Thermal environment estimation device, thermal environment estimation method, and thermal environment estimation program

Abstract

To improve the technology for estimating the thermal environment of living spaces. [Solution] The thermal environment estimation device comprises a thermal environment simulator 124 that simulates the thermal environment in a three-dimensional model representing the shape of the living space using computational fluid dynamics, and a gap arrangement unit 122 that arranges gaps at the boundary of the living space through which air can circulate in the thermal environment simulation by the thermal environment simulator 124.

Description

Technical Field

[0001] The present disclosure relates to a thermal environment estimation device, a thermal environment estimation method, and a thermal environment estimation program for estimating the thermal environment of a living space.

Background Art

[0002] Currently, under the leadership of the government, the spread of reforms to improve living comfort is underway. Specifically, technologies for improving the thermal environment of houses while suppressing energy consumption, and efforts for the spread of net-zero energy houses (ZEH) are being made. A ZEH is a house that aims to achieve "substantially improving the heat insulation performance of the outer skin, etc., introducing a highly efficient equipment system to maintain the quality of the indoor environment while achieving significant energy savings, and then introducing renewable energy to make the annual balance of primary energy consumption zero." In order to achieve the government goal of "aiming for the realization of ZEH on average for newly constructed houses by 2030, and for more than half of the newly constructed detached houses ordered by house manufacturers by 2020," house manufacturers and the like are promoting the development of various technologies for realizing ZEH.

[0003] Since the heat insulation performance of building materials used in houses and the like has been significantly improved compared to before, significant energy savings have been achieved in newly constructed detached houses.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the case of used homes, the building envelope performance, such as insulation and airtightness, may not be adequate. Even in the case of used detached houses that do not use high-performance insulation materials, it is possible to significantly improve the insulation performance of the house by carrying out renovations to install high-performance insulation materials. However, the average homeowner does not have the knowledge about building materials or renovations, so they do not recognize the usefulness of renovations and do not feel that renovations are necessary.

[0006] As a technology that enables customers to deepen their knowledge about housing, a housing information provision system that simulates and presents temperature environments and other factors to the user has been proposed (see, for example, Patent Document 1). The present inventors believed that it would be possible to further improve the technology for estimating the thermal environment of living spaces, and thus came up with the technology disclosed here.

[0007] This disclosure is made in view of these challenges, and its purpose is to improve the technology for estimating the thermal environment of living spaces. [Means for solving the problem]

[0008] To solve the above problems, a thermal environment estimation device according to one aspect of the present disclosure comprises a simulator that simulates the thermal environment in a three-dimensional model representing the shape of a living space using computational fluid dynamics, and a gap arrangement unit that arranges gaps at the boundary of the living space through which air can circulate in the simulation of the thermal environment by the simulator.

[0009] Another aspect of this disclosure is a method for estimating a thermal environment. This method causes a computer to perform the steps of: simulating the thermal environment in a three-dimensional model representing the shape of a living space using computational fluid dynamics; and, in the simulation of the thermal environment in the simulation step, arranging gaps at the boundary of the living space through which air can circulate.

[0010] Another aspect of this disclosure is a thermal environment estimation device. This device comprises a simulator that simulates the thermal environment in a three-dimensional model representing the shape of a living space using computational fluid dynamics, and a learning unit that learns a thermal environment estimation model that takes the shape of a living space as input and outputs the thermal environment in a living space, using the results of simulations of the thermal environment in a plurality of three-dimensional models having different shapes by the simulator as learning data. The learning data includes the three-dimensional model and the simulation results of the thermal environment in the three-dimensional model, rotated or inverted while maintaining the vertical direction.

[0011] Another aspect of this disclosure is also a thermal environment estimation device. This device includes a thermal environment estimation unit that estimates the thermal environment in a living space by inputting data representing the shape of the living space into a trained thermal environment estimation model that takes data representing the shape of the living space as input and outputs data representing the thermal environment in the living space.

