Location aptitude evaluation system, and location aptitude evaluation method

Abstract

To support operations related to evaluation of the aptitude for the location of a plant.SOLUTION: A location aptitude evaluation system stores a location aptitude evaluation model that is a machine learning model trained by using, as learning data, a data set formed with an aerial image as an explanatory variable and information on a feature used for evaluation of the aptitude for the location of a predetermined plant as an objective variable, and inputs an aerial image capturing an evaluation target that is the land or region to be evaluated in terms of the aptitude for the location of the predetermined plant and the periphery environment thereof, to the location aptitude evaluation model as an explanatory variable, to generate information based on a value output from the location aptitude evaluation model as an evaluation result.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a site suitability evaluation system and a site suitability evaluation method. [Background technology]

[0002] Patent Document 1 describes a land suitability evaluation device that appropriately evaluates the suitability of land for agricultural crops. The land suitability evaluation device stores the location, climate, and topography of land in association with each other, acquires the location of a specific piece of land in response to a user operation, acquires the climate and topography of the specific piece of land, and calculates an evaluation value indicating the suitability of the specific piece of land for agricultural crops from the location of the specific piece of land using an evaluation model that has been machine-learned using a score indicating the performance of the agricultural crop and the climatic and topographical conditions of the cultivation area of ​​the agricultural crop as training data.

[0003] Non-Patent Document 1 describes a technology for identifying features captured in an aerial image using a deep learning model (CNN (Convolutional Neural Network)). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-39753 [Non-patent literature]

[0005] [Non-Patent Document 1] Hiroyuki Ohno, “Research on automatic feature extraction using AI”, [online], Research period (April 2018 to March 2023), Geospatial Information Authority of Japan, [Searched on August 7, 2023], Internet<URL:https: / / www.gsi.go.jp / common / 000248621.pdf> Summary of the Invention [Problem to be solved by the invention]

[0006] A plant must be built on suitable land or in an appropriate area, taking into consideration its size, economic impact, and impact on the surrounding environment. For example, thermal and nuclear power plants require large amounts of cooling water to cool the steam that spins the turbines and turn it back into water, so there must be a water source nearby. Nuclear power plants must also be built in places that will have as little impact on the surrounding area as possible, assuming the release of radioactive materials in the unlikely event of an accident. Thus, when assessing the suitability of a land or area for a plant site, a comprehensive judgment of various conditions is required.

[0007] Satellite images could be used to evaluate the suitability of such land or regions for plant locations, but using satellite images to manually identify the presence of features and the surrounding environment in the land or region and to evaluate the suitability of the plant location would require a great deal of time and effort.

[0008] In the above-mentioned Patent Document 1, an evaluation model is machine-learned using a score indicating the performance of the agricultural crop and the climatic and topographical conditions of the cultivation area of ​​the agricultural crop as training data, and an evaluation value indicating the suitability of a specific piece of land for the agricultural crop is calculated from the position data of the specific piece of land. However, the technology described in the document does not evaluate the site suitability of a plant, nor does it evaluate the site suitability based on satellite images. In addition, the technology described in Non-Patent Document 1 merely identifies features captured in an aerial image by deep learning, and does not evaluate the site suitability of a plant.

[0009] The present invention has been made in consideration of the above background, and aims to provide a site suitability evaluation system and a site suitability evaluation method that are capable of supporting work related to evaluating the site suitability of a plant. [Means for solving the problem]

[0010] One of the present inventions for achieving the above-mentioned object is a location suitability evaluation system which stores a location suitability evaluation model, which is a machine learning model trained by using as training data a dataset consisting of aerial images as explanatory variables and information on features used to evaluate the location suitability of a specified plant as the objective variable, and inputs aerial images showing the evaluation target, which is the land or area to be evaluated for the location suitability of a specified plant, and its surrounding environment, into the location suitability evaluation model as explanatory variables, thereby generating information based on the values ​​output by the location suitability evaluation model as evaluation results.

