System and Method for estimating electric power demand through land cover information based on image recognition with terminal for outputting prediction information

Abstract

A power demand forecasting system is disclosed that can selectively apply a standard load density based on object recognition feature information regarding infrastructure constructed within a complex, using image recognition-based land cover information for satellite and aerial images of a newly developed complex. The power demand forecasting system is characterized by comprising an image information providing server that generates image information, and a power forecasting server that generates image recognition-based land cover information regarding the image information via a communication network, and generates power demand forecasting information by selectively applying a standard load density based on object recognition feature information regarding infrastructure constructed within a specific complex using the land cover information.

Description

Technology Field

[0001] The present invention relates to a power demand forecasting technology, and more specifically, to a power demand forecasting system and method using image recognition-based land cover information for a newly developed complex. Background Technology

[0002] Recently, in accordance with the government's policies for balanced national development and economic revitalization, development complexes are being actively established, with the designation and development of new complexes continuously proceeding nationwide. Furthermore, as industrial complexes continue to increase and the designation of residential districts remains steady, there is a need for highly reliable forecasting of electricity demand for these new complexes to plan power facilities and respond to changes in power requirements during development.

[0003] The existing power demand forecasting method establishes and applies criteria for determining the appropriate scale of new electricity supply facilities in areas where new housing and industrial complexes are being developed, and applies mathematical formulas to housing complexes by classifying them into residential and non-residential zones.

[0004] For residential land, standard load densities and diversity ratios are applied according to housing classification (single-family, multi-family) and regional classification (Seoul Metropolitan Area, Metropolitan City, Others); for non-residential land, standard load densities and diversity ratios are applied according to land use and specific building use.

[0005] In addition, since the electricity demand of industrial complexes varies depending on each company's site utilization plan and technological level, making accurate forecasting difficult, standard load density criteria by industry are established and reflected in demand forecasting.

[0006] However, existing power demand forecasting methods that use indicators such as standard load density and diversity factor have unclear selection criteria, scope, and definitions of underlying data, and since results vary each time research on the criteria is conducted, this can lead to a decline in the reliability of power demand forecasting.

[0007] In particular, the standard load density by industry applied to industrial complexes is defined for only about 20 industries, making it difficult to accurately forecast demand for special complexes based on new industries such as knowledge industries and high-tech industries.

[0008] Furthermore, regional investments resulting from changes and new urban planning schemes, as well as infrastructure construction projects, increase population inflow and birth rates. Consequently, since economic revitalization leads to higher sales and can cause an increase in electricity demand, demand forecasting based on urban development factors and electricity usage is necessary.

[0009] In particular, when planning infrastructure facilities for a new development complex in a large-scale area where high power demand is expected, it is common for the construction period of the complex to take 5 to 10 years or more, and accordingly, power facilities and demand forecasting must be planned and responded to with a medium-to-long-term perspective.

[0010] However, actual existing demand forecasts are inevitably limited to estimations based solely on publicly available information regarding the completion date of a development complex, such as the applicable area, purpose of use, and duration relative to the total complex size. Consequently, it is currently difficult to establish sequential demand forecasts and detailed power facility plans for large-scale new development complexes based on their construction phases.

[0011] Figure 1 is a diagram showing the concept of the power demand forecasting process for a new development complex. Referring to Figure 1, power demand is forecasted by applying standard load density and diversity factor to the infrastructure based on public information (development scale, use, period) containing development plan information.

[0012] Based on this, it is possible to establish and proceed with plans for the construction of power facilities, but this approach is based solely on planning information from the initial stages of the development complex. Therefore, in the case of development complexes with a construction period of 5 to 10 years or more, such as actual large-scale complexes, it is impossible to sequentially establish and proceed with mid-to-long-term plans until the completion of the complex.