[0012] Furthermore, any combination of the above components, as well as conversions of the expressions of this disclosure between methods, apparatus, systems, recording media, computer programs, etc., are also valid as aspects of this disclosure. [Brief explanation of the drawing]

[0013] [Figure 1] This is a diagram showing the configuration of the thermal environment estimation system according to the embodiment. [Figure 2] This is a flowchart showing the procedure for the thermal environment estimation method according to the embodiment. [Figure 3] This is a flowchart showing the procedure for the thermal environment estimation method according to the embodiment. [Figure 4] This diagram shows the configuration of the learning device according to the embodiment. [Figure 5] This figure shows the configuration of the thermal environment estimation device according to the embodiment. [Figure 6] This is a diagram showing an example of a living space. [Figure 7] This figure shows an example of a three-dimensional model representing the shape of the living space shown in Figure 6. [Figure 8] This is a diagram showing the result of measuring the temperature distribution of the living space shown in FIG. 6. [Figure 9] This is a diagram showing the result of estimating the temperature distribution in the three-dimensional model shown in FIG. 7 using the learned thermal environment estimation model learned by the thermal environment estimation method of the embodiment.

Embodiments for Carrying Out the Invention

[0014] In an embodiment of the present disclosure, a technique for estimating the thermal environment of a living space such as a house will be described.

[0015] FIG. 1 shows the configuration of a thermal environment estimation system according to an embodiment. The thermal environment estimation system 1 includes a learning device 100, a learning data generation device 150, a thermal environment estimation device 200, a terminal device 2, and the Internet 3 which is an example of a communication network connecting these devices.

[0016] The learning device 100 learns a thermal environment estimation model for estimating the thermal environment of a living space. The learning device 100 learns a thermal environment estimation model which is a surrogate model that substitutes for a numerical fluid dynamics simulation, using the result of simulating the thermal environment in the living space by numerical fluid dynamics as learning data. The thermal environment estimation model inputs data representing the shape of the living space and outputs data representing the thermal environment in the living space.

[0017] The learning data generation device 150 generates learning data for the learning device 100 to learn the thermal environment estimation model. The learning data generation device 150 may be implemented by the same device as the learning device 100 or by a different device. In the following description, the learning data generation device 150 is implemented by the same device as the learning device 100.

[0018] The thermal environment estimation device 200 estimates the thermal environment of the living space using the thermal environment estimation model learned by the learning device 100, and presents the estimated thermal environment to the terminal device 2. The thermal environment of the living space may include the temperature distribution, average temperature, temperature at a specific position, etc. of the living space. The thermal environment estimation device 200 may present by superimposing an image representing the thermal environment of the living space on the image of the living space.

[0019] According to the thermal environment estimation system 1 of the present embodiment, the resident can grasp the thermal environment of his / her living space, so that reforms for improving the thermal environment can be considered. In addition, it is possible to promote reforms for improving the thermal environment of the living space and popularize ZEH.

[0020] The thermal environment estimation device 200 may estimate and present the thermal environment of the living space before and after implementing a reform that changes at least a part of the living space, such as replacing the building materials used in the living space with building materials having higher performance such as heat insulation. Thereby, since the effect of the reform can be presented clearly, the necessity and importance of the reform can be more strongly impressed on the user, and the implementation of the reform can be promoted.

[0021] The learning device 100 and the thermal environment estimation device 200 may be implemented by the same device or by different devices. The thermal environment of the living space estimated by the thermal environment estimation device 200 may be presented to the terminal device 2, or may be presented to a display device or the like provided in the thermal environment estimation device 200.

[0022] FIG. 2 is a flowchart showing the procedure of the thermal environment estimation method according to the embodiment. In this figure, a procedure in which the learning device 100 learns a thermal environment estimation model using learning data generated by simulating the thermal environment in various living spaces using numerical fluid dynamics will be described.

[0023] The learning device 100 generates a three-dimensional model representing the shape of the living space (S10). The living space may be one that actually exists or one that is virtually created. The three-dimensional model may be calculated based on captured images or design drawings of an existing living space. The three-dimensional model may be created using a CAD application or the like.

