10 is a display example of a schematic image showing the operation of the energy data management device according to the embodiment. [Figure 4] 10 is an example of a flowchart illustrating an operation of the energy data management device according to the embodiment. [Figure 5] 10 is a display example of data extracted by the operation of the energy data management device according to the embodiment. [Figure 6] 10 is a display example of data extracted by the operation of the energy data management device according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, embodiments will be described with reference to the drawings. The drawings are schematic or conceptual, and the relationship between the thickness and width of each part, the size ratio between parts, etc. are not necessarily the same as those in reality. Furthermore, even when the same part is shown, the dimensions and ratios may be different depending on the drawing. In the present specification and the drawings, elements similar to those described above with reference to the previous drawings are designated by the same reference numerals, and detailed descriptions thereof will be omitted as appropriate.

[0012] FIG. 1 is a schematic block diagram illustrating an energy data management device according to an embodiment. As shown in FIG. 1 , an energy data management device 1 according to an embodiment includes a data extraction unit 7, a demand prediction calculation unit 9, and a display unit 10. In the example shown in FIG. 1 , the energy data management device 1 also includes the components of the energy data management device of Patent Document 1, as well as an energy database 6 collected and managed by the components. As shown in FIG. 1 , the energy data management device of Patent Document 1 includes a data collection unit 2, a tracking processing unit 3, a power consumption calculation unit 4, and a product basic unit calculation unit 5, and collects and manages the energy consumption of each process constituting a production line for each product. As in this example, the energy data management device of Patent Document 1 calculates energy data for each product and each process using various data acquired from a tracking system 11, a host computer 12, and a remote IO system 13. As long as a desired energy database 6 can be constructed, the energy data collection method can be configured as appropriate.

[0013] The energy database 6 stores the energy data for each product and each process that has been collected and calculated as described above. As will be described later, the energy data stored in the energy database 6 associates product IDs, which are identification numbers for each product, with data on product attributes, manufacturing date and time, and energy consumption.

[0014] The data extraction unit 7 extracts multiple desired energy data items from the energy database 6 based on desired extraction conditions. The data extraction unit 7 has one or more types of statistical processing functions, and statistically processes the multiple energy data items extracted from the energy database 6 to detect energy data items that are outliers. The data extraction unit 7 performs a process to exclude energy data items that have the detected outliers. The data extraction unit 7 performs a process to exclude energy data items that have outliers and updates the energy database 6.

[0015] In the data extraction unit 7, the extraction conditions for energy data are set in advance. The extraction conditions for energy data can be set as appropriate depending on the products to be manufactured, etc. It is preferable that the extraction conditions for energy data are sufficiently subdivided and hierarchical. By sufficiently subdividing and hierarchizing the data extraction conditions, it becomes possible to more accurately predict energy demand for products that have attribute combinations with no or little track record.

[0016] The demand prediction calculation unit 9 calculates a predicted energy demand value by reading the set energy data from the updated energy database 6. The demand prediction calculation unit 9 can calculate a predicted energy demand value according to the production schedule of each product by reading the recipe and production schedule data for each product from an external production information management system 14.

[0017] The display unit 10 not only displays the data extracted by the data extraction unit 7 on a display, but also provides an interface such as a GUI (Graphical User Interface). The display unit 10 cooperates with the data extraction unit 7 to select an analysis means for the statistical processing of the data extraction unit 7 and set criteria for the statistical processing.

[0018] The display unit 10 displays the energy demand prediction calculated by the demand prediction calculation unit 9 using the energy data in the updated energy database 6.

[0019] In the example of Fig. 1, the energy data management device 1 has a production information input unit 8. The production information input unit 8 replaces the production information management system 14. The production information input unit 8 is applied, for example, to a management system for a small-scale plant where the production information management system 14 is not provided. The production information input unit 8 cooperates with the display unit 10 and inputs the production schedule for each product in the demand prediction calculation to the demand prediction calculation unit 9 each time using a GUI or the like.

[0020] The settings and operations of the energy data management device 1 according to the embodiment will be described. FIG. 2 is a display example of a schematic image showing the operation of the energy data management device according to the embodiment. Image P1 in Fig. 2 is an example of an image displayed on a display (not shown) connected to the energy data management device 1. Image P1 shows the setting range of the energy data extraction conditions set in the data extraction unit 7, and is set in advance in the data extraction unit 7. The energy data extraction conditions can be set arbitrarily, for example, manually, or the like, and may be set each time a demand prediction calculation is performed.