[0013] Therefore, in order to plan and respond to adaptive power demand in newly developed complexes, a power demand forecasting method different from the method based on publicly disclosed information according to construction plans during the initial stages of complex development is required. Prior art literature

[0014] 1. Republic of Korea Published Patent No. 10-2022-0011506 The problem to be solved

[0015] The present invention is proposed to resolve the problems associated with the background technology described above, and aims to provide a power demand forecasting system and method capable of selectively applying a standard load density based on object recognition feature information regarding infrastructure constructed within a complex, utilizing land cover information based on image recognition of satellite and aerial images of a newly developed complex.

[0016] In addition, another objective of the present invention is to provide a power demand forecasting system and method capable of efficient power facility construction planning and power demand response by deriving situational information-based power demand forecasts through image recognition results of the development complex periodically during the construction period. means of solving the problem

[0017] To achieve the above-mentioned objectives, the present invention provides a power demand forecasting system capable of selectively applying a standard load density based on object recognition feature information regarding infrastructure constructed within a complex, using image recognition-based land cover information for satellite and aerial images of a newly developed complex.

[0018] The above power demand prediction system is,

[0019] A video information providing server that generates video information; and

[0020] It is characterized by including a power forecasting server that generates land cover information based on image recognition for the above image information through a communication network, and generates power demand forecasting information by selectively applying a standard load density according to object recognition feature information for infrastructure constructed within a specific complex through the above land cover information.

[0021] In addition, the power prediction server comprises: a collection module for collecting image information; a classification module for extracting image information based on location information of a new development complex among the image information and generating land cover information from the extracted image information; an object recognition module for generating object recognition feature information through object recognition and generating power demand prediction information for each object based on the object recognition feature information; and a prediction module for generating overall power demand prediction information for the new development complex by summing the power demand prediction information for each object.

[0022] In addition, the object recognition feature information is characterized by including the area and height of the object.

[0023] In addition, the above land cover information is characterized by being generated by classifying the sections in a vacant land state during the initial development of the new development complex as vacant land areas.

[0024] In addition, the above image information is periodically collected and compared, and power demand prediction information for each object is generated based on changes in the land cover status of the object in the land cover information according to the comparison result.

[0025] In addition, the above change is characterized by being calculated by estimating the added area or height of the object through comparison with past image information from the initial stage of the construction of the new development complex.

[0026] In addition, the power demand forecast information for each object is characterized by being calculated by applying a predetermined standard load density based on public information to the object recognition feature information.

[0027] In addition, the power demand forecast information for each object is characterized by being a value obtained by multiplying the area of ​​the object by the standard load density and dividing it by the inequality factor, or a value obtained by multiplying the height of the object by the standard load density and dividing it by the inequality factor.

[0028] At this time, the diversity factor is characterized by representing the relationship between the power consumption of two or more loads as a predetermined value.

[0029] In addition, the above image information is characterized as being satellite image information or aerial image information.

[0030] On the other hand, another embodiment of the present invention provides a method for predicting power demand through image recognition-based land cover information, characterized by comprising: (a) a step in which an image information providing server generates image information; and (b) a step in which a power prediction server generates image recognition-based land cover information for the image information through a communication network, and selectively applies a standard load density according to object recognition feature information for infrastructure constructed within a specific complex through the land cover information to generate power demand prediction information.

[0031] At this time, the above step (b) is characterized by comprising: a step in which a collection module collects the image information; a step in which a classification module extracts image information corresponding to the location information of the new development complex among the image information and generates land cover information from the extracted image information; a step in which an object recognition module generates object recognition feature information through object recognition within the vacant land area and generates power demand forecast information for each object according to the object recognition feature information; and a step in which a forecast module sums the power demand forecast information for each object to generate overall power demand forecast information for the new development complex. Effects of the invention

[0032] According to the present invention, it is possible to establish efficient plans for facility investment and duration by referring to periodic power demand forecasts according to the development process of a new development complex and responding adaptively to the construction of power facilities required for the complex.

