Information-based knowledge production device, information-based knowledge production method, trained model, trained model generation method, and program

Abstract

One aspect of the present invention is an information-based knowledge production device that comprises: a knowledge production vocabulary management unit that assigns unified meaning to information about different types of power systems in a machine-readable form and manages an ontology that defines a common vocabulary that is needed for knowledge production; and an information-based knowledge production unit that adds metadata that uses the common vocabulary of the ontology to individual pieces of information and produces knowledge from the information.

Description

Information knowledge generation device, information knowledge generation method, trained model, trained model generation method and program

[0001] The present disclosure relates to an information knowledge generation device, an information knowledge generation method, a trained model, a trained model generation method, and a program.

[0002] In recent years, artificial intelligence (AI) has been used to predict power generation and power supply and demand. The prediction results are used to adjust supply and demand among power generation by power generation facilities, power generation by renewable energy power generation facilities, power supply, and power demand at load facilities.

[0003] As a technology for predicting power generation based on weather forecast information, Patent Document 1 discloses a power generation prediction device that obtains a predicted value for power generation using a model constructed by machine learning using explanatory variables that include at least weather forecast information for a mesh that includes a prediction point and multiple surrounding meshes, and a target variable that corresponds to the power generation amount from natural energy.

[0004] International Publication No. 2020 / 203854

[0005] The technology disclosed in Patent Document 1 cannot handle cases where weather forecast information is unavailable for each forecast location and its surrounding areas. Furthermore, when forecasting power generation and demand for generated power at various power generation facilities, such as small-scale solar power generation and hydrogen storage facilities, which are expected to increase in the future, it is necessary to handle a variety of data with different meanings, such as measurement data collected from a variety of fluctuating devices and meteorological observation data. When integrating multiple heterogeneous distributed power systems into a virtual power plant to provide grid power suppression services, current methods target relatively large-scale power systems by centrally controlling individual distributed unit EMSs (Energy Management Systems). This is not suitable for controlling virtual power plants, which target a large number of relatively small distributed unit EMSs, as discussed in the present disclosure. Furthermore, there is no effective method for accurately suppressing overall demand by managing consumers who own a large number of distributed unit EMSs.

[0006] The present disclosure provides an information knowledge generation device that can configure a system that can easily accommodate the diversity of EMS-related equipment and data, as well as changes and additions to them, by assigning a unified meaning to various information related to heterogeneous power systems and converting it into machine-processable knowledge.

[0007] An information knowledge system according to one aspect of the present invention includes a knowledge vocabulary management unit that manages an ontology that defines the common vocabulary required for knowledge system creation by assigning a unified meaning to information related to different types of power systems in a machine-readable form, and an information knowledge system that assigns metadata using the common vocabulary of the ontology to each piece of information and knowledge systems the information. Furthermore, by structuring a learning model using a collection of knowledge-based data and a physical simulator, it is possible to predict the power generation amount of individual power systems installed in a specific region, thereby realizing a highly accurate virtual power plant through request-based control from a Virtual Power Plant - Demand Response Energy Management System (VPP-DR-EMS).

[0008] According to the information knowledge system of the present disclosure, it is possible to configure a system that can easily accommodate the variety of devices and data related to EMS, as well as changes and additions thereto.

[0009] FIG. 1 is a functional configuration diagram of an information knowledge system according to an embodiment of the present invention. FIG. 2 is a configuration diagram of an electric power system realized by an information knowledge system according to an embodiment of the present invention. FIG. 3 is a configuration diagram of a DU-EMS (distributed unit EMS) of an electric power system according to an embodiment of the present invention. FIG. 4 is a hardware configuration diagram of an information knowledge system according to an embodiment of the present invention. FIG. 5 is a diagram showing an example of a class hierarchical structure in an information knowledge system according to an embodiment of the present invention. FIG. 6 is a diagram showing an example of a class hierarchical structure in an information knowledge system according to an embodiment of the present invention. FIG. 7 is a diagram showing an example of a data image of a measurement result information file in the electric power system shown in FIG. 2. FIG. 8 is a diagram showing an example of a schema of metadata of measurement result information in the electric power system shown in FIG. 2. FIG. 9 is a diagram showing an example of a schema of metadata of sensors in the electric power system shown in FIG. 2. FIG. 10 is a diagram showing an example of metadata of measurement result information in the electric power system shown in FIG. 2.

[0010] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following describes the preferred embodiments of the present invention with reference to the accompanying drawings. In the drawings, the same components are designated by the same reference numerals, and redundant explanations may be omitted.

