WO2022249882A1 - Dispositif et procédé de traitement d'informations et programme - Google Patents

Dispositif et procédé de traitement d'informations et programme Download PDF

Info

Publication number
WO2022249882A1
WO2022249882A1 PCT/JP2022/019874 JP2022019874W WO2022249882A1 WO 2022249882 A1 WO2022249882 A1 WO 2022249882A1 JP 2022019874 W JP2022019874 W JP 2022019874W WO 2022249882 A1 WO2022249882 A1 WO 2022249882A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
distribution
target space
model
spatial shape
Prior art date
Application number
PCT/JP2022/019874
Other languages
English (en)
Japanese (ja)
Inventor
友佑 向江
新 高橋
祥宏 金子
Original Assignee
ピクシーダストテクノロジーズ株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by ピクシーダストテクノロジーズ株式会社 filed Critical ピクシーダストテクノロジーズ株式会社
Publication of WO2022249882A1 publication Critical patent/WO2022249882A1/fr

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/50—Control or safety arrangements characterised by user interfaces or communication
    • F24F11/52—Indication arrangements, e.g. displays
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
    • F24F11/63—Electronic processing
    • F24F11/64—Electronic processing using pre-stored data
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F11/00—Control or safety arrangements
    • F24F11/89—Arrangement or mounting of control or safety devices
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00—Measuring or testing not otherwise provided for
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01W—METEOROLOGY
    • G01W1/00—Meteorology
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F2110/00—Control inputs relating to air properties
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F2110/00—Control inputs relating to air properties
    • F24F2110/10—Temperature
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
    • F24F2130/00—Control inputs relating to environmental factors not covered by group F24F2110/00

Definitions

  • the present disclosure relates to an information processing device, an information processing method, and a program.
  • Air conditioning is required to achieve both user comfort and energy efficiency. In order to properly control air conditioning, it is useful to grasp the detailed indoor thermal environment in a timely manner.
  • Patent Document 1 describes a technique related to three-dimensional visualization of indoor temperature distribution by CFD (Computational Fluid Dynamics) analysis.
  • the purpose of the present disclosure is to shorten the time required to estimate the distribution of physical quantities over the target space.
  • One aspect of the present disclosure includes means for acquiring spatial shape information about the spatial shape of a target space; means for acquiring sensing information from sensors in the target space; An information processing apparatus comprising means for generating and means for estimating the distribution of physical quantities over a target space by applying model inputs to a trained model.
  • FIG. 1 is a schematic diagram showing the configuration of a sound wave transmitting device according to a first embodiment
  • FIG. 1 is a schematic diagram showing the configuration of a sound wave receiving device according to a first embodiment
  • FIG. 4 is an explanatory diagram of machine learning for constructing an inference model used in the first embodiment
  • 1 is an explanatory diagram of an outline of a first embodiment
  • FIG. It is a figure which shows the data structure of the spatial data table of 1st Embodiment.
  • FIG. 4 is an explanatory diagram of a filter according to the first embodiment;
  • FIG. 4 is a flowchart of physical quantity estimation processing according to the first embodiment;
  • FIG. 14 is a detailed flowchart of acquisition of propagation time information in FIG. 13;
  • FIG. 10 is an explanatory diagram of machine learning for constructing an inference model used in the second embodiment;
  • FIG. 11 is an explanatory diagram of the overview of the second embodiment;
  • FIG. 11 is an explanatory diagram of the outline of the third embodiment;
  • FIG. 19 is a diagram showing another example of FIG. 18;
  • 10 is a detailed flowchart of analysis processing according to the third embodiment; It is explanatory drawing of the machine learning for building the inference model used in 4th Embodiment. It is an explanatory view of an outline of a 4th embodiment. It is explanatory drawing of the machine learning for building the inference model used in 5th Embodiment. It is an explanatory view of an outline of a 5th embodiment.
  • FIG. 20 is an explanatory diagram of machine learning for constructing an inference model used in the sixth embodiment;
  • FIG. 21 is an explanatory diagram of the overview of the sixth embodiment;
  • FIG. 1 is a block diagram showing the configuration of the air conditioning system of the first embodiment.
  • FIG. 2 is a block diagram showing the detailed configuration of the air conditioning system of the first embodiment.