[0033] In addition, another advantage of the present invention is that it enables efficient and rapid power demand forecasting even for large-scale complexes through satellite imagery of newly developed complexes, and when utilizing an image recognition-based power demand forecasting system, it is possible to estimate periodic and collective power demand forecasts for development complexes being established nationwide. Brief explanation of the drawing

[0034] Figure 1 is an example of the existing power demand forecasting process for a typical new development complex. FIG. 2 is a block diagram of a power demand prediction system based on image recognition-based land cover information according to an embodiment of the present invention. Figure 3 is a detailed configuration block diagram of the power prediction server shown in Figure 2. FIG. 4 is a flowchart showing the process of predicting power demand through image recognition-based land cover information according to an embodiment of the present invention. FIG. 5 is a flowchart showing the process of predicting power demand for an object according to an embodiment of the present invention. FIG. 6 is a flowchart showing a process of periodic power demand prediction through image information according to an embodiment of the present invention. FIG. 7 is a conceptual diagram of extracting features of an object based on image recognition according to an embodiment of the present invention. Specific details for implementing the invention

[0035] The present invention is capable of various modifications and may have various embodiments, and specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the invention to specific embodiments, and it should be understood that the invention includes all modifications, equivalents, and substitutions that fall within the spirit and scope of the invention.

[0036] When describing each drawing, similar reference numerals are used for similar components.

[0037] Terms such as first, second, etc., may be used to describe various components, but said components should not be limited by said terms. Such terms are used solely for the purpose of distinguishing one component from another.

[0038] For example, without departing from the scope of the present invention, the first component may be named the second component, and similarly, the second component may be named the first component. The term "and / or" includes a combination of a plurality of related described items or any of a plurality of related described items.

[0039] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which this invention pertains.

[0040] Terms such as those defined in commonly used dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in this application.

[0041] A power demand prediction system and method based on image recognition-based land cover information according to an embodiment of the present invention will be described in detail below with reference to the attached drawings.

[0042] FIG. 2 is a block diagram of a power demand prediction system (200) based on image recognition-based land cover information according to an embodiment of the present invention. Referring to FIG. 2, the power demand prediction system (200) may be configured to include an image information providing server (210) that generates satellite image information and aerial image information, a communication network (220), a power prediction server (230) that generates image recognition-based land cover information for satellite image information and aerial image information through the communication network (220), and generates power demand prediction information by selectively applying a standard load density according to object recognition feature information for infrastructure being constructed within the complex through the land cover information, and a terminal (240) that receives and outputs power demand prediction information from the power prediction server (230) through the communication network (220).

[0043] The image information providing server (210) performs the function of generating satellite image information and / or aerial image information. It provides a service API (Application Programming Interface) capable of acquiring satellite and / or aerial images. Accordingly, the power prediction server (230) can collect satellite images of the development complex based on the location information of the new development complex in conjunction with this service API. Of course, the image server (210) may provide both location information and map information, or it may provide only image information. In addition, the image information providing server (210) may be linked with a server operated by the National Geographic Information Institute.

[0044] The communication network (220) can connect the phase image server (210), the power prediction server (230), and the terminal (240) via communication. To this end, the communication network (220) refers to a connection structure capable of exchanging information between each node, such as a plurality of terminals and servers, and can be a Public Switched Telephone Network (PSTN), a Public Switched Data Network (PSDN), an Integrated Services Digital Network (ISDN), a Broadband Integrated Services Digital Network (BISDN), a Local Area Network (LAN), a Metropolitan Area Network (MAN), a Wide Area Network (WLAN), etc.

[0045] However, the present invention is not limited thereto and may be wireless communication networks such as CDMA (Code Division Multiple Access), WCDMA (Wideband Code Division Multiple Access), Wibro (Wireless Broadband), WiFi (Wireless Fidelity), DLNA (Digital Living Network Alliance), Zigbee, Z-wave, HSDPA (High Speed ​​Downlink Packet Access) networks, Bluetooth, RFID (Radio Frequency Identification), Infrared Data Association (IrDA), Ultra-wide Band, Wireless USB (Wireless Universal Serial Bus), NFC (Near Field Communication) networks, satellite broadcasting networks, analog broadcasting networks, DMB (Digital Multimedia Broadcasting) networks, etc. Alternatively, it may be a combination of these wired communication networks and wireless communication networks.