[0011] Other problems and solutions disclosed in the present application will be made clear in the detailed description and drawings. Effect of the Invention

[0012] According to the present invention, it is possible to assist in the work of evaluating the suitability of a plant site. [Brief description of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram showing a schematic configuration of a location suitability evaluation system. [Diagram 2] FIG. 2 is a diagram showing main functions of a location suitability evaluation device. [Diagram 3] 1 is an example of learning data. [Figure 4] FIG. 1 is a diagram illustrating a location suitability determination model. [Figure 5A] 13 is a flowchart illustrating a model learning process. [Figure 5B] 13 is a flowchart illustrating a location suitability evaluation process. [Figure 6] 13 is an example of an evaluation result presentation screen. [Figure 7] 1 is an example of a hardware configuration of an information processing device used to realize a satellite image providing device or a location suitability evaluation device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] Hereinafter, an embodiment will be described with reference to the drawings. In the following description, the letter "S" before a reference numeral denotes a processing step.

[0015] 1 shows a schematic configuration of a system for evaluating the suitability of a plant site (hereinafter referred to as "site suitability evaluation system 1"), which will be explained as one embodiment of the present invention. The site suitability evaluation system 1 provides information related to the evaluation of the site suitability of a plant for land or area (hereinafter referred to as "evaluation target") that is to be evaluated for the suitability of a plant site, such as a candidate site for the construction of the plant.

[0016] The location suitability evaluation system 1 includes an information processing device (hereinafter referred to as a "satellite image providing device 3") that provides images of the earth's surface taken from an artificial satellite (hereinafter referred to as a "satellite image"), and an information processing device (hereinafter referred to as a "location suitability evaluation device 100") that is communicatively connected to the satellite image providing device 3 via a communication network 5.

[0017] In the following, we will use as an example a case where the plant is a power plant (a thermal power plant, a nuclear power plant, a hydroelectric power plant, a geothermal power plant, a power plant that uses renewable energy, etc.), but the plant may be something other than a power plant.

[0018] The communication network 5 is a wired or wireless communication infrastructure, such as the Internet, a local area network (LAN), a wide area network (WAN), various public communication networks, and dedicated lines.

[0019] The satellite image providing device 3 is operated by an organization such as a company that provides satellite images. The satellite image providing device 3 provides satellite images taken by an artificial satellite to the location suitability evaluation device 100 via the communication network 5 as image data in a predetermined data format.

[0020] The location suitability evaluation device 100 is operated and used by a person (hereinafter referred to as "user 7") who evaluates the location suitability of a candidate site to be evaluated for a plant when selecting a proposed construction site for the plant. The location suitability evaluation device 100 generates information (hereinafter referred to as "evaluation result") related to the evaluation of the location suitability of the candidate site by inputting a satellite image showing the candidate site designated by the user 7 into a machine learning model, and provides information based on the generated evaluation result to the user.

[0021] 2 shows the main functions of the location suitability evaluation device 100. As shown in the figure, the location suitability evaluation device 100 has the functions of a memory unit 110, an information acquisition management unit 125, a preprocessing unit 130, an annotation information generation unit 135, a learning data generation unit 140, a model learning unit 150, a location suitability evaluation unit 155, and an evaluation result presentation unit 160.

[0022] Of these, the memory unit 110 stores each piece of information (data) including a training satellite image 111, a preprocessed satellite image 112, annotation information 113, training data 114 (including validation data), a location suitability evaluation model 115, an evaluation target satellite image 116, a preprocessed evaluation target satellite image 117, and an evaluation result 118.

[0023] The information acquisition and management unit 125 acquires information necessary to realize the various functions of the location suitability evaluation device 100 , such as information used to generate the learning data 114 , and manages the acquired information in the storage unit 110 .

[0024] For example, the information acquisition management unit 125 acquires satellite images (hereinafter referred to as “training satellite images”) used to generate the training data 114 from the satellite image providing device 3, and manages the acquired training satellite images as training satellite images 111.

[0025] Furthermore, for example, the information acquisition management unit 125 acquires information for adjusting hyperparameters and feature quantities used in learning the location suitability evaluation model 115 from the user 7.

[0026] In addition, for example, the information acquisition management unit 125 accepts specification or input from the user 7 of a satellite image containing the evaluation target (hereinafter referred to as the “evaluation target satellite image”), and manages the accepted satellite image as the evaluation target satellite image 116.