[0021] The example in Figure 2 shows an application to a steel rolling plant, where the extraction conditions are product attributes. In image P1, the product attributes are product thickness (denoted as "Thickness") 20 and product width (denoted as "Width") 21. Other extraction conditions can also be set, such as product length, weight, and material. In the energy database 6, attribute data such as thickness and width are associated with product IDs as energy data. By subdividing and hierarchizing the data as shown in image P1, the extraction conditions for energy data can be set in greater detail. In addition, the energy data associated with a product ID is also associated with the manufacturing date of the product having that product ID. In energy data, the product ID can be associated with the date of production on the line and the date of completion of the product as the manufacturing date, and further associated with the date (time) of production and completion of each process.

[0022] FIG. 3 is a display example of a schematic image showing the operation of the energy data management device according to the embodiment. Image P2 in Fig. 3 is an example of an image displayed on a display connected to the energy data management device 1. Image P2 shows an example of setting extraction conditions for energy data from the energy database 6. The extraction conditions for performance data shown in image P2 are set in the data extraction unit 7 via the display unit 10, for example.

[0023] As described above, in the energy database 6, data on the manufacturing date of a product is associated with the product ID. By specifying the manufacturing date as a data extraction condition as shown in the condition setting field 30 and pressing the extraction button 33, energy data for a product with a desired manufacturing date can be extracted. When collecting energy data, energy data is also collected when an abnormality occurs in any process on the production line. Data collected during such process abnormalities often becomes an outlier that is far removed from data collected during normal times. Therefore, by excluding such periods from the extraction conditions in advance as shown in the condition setting field 31 when extracting data, the data extraction process can be expedited. In the specific example of FIG. 3, a comment regarding the exclusion range can be entered.

[0024] A series of operations of the energy data management device 1 according to the embodiment will be described using a flowchart. FIG. 4 is an example of a flowchart illustrating the operation of the energy data management device according to the embodiment. 5 and 6 are examples of displaying data extracted by the operation of the energy data management device according to the embodiment. 5 and 6 also show the setting of extraction conditions from the energy database 6, and the setting of extraction conditions will be explained first.

[0025] FIG. 5 shows a screen P3 on which extraction conditions are set and the extraction results are displayed. As shown in FIG. 5, a period setting field 50 and an attribute setting field 51 are provided at the top of screen P3. In the period setting field 50, the range of manufacturing dates for the data to be extracted is input. In the attribute setting field 51, the steel type (written as "Steel grade"), thickness range (written as "Thickness"), and width range (written as "Width") can be input. Note that any range can also be input for weight (written as "Weght") and length (written as "Length", but in this example, the entire range (written as "All") is set.

[0026] Results 52 and 53 set according to these extraction conditions are displayed below the period setting field 50 and the attribute setting field 51. Buttons 54 to 57 are GUIs for setting data selection and the like by operating the screen.

[0027] The display data can be set arbitrarily, and in results 52 and 53 in Figure 5, a frequency distribution for the amount of power consumed under the extraction conditions is calculated and displayed, and in result 60 shown in image P4 in Figure 6, a scatter plot of the amount of power consumed against the date of manufacture is calculated and displayed.

[0028] As shown in FIG. 4, in step S1, extraction conditions for energy data to be extracted by the data extraction unit 7 are set via the GUI of the display unit 10 or the like. As explained in relation to FIGS. 2 and 3, the data to be extracted is extracted by setting extraction conditions such as the manufacturing period and product attributes according to the attributes of the product to be manufactured. When setting extraction conditions for product attributes, a more finely classified hierarchy can be used, as explained in relation to FIG. 2. By applying a finely hierarchical classification such as the thickness and width of the product for which demand prediction is to be performed as extraction conditions, it becomes possible to more accurately predict the energy demand for small-lot, high-mix products.

[0029] In step S2, the data extraction unit 7 extracts a plurality of energy data (referred to as performance data) from the energy database 6 according to the set extraction conditions.