[0046] The power prediction server (230) generates land cover information based on image recognition of satellite image information and aerial image information through the communication network (220), and performs the function of generating power demand prediction information by selectively applying a standard load density according to object recognition feature information of infrastructure being constructed within the complex through the land cover information.

[0047] The terminal (240) performs the function of connecting to the power prediction server (230) to receive and output power demand prediction information. Accordingly, the terminal (240) can be a mobile phone, a smartphone, a laptop computer, a digital broadcasting terminal, a PDA (Personal Digital Assistants), a PMP (Portable Multimedia Player), a navigation device, a notepad, etc. Accordingly, the terminal (240) can be configured to include a communication circuit, a display, a microprocessor, memory, etc.

[0048] FIG. 3 is a detailed block diagram of the power prediction server (230) illustrated in FIG. 2. Referring to FIG. 3, the power prediction server (230) may be configured to include: a collection module (310) that collects image information including satellite image information and aerial image information; a classification module (320) that extracts image information according to location information of a new development complex among the image information and classifies a vacant land area, which is a section in the initial vacant land state of the new development complex, to generate land cover information; an object recognition module (330) that generates object recognition feature information through object recognition within the vacant land area and applies a predetermined standard load density according to the land cover information to the object recognition feature information to generate power demand prediction information for each object; a prediction module (340) that generates overall power demand prediction information for the new development complex based on the power demand prediction information for each object; and a display unit (350) that outputs overall power demand prediction information.

[0049] The collection module (310) has the function of collecting image information including satellite image information and aerial image information. To this end, the collection module (310) may be configured to include a communication modem, a microprocessor, a microcomputer, memory, etc., for communication connection with a communication network (220).

[0050] The memory may be memory provided within a microprocessor or microcontroller, or it may be separate memory. Accordingly, it may be composed of a combination of non-volatile memory such as a flash memory disk (SSD: Solid State Disk), hard disk drive, flash memory, EEPROM (Electrically erasable programmable read-only memory), SRAM (Static RAM), FRAM (Ferro-electric RAM), PRAM (Phase-change RAM), MRAM (Magnetic RAM), and / or volatile memory such as DRAM (Dynamic Random Access Memory), SDRAM (Synchronous Dynamic Random Access Memory), and DDR-SDRAM (Double Data Rate-SDRAM).

[0051] The classification module (320) performs the function of generating land cover information by extracting image information based on location information of a new development complex among image information and classifying the vacant land area, which is a section in the initial vacant land state of the new development complex. A map representing the state of the surface can be created using images taken by a satellite or an aircraft, which is called a land cover map. It is divided into a major classification that distinguishes urbanized dry areas, agricultural areas, etc. according to classification items, a medium classification that classifies urbanized dry areas into residential areas, industrial areas, commercial areas, etc. according to the major classification, and a sub-classification that subdivides residential areas into single-family residential areas, public residential areas, etc. according to the medium classification. The sub-classification may consist of a detailed item, a classification code, and a color code.

[0052] Land cover classification is one of the most representative and typical applications of remote sensing data, classifying physical conditions of the Earth's surface, such as forests, grasslands, and concrete pavements. Since such land cover maps best reflect the current state of the Earth's surface, they are utilized as foundational data for various realistic modeling projects, contributing significantly to enhancing the efficiency and scientific rigor of policy formulation.

[0053] The object recognition module (330) generates object recognition feature information through object recognition within the vacant land area and performs the function of generating power demand forecast information for each object by applying a predetermined standard load density based on land cover information to the object recognition feature information. The object recognition feature information can be area and height.