[0024] The learning device 100 places gaps that allow air to circulate at the boundaries of components such as the ceiling, exterior walls, interior walls, floor, windows, and doors of a three-dimensional model representing the shape of the living space (S12). The term "boundary" includes the boundary line and its vicinity. For example, the boundary between the ceiling and the exterior wall means the boundary line where the ceiling and the exterior wall meet and the space in its vicinity. The vicinity space may be, for example, a space of a few millimeters from the boundary line. The gaps may be the same size at all boundaries, or they may be of different sizes depending on the location of the boundary. For example, gaps of different sizes may be placed at the ends of exterior walls that are in contact with the outside air, the ends of interior walls that are not in contact with the outside air, and at the boundaries between windows and exterior walls. The gaps may be placed in a size corresponding to the equivalent gap area (c value), which is an indicator of the level of airtightness. The gaps may be placed at the ends of components such as the ceiling, exterior walls, interior walls, floor, windows, and doors. The gaps may be placed at the boundaries between the surface that makes up the exterior wall in contact with the outside air and other components. This allows for accurate reflection of airflow at the boundary in thermal environment simulations using computational fluid dynamics, as described later, enabling a more accurate simulation of the thermal environment of living spaces.

[0025] The learning device 100 divides the three-dimensional model into meshes to generate mesh data representing the shape of the living space (S14). The mesh size may be adjusted according to the accuracy required in the computational fluid dynamics simulation, the time required for the simulation, etc. The mesh data constituting the three-dimensional model may include metadata representing the attributes of the components of the living space. The components may include at least one of the following: ceiling, floor, interior walls, exterior walls, windows, doors, and heating and cooling equipment. The heating and cooling equipment may include air conditioning equipment, heaters, fans, etc. These attributes are reflected in physical quantities such as thermal conductivity in the computational fluid dynamics simulation.

[0026] The learning device 100 simulates the thermal environment in a three-dimensional model representing the shape of the living space using computational fluid dynamics (S16). The learning device 100 sets calculation conditions for calculating the thermal environment of the living space. The calculation conditions may include, for example, parameters in the computational fluid dynamics simulation, environmental conditions (outside temperature, incident solar radiation, indoor humidity, wind speed, etc.), conditions related to the structure of the living space (wall thickness, location, orientation, etc.), and conditions related to building materials and equipment installed in the living space (presence, type, number, and location of insulation, presence or absence of ventilation by ventilation equipment, ventilation speed, temperature of air supplied by heating and cooling equipment, wind speed, etc.). The learning device 100 may also estimate the structure of the living space and the building materials and equipment installed in the living space from information about the living space and set the calculation conditions. For example, the learning device 100 may estimate and set the specifications of the installed insulation material from the age and structure of the living space. The learning device 100 may also estimate and set the incident solar radiation from the orientation of the room and the specifications of the window glass. The learning device 100 simulates the thermal environment of the living space according to the set calculation conditions. The learning device 100 may also calculate the temperature distribution of the living space, the temperature, humidity, wind speed, oxygen concentration, carbon dioxide concentration at a predetermined location in the living space, etc.

[0027] The learning device 100 performs steps S10 to S16 for a number of living spaces with different shapes to accumulate simulation results of the thermal environment, and generates learning data from the accumulated pairs of three-dimensional models and thermal environment simulation results (S18). In order to efficiently generate a large amount of learning data, the learning device 100 may generate multiple learning data from one set of learning data by rotating or inverting the three-dimensional model and the thermal environment simulation results in the three-dimensional model while maintaining the vertical direction. This allows the thermal environment estimation model to be trained with a large amount of learning data, thereby improving the accuracy of the thermal environment estimation model. When simulating airflow in a three-dimensional model, the vertical direction is affected by air convection, so if the three-dimensional model and simulation results are converted in such a way that the vertical direction of the three-dimensional model changes when generating learning data, the simulation results may not be correct. By converting the three-dimensional model and simulation results while maintaining the vertical direction, a large amount of learning data can be efficiently generated while ensuring the accuracy of the simulation results.