[0011] [Embodiment] <Configuration of information knowledge system 10> Fig. 1 is a functional configuration diagram of an information knowledge system 10 according to one embodiment of the present invention. As shown in the figure, the information knowledge system 10 includes a knowledge system vocabulary management unit 1 and an information knowledge system unit 2. The knowledge system vocabulary management unit 1 assigns a unified meaning to information relating to different types of power systems in a machine-readable form and manages an ontology that defines a common vocabulary required for knowledge system creation. The information knowledge system 2 assigns metadata using the common vocabulary of the ontology to each piece of information, and converts the information into knowledge system.

[0012] In an ontology, vocabularies are defined for classes corresponding to concepts and for properties corresponding to relationships between concepts. In an ontology, hierarchical relationships between classes and properties are defined. More specifically, when classes and properties are defined in an ontology, the hierarchical relationships between classes and properties are specified.

[0013] The information about the heterogeneous power systems may include at least equipment information and measurement data about the power systems.

[0014] 2 is a configuration diagram of a power system 100 implemented by an information knowledge system 10 according to an embodiment of the present invention. As shown in the figure, the power system 100 includes an integrated EMS and a plurality of distributed unit EMSs. In the following, the integrated EMS will be described as the VPP-DR-EMS 3 shown in the figure. Furthermore, the distributed unit EMS will be described as a DU-EMS (Distributed Unit EMS) 4.

[0015] The functions of the knowledge vocabulary management unit 1 and the information knowledge unit 2 included in the information knowledge device 10 may be included in the VPP-DR-EMS 3 or the DU-EMS 4 .

[0016] The information knowledge system 10 can also execute an information knowledge system method, which includes the steps of assigning a unified meaning to information relating to different types of power systems in a machine-readable form and managing an ontology that defines a common vocabulary required for knowledge system creation, and of adding metadata to each piece of information using the common vocabulary of the ontology to create knowledge system.

[0017] The components of the power system 100 may include power generation facilities that use renewable energy. Because the amount of power generated by power generation facilities that use renewable energy fluctuates depending on weather conditions, power generation forecasts are performed to stabilize the power grid. Power utilities create power supply and demand plans for multiple power systems that are aggregated as a virtual power plant. If there is a discrepancy between the amount of power based on the power supply and demand plan and the actual amount of power, a problem occurs in which the expected adjustment capacity for the grid power cannot be obtained. Therefore, the VPP-DR-EMS 3 is required to perform highly accurate power generation forecasts, etc., in order to improve the reliability of the power utility's adjustment capacity for the grid power.

[0018] <Example> Figure 3 is a configuration diagram of a DU-EMS4 of a power system 100 according to one embodiment of the present invention. The DU-EMS4 includes a power transmission network N, multiple input power sources 20 and 30, a main voltage controller (MVC) 40, and sub-voltage controllers (SVCs) 50 and 60. The DU-EMS4 needs to include at least one of the input power sources 20 and 30, the main voltage controller 40, and the sub-voltage controllers 50 and 60, and may include multiple of each. Each of the input power sources 20 and 30, the main voltage controller 40, and the sub-voltage controllers 50 and 60 is connected to a DC bus 70. Note that the solid lines in the figure represent the power network, and the dashed lines represent the data communication network.

[0019] A load 80 is connected to the output of the DU-EMS 4. The load 80 may be a DC load such as a DC motor, or a DC / AC converter that converts DC power into AC power and its AC load. Alternatively, the load 80 may be an AC power system connected to the DC bus 70 via a DC / AC converter.

[0020] The load 80 may be, for example, another power device such as an electric vehicle, etc. The load 80 may also be a plurality of power devices connected in parallel.

[0021] Input power sources 20 and 30, which are renewable energy power generation systems, are connected to the DU-EMS 4. In the illustrated example, the input power source 20 is a solar power generation system and the input power source 30 is a wind power generation system, but they are not limited to these and may be various renewable energy power generation systems.

[0022] The input power source 20 includes a solar cell 21 and a power converter 22. The input power source 30 includes a wind power generator 31 and a power converter 32. The number of input power sources 20 and 30 is not limited to one each, and there may be a plurality of input power sources.

[0023] The renewable energy power supply system connected to the DU-EMS 4 may be one that utilizes energy such as wave power or geothermal power in addition to the power supply systems described above, or may be a power supply system that utilizes hydroelectric power, small hydroelectric power, tidal power, tidal current power, or temperature difference power generation, or may be a combination of the power supply systems described above.