  • the air conditioning system 1 includes a measurement device 10 , a sound wave transmitter 20 , a sound wave receiver 30 , an air conditioner 40 and a thermometer 50 .
  • a measuring device 10 (an example of an “information processing device”) is connected to a sound wave transmitting device 20 , a sound wave receiving device 30 , an air conditioner 40 , and a thermometer 50 .
  • the measuring device 10, the sound wave transmitting device 20, the sound wave receiving device 30, the air conditioner 40, and the thermometer 50 are arranged in a space (hereinafter referred to as "target space") SP that is the target of physical quantity measurement.
  • target space space
  • the measuring device 10 may be arranged outside the target space SP.
  • the measuring device 10 has the following functions.
  • a function of controlling the sound wave transmitter 20 function of acquiring received waveform data from the sound wave receiver 30
  • Air conditioning A function of controlling the device 40
  • the measuring device 10 is, for example, a smart phone, a tablet terminal, or a personal computer.
  • the sound wave transmitting device 20 is configured to transmit a directional sound wave (for example, an ultrasonic beam) under the control of the measuring device 10 . Also, the sound wave transmitter 20 is configured to change the transmission direction of the ultrasonic waves.
  • a directional sound wave for example, an ultrasonic beam
  • the sound wave receiving device 30 is configured to receive the ultrasonic beams transmitted from the sound wave transmitting device 20 and generate received waveform data according to the received ultrasonic beams.
  • the sound wave receiving device 30 is, for example, an omnidirectional microphone or a directional microphone.
  • the air conditioner 40 is configured to adjust the internal environment (for example, temperature, humidity, airflow, air cleanliness, or a combination thereof) of the target space SP under the control of the measuring device 10 .
  • the thermometer 50 is configured to measure the temperature of the target space SP (hereinafter referred to as "reference temperature").
  • the thermometer 50 may be a contact thermometer or a non-contact thermometer (for example, an infrared radiation thermometer).
  • the measuring device 10 includes a storage device 11, a processor 12, an input/output interface 13, and a communication interface 14.
  • the storage device 11 is configured to store programs and data.
  • the storage device 11 is, for example, a combination of ROM (Read Only Memory), RAM (Random Access Memory), and storage (eg, flash memory or hard disk).
  • Programs include, for example, the following programs.
  • ⁇ OS (Operating System) program ⁇ Information processing (for example, information processing for measuring the physical quantity distribution of the target space SP, and information processing for providing feedback to the air conditioner 40 based on the physical quantity distribution of the target space SP) application program that runs
  • the data includes, for example, the following data.
  • ⁇ Data and databases referenced in information processing ⁇ Data obtained by executing information processing (that is, execution results of information processing)
  • ⁇ Data related to sound wave velocity characteristics regarding the speed of sound waves with respect to the temperature of the space ⁇ Spatial shape data (spatial shape information) indicating the shape of the target space
  • the processor 12 is configured to implement the functions of the measuring device 10 by activating programs stored in the storage device 11 and processing data.
  • Processor 12 is an example of a computer.
  • the programs and data stored in the storage device 11 may be provided via a network, or may be provided by being recorded on a computer-readable recording medium. At least part of the functions of the measurement device 10 may be implemented by one or more pieces of dedicated hardware.
  • the input/output interface 13 acquires signals (for example, received waveform data, reference temperature data, or user instructions) from input devices connected to the measuring apparatus 10 and outputs signals to output devices connected to the measuring apparatus 10. (eg, a control signal, or an image signal).
  • the input device is, for example, the sound wave receiver 30, the thermometer 50, a keyboard, a pointing device, a touch panel, or a combination thereof.
  • the output device is, for example, the sonic transmitter 20, the air conditioner 40, the display, or a combination thereof.
  • the communication interface 14 is configured to control communication with an external device (for example, a server not shown).
  • an external device for example, a server not shown.
  • FIG. 3 is a schematic diagram showing the configuration of the sound wave transmitter of the first embodiment.
  • the sound wave transmitting device 20 includes a plurality of ultrasonic transducers (an example of a "vibrating element") 21 and a control circuit 22.
  • the control circuit 22 causes the plurality of ultrasonic transducers 21 to vibrate under the control of the measuring device 10 .
  • the plurality of ultrasonic transducers 21 vibrate, ultrasonic beams are transmitted in the transmission direction (Z-axis direction) perpendicular to the transmission plane (XY plane).