[0028] The learning device 100 learns a thermal environment estimation model using the generated training data (S20). The learning device 100 may learn the thermal environment estimation model by supervised learning using the simulation results as ground truth data. The thermal environment estimation model may consist of a multi-layer neural network, and the learning device 100 may learn the thermal environment estimation model by deep learning. The learning device 100 may learn the thermal environment estimation model by any learning method such as backpropagation.

[0029] Figure 3 is a flowchart illustrating the procedure for the thermal environment estimation method according to an embodiment. This figure describes the procedure by which the thermal environment estimation device 200 estimates the thermal environment of a living space using a learned thermal environment estimation model learned by the learning device 100.

[0030] The thermal environment estimation device 200 acquires a three-dimensional model representing the shape of the living space for which the thermal environment is to be estimated (S30). The living space may be a real-world space or a virtually created space. The thermal environment estimation device 200 may acquire the three-dimensional model representing the shape of the living space from a terminal device 2 or the like, or it may generate it itself. The three-dimensional model may be calculated based on captured images or design drawings of a real-world living space. The three-dimensional model may be created using a CAD application or the like.

[0031] The thermal environment estimation device 200 generates mesh data representing the shape of the living space by meshing the three-dimensional model (S32). The mesh size may be larger than the mesh size when the learning device 100 performs a computational fluid dynamics simulation. When simulating the thermal environment using computational fluid dynamics, reducing the mesh size and representing the shape more finely improves the accuracy of the simulation. However, the inventors' experiments have shown that when estimating the thermal environment using a trained thermal environment estimation model, reducing the mesh size somewhat does not significantly change the estimation accuracy. Therefore, by reducing the mesh size compared to the training time, it is possible to shorten the time required while maintaining the accuracy of estimating the thermal environment of the living space. The mesh data constituting the three-dimensional model may include data representing the attributes of the components of the living space. The components may include at least one of the following: ceiling, floor, interior wall, exterior wall, window, door, and heating and cooling equipment.

[0032] The thermal environment estimation device 200 receives mesh data representing the shape of the living space as input to a trained thermal environment estimation model and outputs the thermal environment of the living space (S34).

[0033] The thermal environment estimation device 200 presents the thermal environment of the living space, output from the thermal environment estimation model, to the terminal device 2 (S36).

[0034] Figure 4 shows the configuration of a learning device 100 according to an embodiment. The learning device 100 comprises a communication device 101, a storage device 110, and a processing device 120. The learning device 100 may be a server device, a personal computer or other device, or a mobile device such as a mobile phone, smartphone, or tablet.

[0035] The communication device 101 communicates with other devices and sends and receives data.

[0036] The storage device 110 stores programs, data, etc., used by the processing device 120. The storage device 110 may be a semiconductor memory, a hard disk, etc. The storage device 110 may be installed inside the learning device 100 or it may be installed on the cloud. The storage device 110 holds the thermal environment estimation model 111. The storage device 110 holds information about the living space, data of a three-dimensional model representing the shape of the living space, mesh data representing the shape of the living space, information about the components of the living space, simulation results of the thermal environment of the living space, training data for training the thermal environment estimation model 111, etc.

[0037] The processing unit 120 comprises a living space three-dimensional model generation unit 121, a gap arrangement unit 122, a mesh data generation unit 123, a thermal environment simulator 124, a learning data generation unit 125, and a learning unit 126. These components are realized hardware-wise by circuits, the CPU of any computer, memory, and other LSIs, and software-wise by programs loaded into memory, but here we are describing functional blocks realized through the cooperation of these components. Therefore, it will be understood by those skilled in the art that these functional blocks can be realized in various forms, such as hardware only or a combination of hardware and software.

[0038] In step S10, the living space three-dimensional model generation unit 121 generates a three-dimensional model representing the shape of the living space.

[0039] In step S12, the gap placement section 122 places gaps that allow air to circulate at the boundaries of constituent elements such as the ceiling, exterior walls, interior walls, floor, windows, and doors of the living space in the three-dimensional model representing the shape of the living space.