[0030] In step S3, the data extraction unit 7 selects the parameters to be analyzed and sets the statistical analysis means via the GUI of the display unit 10 or the like. Examples of statistical analysis means include mean ± σ × coefficient, median ± median absolute deviation × 1.4826 × coefficient, and percentile. One of the analysis means is selected via the GUI or the like, and the outlier determination criteria corresponding to the analysis means are input. For example, in the case of mean ± σ × coefficient, the outlier determination criteria is set to what value the "coefficient" should be.

[0031] When the analysis parameters and analysis means are selected, as shown in FIG. 5, the judgment criteria ranging from ±σ to ±3σ are displayed on screen P3, and the operator selects, for example, ±3σ using a GUI or the like. The selected ±3σ becomes the criterion for judging outliers, and data outside the range of the mean value ±3σ is deemed to be outliers, and the displayed plot color, for example, changes. In the example of FIG. 6, the mean absolute deviation is introduced into the analysis means along with the standard deviation σ. In this way, the data extraction unit 7 is not limited to applying a single statistical analysis means, and multiple analysis means may be applied to judge outliers.

[0032] The data extraction unit 7 and display unit 10 display the analysis results obtained by the statistical processing, and in step S4, the operator determines whether or not there are any outliers to be excluded in the analysis results. If the operator determines that there are no outliers to be excluded, the process ends. If the operator determines that there are outliers to be excluded, the process proceeds to step S5. Note that the determination in step S4 may be made automatically under preset conditions, rather than by the operator.

[0033] In step S5, the data extraction unit 7 excludes data that falls outside the range of the judgment criteria determined by the statistical analysis means via the GUI of the display unit 10 or the like. The excluded data may be deleted from the energy database 6, or an excluded data flag may be set to active. In this way, the energy database 6 is updated.

[0034] Thereafter, the demand prediction calculation unit 9 acquires data on the product production schedule from the production information management system 14 or the production information input unit 8, and acquires energy data by process for each product from the updated energy database 6. Using this data, it is possible to calculate an energy demand prediction for each product according to the production schedule and output it via the display unit 10.

[0035] The effects of the energy data management device 1 according to the embodiment will be described. The energy data management device 1 according to the embodiment can read desired energy data from an energy database 6 that stores data on energy consumption for each process for each product, and can therefore perform energy demand prediction according to a production schedule. The energy data management device 1 is equipped with a data extraction unit 7, and can use statistical means to remove outliers that could be a source of errors when performing energy demand prediction on the data stored in the energy database 6. Therefore, even if the energy database 6 contains data containing outliers due to process or other abnormalities, this data can be removed in advance to perform energy demand prediction, thereby enabling more accurate demand prediction.

[0036] The data extraction unit 7 can arbitrarily set extraction conditions for detecting and excluding abnormal values. Extraction conditions can be finely hierarchized according to the characteristics of the product, ensuring reliable detection of outliers when forecasting demand for high-mix, low-volume products.

[0037] In this way, it is possible to realize an energy data management device that utilizes energy data managed for each product to accurately predict energy demand.

[0038] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0039] 1...energy data management device, 2...data collection unit, 3...tracking processing unit, 4...power consumption calculation unit, 5...product basic unit calculation unit, 6...energy database, 7...data extraction unit, 8...production information input unit, 9...demand prediction calculation unit, 10...display unit, 11...tracking system, 12...host computer, 13...remote IO system, 14...production information management system

Claims

1. a data extraction unit that extracts two or more energy data from a plurality of energy data representing actual energy consumption data of a plurality of products based on one or more extraction conditions, detects energy data having outliers from the two or more extracted energy data based on one or more statistical processing functions and criteria for the one or more statistical processing functions, and excludes the energy data having outliers; a demand prediction calculation unit that calculates an energy demand prediction for the product planned to be manufactured based on an average value or a median value of the remaining plurality of energy data after excluding the energy data having the outlier from the plurality of energy data, and production information data including attributes and a manufacturing schedule of the product planned to be manufactured; An energy data management device comprising:

2. The energy data management device according to claim 1 , wherein the data extraction unit subdivides the one or more extraction conditions into a plurality of categories.

3. The energy data management device according to claim 1 or 2, wherein the data extraction unit includes, as the one or more statistical processing functions, any one of mean ± σ × coefficient, median ± median absolute deviation × 1.4826 × coefficient, and percentile.

Images