[0027] The preprocessing unit 130 performs preprocessing (resizing, noise removal, contrast adjustment, etc.) on the learning satellite image, and manages the preprocessed image as the preprocessed satellite image 112.

[0028] The preprocessing unit 130 also performs preprocessing (resizing, noise removal, contrast adjustment, etc.) on the satellite image to be evaluated, and manages the preprocessed image as the satellite image to be evaluated after preprocessing 117 .

[0029] The learning satellite image and the evaluation target satellite image are, for example, image data cut out as rectangular meshes of a predetermined size with a common scale.

[0030] The learning satellite images include, for example, satellite images that show an existing plant (e.g., a rectangular image of a specified size that shows a plant in the center and a specified area surrounding the plant) and satellite images that do not show existing plants (e.g., a rectangular image of a specified size that shows a vacant lot in the center and a specified area surrounding the vacant lot).

[0031] The satellite image to be evaluated is a rectangular image of a given size, with the evaluation target in the center and a given area around the evaluation target. By matching the scale and size of the satellite image to be evaluated with those of the training satellite image, it is possible to appropriately determine whether or not a given feature is around the evaluation target (within a given distance range).

[0032] The annotation information generation unit 135 acquires information used for annotation (information used to generate the learning data 114 (information used as explanatory variables and objective variables (labels)) hereinafter referred to as "annotation information") from the satellite image providing device 3 and the user 7, and manages the acquired information as annotation information 113. Note that the annotation information generation unit 135 may generate the above information using, for example, a known object detection mechanism (such as R-CNN (Regional CNN (Convolutional Neural Network)), YOLO (You only Look once), SSD (Single Shot Detector), DETR (DEtection TRansformer), etc.).

[0033] The training data generation unit 140 generates a data set in which labels are associated with the training satellite images 111, and manages the generated data set as training data 114. The training data generation unit 140 generates labels, for example, by using the annotation information 113. In addition, the training data generation unit 140 expands the training data 114 by increasing the number of training satellite images 111, for example, by using a Generative Adversarial Network (GAN) or various image processing techniques (up / down / left / right flipping, vertical / horizontal shift, rotation, channel shift, image deformation, etc.).

[0034] 3 shows an example of the learning data 114. The learning data 114 shown in the example includes a plurality of records (datasets) in which an identifier of a learning satellite image (hereinafter referred to as "image data ID"), which is an explanatory variable, is associated with a target variable (presence or absence of a power plant, type of power plant (0: thermal power plant, 1: nuclear power plant, ~), presence or absence of a water intake source (1: present, 0: absent), presence or absence of an apartment building (1: present, 0: absent), presence or absence of an industrial area (1: present, 0: absent), presence or absence of a railway station (1: present, 0: absent), etc.). In the case of the learning data 114 shown in the figure, the entity of the satellite image (image data) is associated with the image data ID and managed separately.

[0035] The model learning unit 150 uses the learning data 114 to learn the location suitability evaluation model 115, which is a machine learning model. For example, a deep neural network (DNN), a convolutional neural network (CNN), a multilayer perceptron (MLP), a support vector machine, a decision tree, a random forest, gradient boosting, or the like is used as the location suitability evaluation model 115. The location suitability evaluation model 115 may be configured by combining a plurality of machine learning models in stages. For example, when using a DNN or a CNN, for example, a ReLU or a softmax function is used as an activation function, and for example, a cross-entropy is used as a loss function. The model learning unit 150 verifies the learned location suitability evaluation model 115 by an appropriate verification method (hold-out method, cross-validation method, or the like).

[0036] FIG. 4 is a diagram showing an example of a location suitability evaluation model 115 (neural network structure) configured using DNN. As shown in the figure, the feature quantities of the satellite image to be evaluated, which are explanatory variables, are input to an input layer 411 of the location suitability evaluation model 115. The intermediate layer 412 includes one or more hidden layers consisting of one or more nodes including parameters adjusted by learning. The intermediate layer 412 obtains one or more predicted values ​​(probabilities) for the output layer 413 based on the input feature quantities. The output layer 413 outputs values ​​(probability values) of various objective variables (such as the location suitability for each type of power plant, the presence or absence of a water intake source, the presence or absence of an apartment building, the presence or absence of an industrial area, the presence or absence of a train station, etc.).