[0040] In step S14, the mesh data generation unit 123 divides the three-dimensional model into meshes to generate mesh data representing the shape of the living space.

[0041] In step S16, the thermal environment simulator 124 simulates the thermal environment in a three-dimensional model representing the shape of the living space using computational fluid dynamics.

[0042] In step S18, the learning data generation unit 125 generates learning data from a set of a three-dimensional model and the simulation results of the thermal environment. The learning data generation unit 125 generates multiple learning data from one set of learning data by rotating or inverting the three-dimensional model and the simulation results of the thermal environment in the three-dimensional model while maintaining the vertical direction.

[0043] In step S20, the learning unit 126 trains the thermal environment estimation model 111 using the generated training data. The trained thermal environment estimation model 111 is then provided to the thermal environment estimation device 200.

[0044] Figure 5 shows the configuration of a thermal environment estimation device 200 according to an embodiment. The thermal environment estimation device 200 comprises a communication device 201, a storage device 210, and a processing device 220. The thermal environment estimation device 200 may be a server device, a device such as a personal computer, or a mobile device such as a mobile phone terminal, smartphone, or tablet terminal.

[0045] The communication device 201 communicates with other devices and sends and receives data.

[0046] The storage device 210 stores programs, data, etc., used by the processing device 220. The storage device 210 may be a semiconductor memory, a hard disk, etc. The storage device 210 may be installed inside the thermal environment estimation device 200 or it may be installed on the cloud. The storage device 210 holds the thermal environment estimation model 111. The storage device 210 holds information about the living space, data of a three-dimensional model representing the shape of the living space, mesh data representing the shape of the living space, information about the components of the living space, etc.

[0047] The processing unit 220 comprises a living space three-dimensional model acquisition unit 221, an input data generation unit 222, a thermal environment estimation unit 223, and a thermal environment presentation unit 224. These components are implemented hardware-wise by circuits, the CPU of any computer, memory, and other LSIs, and software-wise by programs loaded into memory, but here we are describing functional blocks that are realized through the cooperation of these components. Therefore, it will be understood by those skilled in the art that these functional blocks can be implemented in various forms, such as hardware only or a combination of hardware and software.

[0048] In step S30, the living space three-dimensional model acquisition unit 221 acquires a three-dimensional model representing the shape of the living space for which the thermal environment is to be estimated.

[0049] In step S32, the input data generation unit 222 generates input data to be input to the thermal environment estimation model 111. The input data generation unit 222 may also generate mesh data representing the shape of the living space by meshing the three-dimensional model acquired by the living space three-dimensional model acquisition unit 221. The mesh size may be larger than the mesh data size when the learning device 100 performs a computational fluid dynamics simulation.

[0050] In step S34, the thermal environment estimation unit 223 inputs the input data generated by the input data generation unit 222 into the trained thermal environment estimation model 111 to output the thermal environment of the living space.

[0051] In step S36, the thermal environment presentation unit 224 presents the thermal environment of the living space, estimated by the thermal environment estimation unit 223, to the terminal device 2 or the like.

[0052] Figure 6 shows an example of a living space. A living space has components such as a ceiling, floor, exterior walls, interior walls, windows, doors, and heating and cooling equipment.

[0053] Figure 7 shows an example of a three-dimensional model representing the shape of the living space shown in Figure 6. The shapes of the constituent elements present in the living space are modeled.

[0054] Figure 8 shows the results of measuring the temperature distribution of the living space shown in Figure 6. Figure 9 shows the results of estimating the temperature distribution in the three-dimensional model shown in Figure 7 using the trained thermal environment estimation model 111, which was trained using the thermal environment estimation method of this embodiment. In both figures, darker colors indicate lower temperatures. In the measurement results shown in Figure 8, the temperature is low not only in places exposed to the outside air, such as windows and doors, but also at the boundary between the exterior wall and the floor. In the estimation results shown in Figure 9, the temperature is also low at the boundary between the exterior wall and the floor, indicating that the actual temperature distribution can be estimated with good accuracy.