[0024] The primary stabilization device 40 includes a secondary battery 41, which is a first charge / discharge element, and a power converter 42, which is a first power converter. The number of primary stabilization devices 40 is not limited to one, and multiple primary stabilization devices 40 may be provided. The primary stabilization device 40 sets a variable bus voltage target value within a predetermined allowable range centered on the voltage of the DC bus 70, and controls the charge / discharge of the secondary battery 41 by operating the power converter 42 so that the output voltage on the DC bus 70 side matches the bus voltage target value.

[0025] The semi-stabilizing device 50 includes a hydrogen storage facility 51, which is a second charging / discharging element, and a power converter 52, which is a second power converter. The hydrogen storage facility 51 is a charging element that produces and stores hydrogen using surplus power generated in the power system 100. In the semi-stabilizing device 50, the power converter 52 performs DC / DC conversion so that the output current matches an output current target value calculated based on the voltage of the DC bus 70. The number of semi-stabilizing devices 50 is not limited to one, and multiple semi-stabilizing devices 50 may be used.

[0026] The semi-stabilizer 60 includes a fuel cell 61, which is a second charge / discharge element, and a power converter 62, which is a second power converter. The semi-stabilizer 60 supplies DC power generated by an electrochemical reaction in the fuel cell 61, which is a discharge element, to the DC bus 70 via the power converter 62. In the semi-stabilizer 60, the power converter 52 performs DC / DC conversion so that the output current matches an output current target value calculated based on the voltage of the DC bus 70. The number of semi-stabilizers 60 is not limited to one, and multiple semi-stabilizers 60 may be used.

[0027] Furthermore, the DU-EMS 4 includes a monitoring and command device 90. The monitoring and command device 90 collects status information of each of the input power source 30, the primary stabilization device 40, the semi-stabilization device 50, and the semi-stabilization device 60 (hereinafter referred to as "primary stabilization devices, etc.") to monitor their status and operation. The status information includes, for example, voltage, current, temperature, etc. The monitoring and command device 90 also generates operation commands and charge / discharge threshold commands for each component based on the results of monitoring the primary stabilization devices, etc. Note that communication between the monitoring and command device 90 and the primary stabilization devices, etc., can be performed via wire or wirelessly.

[0028] 4 is a hardware configuration diagram of an information knowledge system 10 according to an embodiment of the present invention. The information knowledge system 10 functions as a computer. The information knowledge system 10 includes a CPU 11, a RAM 12, a ROM 13, and an I / O 14, which are interconnected by a bus.

[0029] The CPU 11 uses the RAM 12 as a work memory and executes the program 15 to control the entire information-knowledge system 10. The ROM 13 is a non-volatile memory such as a flash memory, and stores the program 15. The CPU 11 executes the program 15 to provide the following functions.

[0030] The I / O 14 is an input / output interface. The program 15 may include a learned model trained by machine learning to perform segmentation processing that functions as the knowledge-based vocabulary management unit 1 and the information knowledge unit 2.

[0031] 5 and 6 are diagrams showing an example of a class hierarchical structure in an information processing device according to an embodiment of the present invention. The illustrated hierarchical structure shows a part of the hierarchical structure of ontology classes, and may be a tree structure as shown. The illustrated hierarchical structure is a hierarchical definition of classes, which are sets of data that generalize components in the power system 100, by the knowledge vocabulary management unit 1 of the information knowledge device 10. The definition of each class may be defined in advance or may be generated by a trained model.

[0032] When defining each class and property, identification information such as a Uniform Resource Identifier (URI) is assigned to eliminate vocabulary ambiguity. It is also possible to add and arrange each class and property in a tree structure. The objects of the illustrated hierarchical structure are "System," "SystemComponent," and their subordinate classes. The superior-subordinate relationship between classes can be defined using, for example, the rdfs:subClassOf property.

[0033] The characteristics of the devices that make up the system may be defined as device requirements such as function and performance. As an example of a definition, if the "class" is "sensor," various definitions are possible, such as the "function" being "measurement" and the "measurement target" being "pressure."

[0034] FIG. 7 is a diagram showing an example of a data image of a measurement result information file in the power system 100 shown in FIG. 2. The illustrated example aggregates the measurement results of the voltage or current of individual sensors at a specific date and time, with each column representing the measurement result of an individual sensor. That is, the values ​​in the BatteryVoltage, BatteryCurrent, FCVoltage, and FCCurrent columns are the measurement values ​​of the battery voltage, battery current, fuel cell voltage, and fuel cell current, respectively. The data format of the measurement result information file may be various formats, such as CSV.