[0037] Returning to Figure 2, the location suitability evaluation unit 155 inputs the evaluation target satellite image 116 received by the information acquisition management unit 125 from the user 7 into the location suitability evaluation model 115, thereby generating information based on the values ​​output by the location suitability evaluation model 115, and manages the generated information as evaluation result 118.

[0038] The evaluation result presenting unit 160 presents the evaluation result 118 to the user 7. The evaluation result presenting unit 160 generates a screen on which the evaluation result 118 is described, and presents it to the user 7, for example.

[0039] 5A is a flowchart for explaining an example of processing (hereinafter referred to as "model learning processing S500") that the location suitability evaluation device 100 performs when learning the location suitability evaluation model 115. Hereinafter, the model learning processing S500 will be explained with reference to the same figure.

[0040] First, the information acquisition management unit 125 acquires a training satellite image from the satellite image providing device 3, and manages the acquired training satellite image as the training satellite image 111 (S511).

[0041] Next, the preprocessing unit 130 preprocesses the learning satellite image acquired by the information acquisition management unit 125, and manages the preprocessed satellite image (hereinafter referred to as the "preprocessed satellite image") as the preprocessed satellite image 112 (S512).

[0042] Next, the annotation information generating unit 135 generates annotation information based on the preprocessed satellite image 112, and manages the generated annotation information as annotation information 113 (S513).

[0043] Next, the training data generation unit 140 generates a data set based on the preprocessed satellite image 112 and the annotation information 113, and manages the generated data set as training data 114 (S514).

[0044] Next, the model learning unit 150 uses the learning data 114 to learn the location suitability evaluation model 115 (S515).

[0045] This ends the model learning process S500.

[0046] 5B is a flowchart for explaining an example of processing (hereinafter referred to as "location suitability evaluation processing S550") performed when the location suitability evaluation device 100 generates an evaluation result of an evaluation target using the location suitability evaluation model 115. Hereinafter, the location suitability evaluation processing S550 will be explained with reference to the same figure.

[0047] First, the information acquisition management unit 125 accepts a designation or input of an evaluation target satellite image from the user 7, and manages the accepted evaluation target satellite image as the evaluation target satellite image 116 (S551).

[0048] Next, the preprocessing unit 130 preprocesses the evaluation target satellite image acquired by the information acquisition management unit 125, and manages the preprocessed evaluation target satellite image as the preprocessed evaluation target satellite image 117 (S552).

[0049] Next, the location suitability evaluation unit 155 inputs the preprocessed evaluation target satellite image 117 to the location suitability evaluation model 115, and manages the information output by the location suitability evaluation model 115 as the evaluation result 118 (S553).

[0050] Next, the evaluation result presentation unit 160 generates a screen (hereinafter referred to as the "evaluation result presentation screen 600") that provides the user with information based on the evaluation result 118, and outputs the generated evaluation result presentation screen 600 (S554).

[0051] This ends the location suitability evaluation process S550.

[0052] 6 is an example of a screen (hereinafter referred to as an "evaluation result presentation screen 600") that is displayed when the location suitability evaluation device 100 provides a user with information based on the evaluation result 118. As shown in the figure, the illustrated evaluation result presentation screen 600 has a designation field 611 for the satellite image to be evaluated, an evaluation execution button 612, a display field 613 for the satellite image to be evaluated, and a display field 614 for the evaluation result.

[0053] When the user 7 specifies image data of the satellite image to be evaluated in the specification field 611 of the satellite image to be evaluated and operates the evaluation execution button 612, an image of the satellite image to be evaluated is displayed in the display field 613 of the satellite image to be evaluated, and information based on the evaluation result 118 is displayed in the display field 614 of the evaluation result. The user 7 can, for example, efficiently evaluate the location suitability of candidate sites by referring to the displayed information and select land or area on which to build the plant.