[0054] Once object recognition is performed, the area and height of objects within the region (e.g., buildings) can be estimated through a pre-trained machine-learned object recognition model, or the area and height of newly added objects can be estimated by comparing them with historical image information from the initial stages of construction. The object recognition model is an object detection algorithm that detects predetermined key objects and distinguishes them by marking a bounding box centered on them.

[0055] The area can be calculated using the number of pixels within the bounding box (for example, assuming 5 cm per pixel). For instance, it can be calculated by multiplying the number of pixels on the x-axis by the number of pixels on the y-axis. Of course, it is also possible to calculate it by a ratio.

[0056] In addition, machine learning is performed using neural networks, which have a structure with multiple hidden layers between the input and output layers. This is a method in which the machine learns processing rules through data on its own, without humans providing procedures or rules for data processing. Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN) can be used.

[0057] In addition, the height characteristics of objects (e.g., buildings) can also be estimated through 3D satellite imagery, and by reflecting the power demand characteristics according to the objects, it is possible to derive a power demand forecast based on the entire development complex.

[0058] The prediction module (340) performs the function of generating overall power demand prediction information for the new development complex based on object-specific power demand prediction information. Specifically, it generates comprehensive overall power demand prediction information for the new development complex by summing the object-specific power demand prediction information generated for each object by the object recognition module (330).

[0059] The display unit (350) performs the function of outputting overall power demand forecast information. It also outputs information processing screens, setting screens, information guidance screens, etc. To this end, the display unit (350) may be configured to include a display and a sound system. The display may be an LCD (Liquid Crystal Display), an LED (Light Emitting Diode) display, a PDP (Plasma Display Panel), an OLED (Organic LED) display, a touch screen, a CRT (Cathode Ray Tube), a flexible display, a micro LED, a mini LED, etc. In the case of a touch screen, it may also be used as an input means.

[0060] The classification module (320), object recognition module (330), prediction module (340), etc., illustrated in FIG. 3 represent a unit that processes at least one function or operation, and this can be implemented in software and / or hardware. In hardware implementation, it can be implemented as an application-specific integrated circuit (ASIC), digital signal processing (DSP), programmable logic device (PLD), field programmable gate array (FPGA), processor, microprocessor, other electronic unit, or a combination thereof designed to perform the above-mentioned function.

[0061] In software implementation, software constituent components (elements), object-oriented software constituent components, class constituent components and task constituent components, processes, functions, attributes, procedures, subroutines, segments of program code, drivers, firmware, microcode, data, databases, data structures, tables, arrays, and variables may be included. Software, data, etc. may be stored in memory and executed by a processor. Memory or processors may employ various means well known to those skilled in the art.

[0062] FIG. 4 is a flowchart illustrating a process of predicting power demand through image recognition-based land cover information according to an embodiment of the present invention. Referring to FIG. 4, image information of a development complex is collected in conjunction with a service API provided to an image information providing server (210) based on location information of the development complex (steps S410, S420). In addition, image information of the development complex based on location information of the development complex is extracted from the collected image information.

[0063] Subsequently, image processing is performed on this image information (step S430). Specifically, using image processing techniques, the sections in the initial vacant land state of the area are classified, and objects within the new development complex are recognized (steps S431, S432). Specifically, regarding the image information (satellite imagery or aerial imagery) of the new development complex provided by the service API, the development area and sections are distinguished using image processing techniques, and the building area within the classified area of ​​the new development complex is estimated through a pre-trained object recognition model for object recognition, or the additional building area is estimated by comparing with past images from the initial development period.

[0064] In addition, the characteristics of the height of the building (object) can also be estimated through 3D satellite imagery, and by reflecting the power demand characteristics according to the object, it is possible to derive a power demand forecast based on the entire development complex.

[0065] In addition, the land cover status of the object is analyzed from the land cover information through comparison and image recognition of periodically collected satellite images, and the predicted power demand for each object (e.g., building) is estimated based on changes in the object (i.e., changes in height, changes in area) (step S440).