  • FIG. 4 is a schematic diagram showing the configuration of the sound wave receiving device of the first embodiment.
  • the sound wave receiving device 30 includes an ultrasonic transducer 31 and a control circuit 32.
  • the ultrasonic transducer 31 vibrates when receiving the ultrasonic beam transmitted from the sound wave transmitting device 20 .
  • the control circuit 32 is configured to generate received waveform data corresponding to vibration of the ultrasonic transducer 31 .
  • FIG. 5 is a diagram showing the mesh structure of the target space.
  • FIG. 6 is an explanatory diagram of machine learning for constructing an inference model used in the first embodiment.
  • FIG. 7 is an explanatory diagram of the outline of the first embodiment.
  • sound wave transmitters 20a to 20b and sound wave receivers 30a to 30b are arranged in the target space SP.
  • the measuring device 10 can be connected to the sound wave transmitting device 20 and the sound wave receiving device 30 .
  • the measurement device 10 controls the sound wave transmitters 20a-20b to transmit sound waves (eg, ultrasonic beams).
  • the measuring device 10 acquires received waveform data relating to the waveforms of received sound waves from the sound wave receiving devices 30a and 30b.
  • the measuring device 10 specifies the propagation time of sound waves in each path in the target space SP based on the received waveform data.
  • the object space SP can be virtually divided into a plurality of meshes (an example of a "virtual partition") Mi (i is an argument).
  • Each mesh Mi is a two-dimensional or three-dimensional region.
  • the physical quantity of each mesh can be estimated as described below.
  • the path P200 is a path from the sound wave transmitter 20a to the sound wave receiver 30a through the mesh M1.
  • a path P201 is a path from the sound wave transmitting device 20b to the sound wave receiving device 30b through the mesh M1.
  • the propagation properties of sound waves (eg speed of sound) vary depending on the physical properties of the air in the path (eg temperature, wind direction and speed). That is, the propagation properties of sound waves on multiple paths through a mesh are all affected by the physical properties of the air within the mesh.
  • the physical quantity in the mesh M1 can be estimated based on the propagation distance (that is, path length) and propagation time of the sound wave on the path P200 and the propagation distance and propagation time of the sound wave on the path P201.
  • the measuring device 10 also changes the propagation paths of sound waves from the sound wave transmitters 20a and 20b to the sound wave transmitters 20a and 20b, and measures the propagation time of the sound waves on the changed paths. This makes it possible to estimate physical quantities in other meshes present on the route after the change.
  • the measuring device 10 uses an inference model (an example of a “learned model”) constructed by machine learning to obtain information (hereinafter referred to as “distribution information”) about the distribution of physical quantities over the target space.
  • distributed information information about the distribution of physical quantities over the target space.
  • MLP multilayer perceptron
  • CNN convolutional neural network
  • RNN recurrent neural network
  • LSTM long short memory
  • ResNet residual network
  • VAE generative adversarial network
  • Deep learning such as Transfer learning, DQN (Deep Q-Network), or CapsNet
  • DQN Deep Q-Network
  • CapsNet Deep learning such as Transfer learning, DQN (Deep Q-Network), or CapsNet
  • DQN Deep Q-Network
  • CapsNet Deep learning such as Transfer learning, DQN (Deep Q-Network), or CapsNet
  • the distribution information corresponds to binary information indicating whether the physical quantity at each point exceeds the reference value
  • logistic regression support vector machine (SVM), decision tree, naive Bayes, or k-nearest neighbor method ( KNN) can also be used.
  • SVM support vector machine
  • KNN k-nearest neighbor method
  • the inference model MD10 used by the measuring device 10 is constructed by machine learning (supervised learning) using learning data including learning input and correct data.
  • Machine learning may be performed by the measuring device 10 or may be performed by an external device (for example, a learning server).
  • the learning input is generated based on the spatial shape information SS10 and the propagation time information PT10.
  • Correct data is generated based on the distribution information DI10.
  • the spatial shape information SS10 is information about the shape of the space SP10 (not shown) for collecting learning data.
  • the space SP10 is not limited to one space, and may be multiple spaces.
  • the space SP10 may or may not include the target space.
  • the spatial shape information SS10 is obtained, for example, according to a user's instruction to an input device.