[0054] As described above, according to the location suitability evaluation system 1 of this embodiment, the user 7 can obtain information on the location suitability of the plant for the evaluation target by specifying or inputting the satellite image of the evaluation target to the location suitability evaluation device 100, and can use the obtained information as reference information when determining the location suitability of the plant. In this way, the location suitability evaluation system 1 of this embodiment can support the work related to the evaluation of the location suitability of the plant.

[0055] <Example of information processing device> FIG. 7 shows an example of a hardware configuration of an information processing device used to realize the satellite image providing device 3 and the location suitability evaluation device 100. The illustrated information processing device 10 includes a processor 11, a main memory device 12, an auxiliary memory device 13, an input device 14, an output device 15, and a communication device 16. Specific examples of the information processing device 10 include personal computers, office computers, various server devices, general-purpose machines, and the like. The information processing device 10 may be realized in whole or in part using virtual information processing resources provided using a virtualization technology, a process space separation technology, or the like, such as a virtual server provided by a cloud system. The satellite image providing device 3 and the location suitability evaluation device 100 may be realized using a plurality of information processing devices 10 connected to each other so as to be able to communicate with each other.

[0056] In the figure, the processor 11 is configured using, for example, a CPU (Central Processing Unit), an MPU (Micro Processing Unit), a GPU (Graphics Processing Unit), an FPGA (Field Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), an AI (Artificial Intelligence) chip, etc.

[0057] The main memory device 12 is a device that stores programs and data, and is, for example, a Read Only Memory (ROM), a Random Access Memory (RAM), or a non-volatile memory (Non Volatile RAM (NVRAM)).

[0058] The auxiliary storage device 13 is, for example, an SSD (Solid State Drive), a hard disk drive, an optical storage device (CD (Compact Disc), DVD (Digital Versatile Disc), etc.), a storage system, an IC card, a read / write device for a recording medium such as an SD card or an optical recording medium, a storage area of ​​a cloud server, etc. Programs and data can be read into the auxiliary storage device 13 via a recording medium reader or a communication device 16. The programs and data stored (memorized) in the auxiliary storage device 13 are read into the main storage device 12 as needed.

[0059] The input device 14 is an interface that accepts input from the outside, and is, for example, a keyboard, a mouse, a touch panel, a card reader, a pen-input tablet, a voice input device, or the like.

[0060] The output device 15 is an interface that outputs various information such as the progress of processing and the results of processing. The output device 15 is, for example, a display device (LCD (Liquid Crystal Display), a graphic card, etc.) that visualizes the various information described above, a device that converts the various information described above into voice (voice output device (speaker, etc.)), or a device that converts the various information described above into text (printer, etc.). Note that, for example, the information processing device 10 may be configured to input and output information to and from other devices via the communication device 16.

[0061] The input device 14 and the output device 15 constitute a user interface for receiving information from the user and presenting information.

[0062] The communication device 16 is a device that realizes communication (wired communication or wireless communication) with other devices via a communication infrastructure such as a communication network 5, and is configured using, for example, a NIC (Network Interface Card), a wireless communication module, a USB module, etc.

[0063] The information processing device 10 may be implemented with, for example, an operating system, a file system, a DBMS (DataBase Management System) (relational database, NoSQL, etc.), a KVS (Key-Value Store), etc.

[0064] The functions of the satellite image providing device 3 and the location suitability evaluation device 100 are realized by the processor 11 of the information processing device 10 reading and executing a program stored in the main memory device 12, or by the functions of the hardware (FPGA, ASIC, AI chip, etc.) that constitutes the satellite image providing device 3 and the location suitability evaluation device 100. The satellite image providing device 3 and the location suitability evaluation device 100 store the various information (data) described above, for example, as a table in a database or a file managed by a file system.

[0065] Although the embodiment of the present invention has been described above in detail, the above description is for the purpose of facilitating understanding of the present invention and does not limit the present invention. The present invention may be modified or improved without departing from the spirit thereof, and the present invention naturally includes equivalents thereof. For example, the above embodiment has been described in detail to explain the present invention in an easily understandable manner, and is not necessarily limited to those having all of the configurations described. In addition, it is possible to add, delete, or replace part of the configuration of the above embodiment with other configurations.