[0035] FIG. 8 is a diagram showing an example of a metadata schema for converting measurement result information in the power system 100 shown in FIG. 2 into knowledge. The illustrated schema defines the type and format of information for describing measurement result information as metadata using vocabulary defined by ontology. The schema may be written using XML, but is not limited to this. Metadata for each piece of measurement result information is written according to the above schema.

[0036] FIG. 9 is a diagram illustrating an example of a schema of metadata for converting sensor information in the power system 100 illustrated in FIG. 2 into knowledge. The illustrated schema, like the example of FIG. 8, defines the type and format of information to be described as metadata using vocabulary defined by ontology. The schema may be described using XML, but is not limited to this. The metadata of each sensor is described according to the above schema.

[0037] FIG. 10 is a diagram showing an example of metadata for measurement result information in the power system 100 shown in FIG. 2. The table in FIG. 10 is an example in which metadata using common vocabulary of ontology is assigned to each piece of information. In the figure, "2405230800001" of "file" indicates an individual of the file class. Also, in the figure, "MeasurementResultByTime_202405230800001-001" indicates an individual of the "measurement result information by measurement date and time class," and "MeasurementResultInfoBySensor_202405230800001-s1" indicates an individual of the "measurement result information by sensor class."

[0038] An example of using the metadata added in the information knowledge system 10 according to this embodiment will be described below.

[0039] The information knowledge system 10 according to this embodiment allows data collected from different information sources to be integrated and organized using metadata, and then stored and utilized. As an example, the information knowledge system 2 assigns metadata of a common schema, such as the type of data, units, measurement time, and measurement location, to measurement data collected from a large number of different sensors. For example, the data is stored after aligning descriptions such as the units and values ​​of measurement values ​​and the notation of measurement time, and data can be searched and extracted as needed.

[0040] Another example is grouping various data according to data patterns. For example, the information knowledge unit 2 assigns metadata to measured values ​​of meteorological data for each region in the country. In this case, fluctuations in measured values ​​over a predetermined period can be analyzed to group regions with similar patterns. By grouping in this way, for example, to address missing meteorological data in a certain region, such as a lack of sunshine, the missing data can be substituted or supplemented with estimated values ​​based on meteorological data from other regions.

[0041] Furthermore, for example, the information knowledge generation unit 2 assigns metadata to the sensor measurement values. In this case, it is possible to analyze fluctuations in the measurement values ​​over a predetermined period and group sensors with similar patterns. By grouping in this way, for example, when data from a certain sensor is lacking, it is possible to substitute or supplement the data with estimated values ​​based on data from other sensors with similar patterns.

[0042] Furthermore, the information knowledge unit 2 can analyze the amount of power usage and the usage status of the power system 100 over a predetermined period, and group users according to how they use the power system 100. Based on such grouping, further grouping by regional characteristics can be performed, making it possible to clarify the differences between how the power system 100 is used in each region and how it is used nationwide.

[0043] Furthermore, the information knowledge unit 2 can group users based on a combination of metadata related to user attributes such as the amount of electricity used, the usage status of the power system 100, charges, regional characteristics, and whether the user has a morning or evening lifestyle, etc. By grouping in this way, for the conditions of the specified combination of metadata, metadata such as the region where the user lives, user attributes, and the configuration and settings of the power system 100 can be used as source data for setting initial values ​​and a model case for how to use the power system 100.

[0044] As yet another example, the information knowledge generation unit 2 can analyze the pattern of change in measurement values ​​based on the measurement values ​​for a predetermined period to which metadata is attached. A range of normal values ​​can be defined in the metadata schema, and the detection of an abnormal value outside the range among the measurement values ​​can be used as a trigger to transition the operation of the power system 100 to an abnormal mode.

[0045] The information knowledge generation device 10 according to the present embodiment includes a trained model. The trained model according to the embodiment acquires information about different types of power systems, assigns metadata to each piece of information using a common vocabulary of an ontology, and converts the information into knowledge.

[0046] Furthermore, the information knowledge generation device 10 according to the present embodiment includes an embodiment of a method for generating a trained model. The method for generating a trained model according to the embodiment includes the steps of acquiring multiple types of data related to power systems. The method for generating a trained model according to the embodiment includes the steps of acquiring information about different types of power systems, assigning a unified meaning to the information about the different types of power systems in a machine-readable form, managing an ontology that defines a common vocabulary required for knowledge generation, and assigning metadata using the common vocabulary of the ontology to each piece of information to generate knowledge about the information.