  • the spatial shape information SS10 may be content based on survey results, or data of a spatial model corresponding to the target space SP (for example, BIM (Building Information Modeling) data, or other CAD (Computer-Aided Design) data ) may be based on.
  • the spatial shape information SS10 may include information indicating the positions of the sound wave transmitting device 20 and the sound wave receiving device 30 installed in the space SP10.
  • the spatial shape information SS10 may include information indicating the propagation path of sound waves transmitted from the sound wave transmitting device 20 installed in the space SP10 to the sound wave receiving device 30.
  • the propagation time information PT10 is information relating to the propagation time of sound waves on paths within the space SP10.
  • the propagation time information PT10 is collected by installing the sound wave transmitting device 20 and the sound wave receiving device 30 in the space SP10 and performing actual measurements.
  • the distribution information DI10 is the distribution information of the space SP10.
  • the distribution information DI10 may be collected by actually measuring physical quantities at each point in the space SP10 (for example, by actually measuring with a thermocouple or wind speed sensor), or by estimating the distribution of physical quantities over the space SP10 by, for example, CFD simulation. Results may be used. Moreover, a result obtained by performing predetermined information processing (for example, filter processing such as a Kalman filter) on the estimation result by the CFD simulation may be used as the distribution information DI10. Information about the propagation time and propagation distance of sound waves measured using the sound wave transmitting device 20 and the sound wave receiving device 30 may be used for the filtering process.
  • predetermined information processing for example, filter processing such as a Kalman filter
  • the inference model MD10 can reasonably estimate the distribution of physical quantities over the given space from the spatial shape of the given space and the propagation time of sound waves along the path in the given space. can be estimated with high accuracy.
  • the measuring device 10 generates model inputs by referring to spatial shape information SS11 and propagation time information PT11.
  • the measurement device 10 provides model inputs to the inference model MD10.
  • the inference model MD10 makes inferences based on model inputs and returns distribution information DI11.
  • the spatial shape information SS11 is information about the shape of the target space SP.
  • the spatial shape information SS11 is obtained, for example, in response to a user's instruction to an input device.
  • the spatial shape information SS11 may be the content based on the survey result, or may be the content based on the data of the space model corresponding to the target space SP.
  • the spatial shape information SS11 may include information indicating the positions of the sound wave transmitting device 20 and the sound wave receiving device 30 installed in the target space SP.
  • the spatial shape information SS11 may include information indicating the propagation path of sound waves transmitted from the sound wave transmitting device 20 installed in the target space SP to the sound wave receiving device 30.
  • the propagation time information PT11 is information relating to the propagation time of the sound wave on the path within the target space SP.
  • the propagation time information PT11 is collected by actual measurement by the measurement device 10 using the sound wave transmitter 20 and the sound wave receiver 30 installed in the target space SP.
  • the propagation time information PT11 may be data at one point in time or may be time-series data.
  • the distribution information DI11 is the estimation result of the physical quantity distribution over the target space SP by the inference model MD10.
  • the measuring device 10 of the first embodiment uses the inference model MD10 constructed by machine learning to estimate physical quantities related to air characteristics in the target space. Therefore, according to this measuring device 10, the distribution of physical quantities over the target space SP can be estimated in a short time compared to a normal CFD simulation. Furthermore, according to this measuring device 10, even if the boundary conditions regarding the target space SP are unknown, the distribution of physical quantities over the target space can be estimated.
  • FIG. 8 is a diagram showing the data structure of the spatial data table of the first embodiment.
  • a spatial data table is an example of spatial shape information used by the measuring device 10 .
  • the spatial shape information is not limited to the example shown in FIG. 8 as long as it represents the shape (two-dimensional shape or three-dimensional shape) of the target space.
  • the spatial shape information is acquired by, for example, measuring the target space in advance with a sensor such as a LiDAR (Light Detection and Ranging) scanner, and stored in the storage device 11 .
  • the method of acquiring the spatial shape information is not limited to this. good.
  • the spatial data table in FIG. 8 stores spatial information about the target space.
  • the spatial data table includes a 'coordinate' field and a 'reflection property' field. Each field is associated with each other.
  • the "coordinates” field stores the coordinates of the reflecting members existing in the target space (hereinafter referred to as "reflecting member coordinates").