[0066] Buildings expected to have power demand are generated by calculating individual demand forecasts by reflecting power demand characteristics based on each object result through image-based object recognition, and finally, by estimating the total power demand forecast for the new development complex through the summation of power demand forecasts for each object.

[0067] FIG. 5 is a flowchart illustrating a power demand prediction process for an object according to an embodiment of the present invention. Referring to FIG. 5, image information of a development complex is collected in conjunction with a service API provided to an image information providing server (210) based on location information of the development complex, and image processing is performed on this image information to classify the region of the development complex (step S510).

[0068] Subsequently, using image processing techniques, the sections in the initial vacant land state of the area are classified, and objects within the newly developed complex are recognized (step S511).

[0069] Afterwards, object recognition feature information is calculated for the recognized object (step S513).

[0070] Meanwhile, public information regarding the new development complex is obtained to verify the use for residential, non-residential, industrial complexes, etc. (steps S520, S521).

[0071] Here, while land use classification based on publicly disclosed information generally allows for the identification of a single use for the entire development complex, land cover information subdivided into unit sections enables the classification of land use according to various sections within a large complex; therefore, when applying various land uses, the results of power demand forecasting based on land classification can be utilized more specifically.

[0072] Once it is confirmed whether the newly developed complex corresponds to residential, non-residential, or industrial land, the standard load density by use is calculated, and this calculated standard load density by use is applied to object recognition feature information to generate power demand forecast information by object (steps S523, S530). In addition, by applying the standard load density by use to the area and height information extracted through object recognition, the power demand forecast value by object can be estimated by modifying and transforming the existing mathematical formula for power demand forecasting. The existing mathematical formula for power demand forecasting is as follows.

[0073]

[0074] Here, the standard load density for each application is a predetermined value. Diversity factor is a predetermined value that represents the relationship between power consumption of two or more loads and indicates the degree of simultaneous use of power-consuming devices.

[0075] To elaborate, the object's power demand (W) is the product of the object's area and its standard load density divided by the diversity factor. Alternatively, it is the product of the object's height and its standard load density divided by the diversity factor.

[0076] Finally, the power demand forecast for the new development complex is derived by summing the power demand forecasts for each object (step S540).

[0077] FIG. 6 is a flowchart illustrating a process of periodic power demand forecasting using image information according to an embodiment of the present invention. Referring to FIG. 6, a development plan for a new development complex is established, and public information regarding the new development complex is obtained (steps S610, S620).

[0078] Subsequently, a power demand forecast is calculated based on public information (step S630). The power demand forecast is calculated by acquiring image information, using this image information to calculate the power demand forecast, and constructing power facilities accordingly (steps S631, S632, S633).

[0079] Of course, at each planning stage, satellite imagery of the development complex can be periodically collected from an image service API to estimate power demand forecasts based on image recognition-based land cover information.

[0080] Therefore, unlike the method of estimating power demand forecasts at the time of completion based solely on existing publicly available information, satellite or aerial imagery (i.e., image information measured using aircraft) can be acquired at predetermined intervals or key points to frequently review power demand forecasts for the relevant region, and based on this, power facility construction plans and responses in accordance with demand forecasts can be adaptively executed.

[0081] FIG. 7 is a conceptual diagram of extracting object features based on image recognition according to an embodiment of the present invention. Referring to FIG. 7, it is an example showing object recognition results according to the progress of construction within a new development complex (steps S710, S720, S730) through a pre-trained object recognition model.

[0082] Even when the foundation work is completed, the power demand forecast for each object can be calculated by referring to the placement location of each object in the zone (701), and the power demand forecast for the new development complex can be calculated by summing the power demand forecasts for each object (steps S710, S711, S713).