[0047] <Effects of the Information Knowledge System 10 According to the Embodiment> The information knowledge system 10 assigns a unified meaning to information relating to different types of power systems in a machine-readable form, manages an ontology that defines the common vocabulary required for knowledge generation, and adds metadata using the common vocabulary of the ontology to each piece of information to convert the information into knowledge. This makes it possible to integrate multiple types of data. This reduces the complexity involved in handling diverse data, as occurs when managing data in a database, for example. Therefore, the information knowledge system 10 according to the present embodiment makes it possible to configure a system that can easily accommodate the diversity of EMS-related devices and data.

[0048] Furthermore, the power system 100 including the information knowledge device 10 is capable of controlling the supply and demand of power for a large number of power generation systems, which makes it easier to respond to demand response, in which a power company requests a large number of consumers to reduce their power demand when there is a drop in solar power generation output or when there is a power shortage.

[0049] Therefore, the power system 100 can easily accommodate the addition of power systems with various power equipment configurations in various locations, and can control these facilities, predict power generation amount, and predict demand for the generated power. Also, the configuration of the power system 100 can be defined not as a combination of specific equipment, but as requirements for the equipment included in the power system 100. Furthermore, the information knowledge device 10 according to this embodiment can be used for various simulations based on a logical structure corresponding to the physical structure of the system.

[0050] The information knowledge system 10 according to the present embodiment has been described as realizing the power system 100, but it can be used in a variety of fields, not limited to power systems with various power facility configurations.

[0051] Although the embodiments have been described above, the present invention is not limited to the above-described embodiments, and various modifications and improvements are possible within the scope of the present invention.

[0052] Each function of the above-described embodiments can be realized by one or more processing circuits. Here, the term "processing circuit" includes a processor programmed to perform each function by software, such as a processor implemented by an electronic circuit, as well as devices such as an ASIC (Application Specific Integrated Circuit), a DSP (Digital Signal Processor), an FPGA (Field Programmable Gate Array), or a conventional circuit module designed to perform each of the above-described functions.

[0053] This application claims priority from Japanese Patent Application No. 2024-112914, filed on July 12, 2024, with the Japan Patent Office, the entire contents of which are incorporated herein by reference.

[0054] 1...knowledge vocabulary management unit, 2...information knowledge unit, 10...information knowledge device, 20, 30...input power source, 40...main stabilizer, 50, 60...quasi stabilizer, 70...DC bus, 80...load, 90...monitoring and indicating device, 100...power system, N...power transmission network

Claims

1. An information knowledge device comprising: a knowledge vocabulary management unit that assigns unified meaning to information relating to different types of power systems in a machine-readable form and manages an ontology that defines the common vocabulary necessary for knowledge creation; and an information knowledge unit that assigns metadata using the common vocabulary of the ontology to each piece of information and knowledgeizes the information.

2. The information knowledge system according to claim 1, wherein the information relating to the different types of power systems includes at least equipment information and measurement data relating to the power systems.

3. The information knowledge device according to claim 1 or 2, wherein the power system includes an integrated EMS and distributed unit EMSs, each of which includes at least one of a main stabilizing device having a first charge / discharge element and a first power converter, a second charge / discharge element, at least one semi-stabilizing device having a charge element or discharge element and a second power converter, and an input power source which is a renewable energy power source system, and a monitoring and instruction device which monitors status information of the distributed unit EMSs and gives instructions.

4. An information knowledge method performed by an information knowledge device, comprising the steps of: assigning unified meaning to information relating to different types of power systems in a machine-readable form and managing an ontology that defines the common vocabulary required for knowledge creation; and assigning metadata to each piece of information using the common vocabulary of the ontology, and knowledge-creating the information.

5. A trained model that acquires information about heterogeneous power systems, assigns a unified meaning to the information about said heterogeneous power systems in a machine-readable form, manages an ontology that defines the common vocabulary necessary for knowledge generation, and assigns metadata to each piece of information using the common vocabulary of the ontology to generate knowledge about the information.

6. A method for generating a trained model, comprising the steps of: acquiring information relating to heterogeneous power systems; assigning a unified meaning to the information relating to the heterogeneous power systems in a machine-readable form and managing an ontology that defines the common vocabulary required for knowledge generation; and assigning metadata to each piece of information using the common vocabulary of the ontology and knowledge generation of the information.

7. A non-transitory computer-readable medium storing a program that causes an information knowledge generation device to function, the non-transitory computer-readable medium storing a program for executing the following processes: acquiring information about heterogeneous power systems; assigning a unified meaning to the information about the heterogeneous power systems in a machine-readable form and managing an ontology that defines the common vocabulary necessary for knowledge generation; and assigning metadata to each piece of information using the common vocabulary of the ontology to generate knowledge about the information.

Images