  • Reflecting member coordinates are represented by a coordinate system (hereinafter referred to as “space coordinate system”) with an arbitrary reference point in the target space as the origin. Coordinates represented by the spatial coordinate system are not limited to three-dimensional coordinates, and may be two-dimensional coordinates.
  • the "reflection property" field stores reflection property information related to the reflection property of the reflecting member.
  • the 'reflection property' field includes a 'reflection type' field, a 'reflectance' field, and a 'normal angle' field.
  • the reflection type is one of the following. ⁇ Diffuse reflection ⁇ Specular reflection
  • the "reflectance" field stores the reflectance value of the reflective member.
  • the "normal angle” field stores the value of the normal angle of the reflecting surface of the reflecting member.
  • the spatial shape information includes reflection characteristic information as the attribute information of each three-dimensional position. Attributes (eg, ventilation characteristics) related to outlets, exterior openings (eg, windows, vents), air conditioners 40, fans (eg, fans), etc.).
  • FIG. 9 is a diagram showing the data structure of the sensor data table of the first embodiment.
  • the sensor data table may be included in the spatial shape information used by the measuring device 10 .
  • the sensor data table stores information about the sound wave transmitter 20 and the sound wave receiver 30 (hereinafter referred to as "sensor information").
  • the sensor data table includes a "sensor ID” field, a “coordinates” field, and a “sensor type” field. Each field is associated with each other.
  • the "sensor ID” field stores sensor identification information that identifies the sound wave transmitting device 20 or the sound wave receiving device 30.
  • the “coordinates” field stores coordinates indicating the position of the sound wave transmitting device 20 or the sound wave receiving device 30 (hereinafter referred to as "sensor coordinates"). Sensor coordinates are expressed in a spatial coordinate system.
  • the "sensor type” field stores the tag "transmission” indicating that it is the sound wave transmitting device 20 or the tag "receiving” indicating that it is the sound wave receiving device 30.
  • FIG. 10 is a diagram showing the data structure of the route data table of the first embodiment.
  • the route data table may be included in the spatial shape information used by the measuring device 10 .
  • the route data table stores route information about routes.
  • the route data table includes a “route ID” field, a “transmitting sensor” field, and a “receiving sensor” field.
  • the "route ID" field stores route identification information that identifies the route.
  • the "transmission sensor” field stores the sensor identification information of the sound wave transmitters 20 that make up the route.
  • the "receiving sensor” field stores the sensor identification information of the sound wave receiving device 30 that constitutes the route.
  • FIG. 11 is a diagram showing the data structure of the mesh data table of the first embodiment.
  • FIG. 12 is an explanatory diagram of the filter of the first embodiment.
  • the mesh data table may be included in the spatial shape information used by the measuring device 10 .
  • the mesh data table stores mesh information about meshes.
  • the mesh data table includes a "mesh ID” field, a "coordinate” field, a "path ID” field, and a "filter” field.
  • the "mesh ID" field stores mesh identification information that identifies the mesh.
  • the "coordinates" field stores mesh coordinates indicating the position of the mesh.
  • Mesh coordinates are expressed in a spatial coordinate system.
  • Mesh coordinates or other parameters may be defined with reference to BIM (Building Information Modeling) or other CAD (Computer-Aided Design) data.
  • the "path ID” field stores the path identification information of the sound wave propagation path through the mesh.
  • the "filter” field stores filter information regarding a filter for extracting a specific waveform from the waveform of the ultrasonic beam reproduced by the reception waveform data received by the sound wave receiving device 30.
  • the filter information is associated with the route identification information stored in the "route ID" field.
  • the 'filter' field includes a 'temporal filter' field and an 'amplitude filter' field. Filter information may be stored in the route data table instead of the mesh data table.
  • the "temporal filter” field stores information about a temporal filter for extracting a specific waveform along the time axis.
  • Temporal filters are, for example, at least one of the following (FIG. 12).
  • ⁇ Lower limit time threshold THtb ⁇ Upper limit time threshold THtt
  • the "amplitude filter” field stores information about an amplitude filter for extracting a particular waveform along the amplitude axis.
  • the amplitude filter is, for example, at least one of the following (FIG. 12).
  • ⁇ Lower limit amplitude threshold value THab ⁇ Upper limit amplitude threshold THat
  • FIG. 13 is a flowchart of physical quantity estimation processing according to the first embodiment.