[0055] Thus, according to the thermal environment estimation system 1 of this embodiment, the thermal environment of a living space can be estimated quickly and accurately.

[0056] The present invention has been described above based on embodiments, but embodiments merely illustrate the principles and applications of the present invention. Furthermore, many modifications and changes in arrangement are possible in the embodiments, without departing from the spirit of the present invention as defined in the claims.

[0057] Although the above embodiment mainly described a residential building, the technology of this embodiment can also be applied to living spaces other than residential buildings. [Explanation of symbols]

[0058] 1 Thermal environment estimation system, 2 Terminal device, 100 Learning device, 111 Thermal environment estimation model, 121 Living space 3D model generation unit, 122 Gap arrangement unit, 123 Mesh data generation unit, 124 Thermal environment simulator, 125 Learning data generation unit, 126 Learning unit, 150 Learning data generation device, 200 Thermal environment estimation device, 221 Living space 3D model acquisition unit, 222 Input data generation unit, 223 Thermal environment estimation unit, 224 Thermal environment presentation unit.

Claims

1. A simulator that uses computational fluid dynamics to simulate the thermal environment in a three-dimensional model representing the shape of a living space, In the simulation of the thermal environment using the aforementioned simulator, a gap arrangement section is provided at the boundary of the living space, where gaps through which air can circulate are arranged. A thermal environment estimation device equipped with the following features.

2. The system includes a learning unit that learns a thermal environment estimation model that takes the shape of a living space as input and outputs the thermal environment in a living space, using the results of simulating the thermal environment in multiple three-dimensional models having different shapes using the simulator as training data. The thermal environment estimation device according to claim 1.

3. The aforementioned training data includes the three-dimensional model and the simulation results of the thermal environment in the three-dimensional model, rotated or inverted while maintaining the vertical direction. The thermal environment estimation device according to claim 2.

4. The mesh data constituting the three-dimensional model includes data representing the attributes of the components of the living space. A thermal environment estimation device according to any one of claims 1 to 3.

5. The aforementioned components include at least one of the following: ceiling, floor, interior wall, exterior wall, window, door, and heating and cooling equipment. The thermal environment estimation device according to claim 4.

6. On the computer, The steps involve simulating the thermal environment in a three-dimensional model representing the shape of a living space using computational fluid dynamics, In the simulation of the thermal environment in the aforementioned simulation step, the step of arranging gaps that allow air to circulate at the boundary of the living space, A method for estimating the thermal environment to perform this operation.

7. Computers, A simulator that uses computational fluid dynamics to simulate the thermal environment in a three-dimensional model representing the shape of a living space, In the simulation of the thermal environment using the aforementioned simulator, a gap arrangement section is provided at the boundary of the living space, where gaps through which air can circulate are arranged. A thermal environment estimation program designed to function as such.

8. A simulator that uses computational fluid dynamics to simulate the thermal environment in a three-dimensional model representing the shape of a living space, A learning unit learns a thermal environment estimation model that takes the shape of a living space as input and outputs the thermal environment in a living space, using the results of simulating the thermal environment in multiple three-dimensional models having different shapes using the simulator as training data. Equipped with, The aforementioned training data includes the three-dimensional model and the simulation results of the thermal environment in the three-dimensional model, rotated or inverted while maintaining the vertical direction. Thermal environment estimation device.

9. The system includes a thermal environment estimation unit that estimates the thermal environment in a living space by inputting data representing the shape of the living space into a trained thermal environment estimation model that takes data representing the shape of the living space as input and outputs data representing the thermal environment in the living space. Thermal environment estimation device.

10. The aforementioned thermal environment estimation model is trained using a three-dimensional model representing the shape of the living space and the results of simulating the thermal environment in the living space using computational fluid dynamics as training data. The thermal environment estimation device according to claim 9.

11. The mesh size of the mesh data representing the shape of the living space that the thermal environment estimation unit inputs to the thermal environment estimation model is larger than the mesh size of the mesh data representing the shape of the living space used in the computational fluid dynamics simulation when the thermal environment estimation model was trained. The thermal environment estimation device according to claim 10.

Images