[0083] It is possible to extract features of objects (702) for an area (701) where development is in progress (i.e., object recognition feature information) through an object recognition model that has been pre-trained with an image DB (Database) for residential and industrial areas, etc., which is used for extracting land cover information based on image recognition (step S720). Of course, in this case, the power demand forecast for the new development complex can also be calculated by summing the power demand forecasts for each object (steps S721, S723).

[0084] When extracting the area by zone, the integrated area (703) can be estimated by combining only the outer boundary lines using the extracted area and location information for each object (step S730). Of course, in this case, the power demand forecast for the new development complex can also be calculated by summing the power demand forecasts for each object (steps S731, S733).

[0085] When power demand forecasts are calculated even while development is in progress, a plan for constructing power facilities is established based on this, and a response is made according to this plan (Step S740).

[0086] Additionally, the steps of the method or algorithm described in connection with the embodiments disclosed herein may be implemented in the form of program instructions that can be executed through various computer means, such as a microprocessor, a processor, a CPU (Central Processing Unit), etc., and recorded on a computer-readable medium. The computer-readable medium may include program (instruction) code, data files, data structures, etc., either alone or in combination.

[0087] The program (instruction) code recorded on the above medium may be those specifically designed and configured for the present invention, or those known and available to those skilled in the art of computer software. Examples of computer-readable recording media may include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs, DVDs, and Blu-rays; and semiconductor memory devices specifically configured to store and execute program (instruction) code, such as ROM (Read Only Memory), RAM (Random Access Memory), and flash memory.

[0088] Here, examples of program (instruction) code include not only machine code, such as that generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter, etc. The aforementioned hardware device may be configured to operate as one or more software modules to perform the operation of the present invention, and vice versa. Explanation of the symbols

[0089] 200: Power Demand Forecasting System 210: Video Information Provision Server 220: Communication network 230: Power Prediction Server 240: Terminal 310: Collection Module 320: Classification Module 330: Object Recognition Module 340: Prediction Module 350: Display unit 701: Zone 702: Object 703: Integrated Area

Claims

Claim 1 The system includes: a video information providing server (210) that generates video information; and a power forecasting server (230) that generates video recognition-based land cover information for the video information through a communication network (220), and generates power demand forecasting information by selectively applying a standard load density according to object recognition feature information for infrastructure being constructed within a specific complex through the land cover information. The video information is satellite image information or aerial image information, the video information is collected periodically, and changes in the object recognition feature information are analyzed from the land cover information. The power forecasting server (230) includes: a collection module (310) that collects the video information; a classification module (320) that extracts the corresponding video information according to the location information of the new development complex among the video information and generates land cover information from the extracted corresponding video information; an object recognition module (330) that generates object recognition feature information through the object recognition and generates power demand forecasting information for each object according to the object recognition feature information; and a forecasting module (340) that generates overall power demand forecasting information for the new development complex by summing the power demand forecasting information for each object.It includes, wherein the object recognition feature information includes the area and height of the object, and the object is a building where power demand is expected, and the land cover information is generated by classifying the section in the initial state of vacant land of the new development complex as a vacant land area, and the image information is periodically collected and compared, and according to the comparison result, power demand prediction information for each object is generated based on the change according to the land cover state of the object in the land cover information, and the change is calculated by estimating the added area or height of the object through comparison with past image information from the initial state of the new development complex, and the power demand prediction information for each object is calculated by applying a predetermined standard load density according to public information to the object recognition feature information, and the power demand prediction information for each object is a value obtained by multiplying the area of ​​the object by the standard load density and dividing it by the diversity factor, or a value obtained by multiplying the height of the object by the standard load density and dividing it by the diversity factor, and the diversity factor is a predetermined value representing the relationship of power consumption between two or more loads, and the power prediction server (230) outputs the total power demand prediction information. A power demand prediction system based on image recognition and land cover information having a terminal that outputs prediction information, further comprising a display unit (350); and a terminal (240) that receives and outputs power demand prediction information from the power prediction server (230) through the communication network (220).

Images