[0066] For example, in the above, an example was given of using satellite images as explanatory variables, but images of the earth's surface taken from an aircraft (aerial images), or images of the earth's surface taken from the sky using methods other than artificial satellites (hereinafter referred to as "aerial images"), may also be used as explanatory variables.

[0067] 6, the user 7 specifies or inputs the satellite image to be evaluated, but the location suitability evaluation device 100 may automatically evaluate each of a large number of image data previously taken of a large number of lands and regions by the above-mentioned mechanism, and the location suitability evaluation device 100 may automatically select image data with a high overall evaluation (excellent location suitability) based on each evaluation result 118 and present the optimal solution to the user 7 (or list the lands and regions in order of highest evaluation). In that case, the above-mentioned comprehensive evaluation may be performed by taking into account weighting set by the user 7, etc. for each of the values ​​(probability values) of the objective variables. [Explanation of symbols]

[0068] 1. Location Suitability Evaluation System 2 Satellite image provider 3 Satellite image providing device 5. Communication Network 7 Users 100 Location Suitability Evaluation Device 110 Storage section 111 Educational Satellite Images 112 Pre-processed satellite image 113 Annotation Information 114 Training Data 115 Location Suitability Evaluation Model 116 Satellite images to be evaluated 117 Satellite images to be evaluated after pre-processing 118 Evaluation Results 125 Information Acquisition Management Department 130 Pre-processing section 135 Annotation Information Generation Unit 140 Learning Data Generation Unit 150 Model Learning Department 155 Location Suitability Evaluation Department 160 Evaluation result presentation section S500 Model learning process S550 Location suitability evaluation processing 600 Evaluation result display screen

Claims

1. The information processing device includes a processor and a storage device. A location suitability evaluation model is stored, which is a machine learning model trained by using, as training data, a data set configured with an aerial image as an explanatory variable and information on features used to evaluate the location suitability of a specified plant as a target variable; aerial images showing an evaluation target, which is a land or area to be evaluated for the site suitability of a specified plant, and its surrounding environment are input as explanatory variables into the site suitability evaluation model, and information based on values ​​output by the site suitability evaluation model is generated as an evaluation result; Location suitability evaluation system.

2. The location suitability evaluation system according to claim 1, The objective variable of the learning data includes information indicating the presence or absence of each type of the predetermined plant and the presence or absence of each type of feature existing around the predetermined plant. Location suitability evaluation system.

3. The location suitability evaluation system according to claim 1, The predetermined plant is a power plant. Location suitability evaluation system.

4. The location suitability evaluation system according to claim 3, The power plant is at least one of a thermal power plant, a nuclear power plant, a hydroelectric power plant, a geothermal power plant, and a power plant that uses renewable energy. Location suitability evaluation system.

5. The location suitability evaluation system according to claim 4, The feature is at least one of a water intake, a housing complex, an industrial area, and a railway station; Location suitability evaluation system.

6. The location suitability evaluation system according to claim 1, The aerial images used as the explanatory variables of the learning data are The method includes an aerial image showing the evaluation target and its surrounding environment, and an aerial image showing land or an area where the specified plant does not exist and its surrounding environment. Location suitability evaluation system.

7. The location suitability evaluation system according to claim 1, The aerial image is at least one of a satellite image of the earth's surface taken from an artificial satellite and an aerial image of the earth's surface taken from an aircraft. Location suitability evaluation system.

8. An information processing device having a processor and a storage device, A step of storing a location suitability evaluation model, which is a machine learning model trained by using a data set configured with an aerial image as an explanatory variable and information on features used to evaluate the location suitability of a predetermined plant as a target variable, as training data; and a step of inputting an aerial image of an evaluation target, which is a land or area to be evaluated for the site suitability of a specified plant, and its surrounding environment into the site suitability evaluation model as an explanatory variable, thereby generating information based on values ​​output by the site suitability evaluation model as an evaluation result; A method for evaluating location suitability.

9. The method for evaluating site suitability according to claim 8, The objective variable of the learning data includes information indicating the presence or absence of each type of the predetermined plant and the presence or absence of each type of feature existing around the predetermined plant. Methods for assessing site suitability.

10. The method for evaluating site suitability according to claim 8, The predetermined plant is a power plant. Methods for assessing site suitability.

Images