  • FIG. 14 is a diagram showing an example of sensor arrangement according to the first embodiment.
  • FIG. 15 is a detailed flow chart of acquisition of propagation time information in FIG.
  • the measuring device 10 acquires spatial shape information (S1100). Specifically, the processor 12 stores spatial shape information stored in the storage device 11 (for example, a spatial data table (FIG. 8), a sensor data table (FIG. 9), a path data table (FIG. 10), and a mesh data table. (FIG. 11)).
  • a spatial data table for example, a spatial data table (FIG. 8), a sensor data table (FIG. 9), a path data table (FIG. 10), and a mesh data table. (FIG. 11)).
  • the measuring device 10 acquires propagation time information (S1101). Specifically, the processor 12 acquires propagation time information of sound waves for the target path in the target space SP.
  • a target route is a route whose propagation time is to be measured.
  • a plurality of sound wave transmitters 20a to 20e and a plurality of sound wave receivers 30a to 30e are arranged in the target space SP.
  • the plurality of sound wave transmitters 20a-20e are opposed to the plurality of sound wave receivers 30a-30e, respectively.
  • the sound wave transmitter 20a faces the sound wave receiver 30a.
  • the sound wave transmitter 20a and the sound wave receiver 30a form a sensor pair.
  • five sensor pairs are formed.
  • At least one path is defined for each sensor pair.
  • the path is not limited to the propagation path of direct waves, and may include the propagation paths of sound waves that have been reflected one or more times. That is, multiple paths can be defined for the same sensor pair.
  • the route information stored in the route data table can include information about the orientation of the transmitting sensor or receiving sensor (that is, the transmitting or receiving direction of sound waves). Furthermore, by using reflection, it is also possible to define a path between the sound wave transmitting device 20 and the sound wave receiving device 30 that are not opposed to each other. In short, each path corresponds to a combination of the positions and orientations of the sound wave transmitting device 20 and the sound wave receiving device 30 (that is, the sound wave transmitting direction or the sound wave receiving direction).
  • the measuring device 10 acquires propagation time information (S1101) according to FIG. First, the measuring device 10 determines a target route (S1110). Specifically, the processor 12 determines a route whose propagation time is to be measured (hereinafter referred to as a “target route”). As an example, processor 12 identifies the route identification information for at least some of the records stored in the route data table (FIG. 10).
  • the measuring device 10 outputs an ultrasonic beam (S1111).
  • processor 12 refers to the route data table (FIG. 10) and selects one of the route identification information specified in step S1110 that has not yet been selected.
  • the processor 12 stores the information in the "transmitting sensor" field associated with the selected route identification information (that is, the sound wave transmitting device to be controlled (hereinafter referred to as “target sound wave transmitting device") 20) and the "receiving sensor” Field information (that is, the sound wave receiving device to be controlled (hereinafter referred to as “target sound wave receiving device”) 30) is specified.
  • the processor 12 transmits ultrasound control signals to the target ultrasound transmitter 20 .
  • the target sound wave transmitting device 20 transmits an ultrasonic beam according to the ultrasonic control signal transmitted from the measuring device 10 .
  • the plurality of ultrasonic transducers 21 vibrate simultaneously according to the ultrasonic control signal.
  • an ultrasonic beam traveling in the transmission direction is transmitted from the target sound wave transmitting device 20 toward the target sound wave receiving device 30 .
  • the measuring device 10 acquires received waveform data (S1112). Specifically, the ultrasonic transducer 31 of the target sound wave receiving device 30 vibrates by receiving the ultrasonic beam transmitted from the target sound wave transmitting device 20 in step S1111.
  • the control circuit 32 generates reception waveform data (FIG. 12) according to vibration of the ultrasonic transducer 31 .
  • the control circuit 32 transmits the generated received waveform data to the measuring device 10 .
  • the processor 12 acquires received waveform data transmitted from the sound wave receiving device 30 .
  • the processor 12 may perform signal processing such as amplification and band limiting processing on the acquired received waveform data.
  • the measuring device 10 performs filtering (S1113). Specifically, the processor 12 identifies the “filter” field associated with the route identification information of the target route being selected, for example, by referring to the mesh data table (FIG. 11). For example, when the route identification information of the target route is "P001", the processor 12 identifies filter information of "not less than the time threshold THtb1 and less than the time threshold THtt1, and not less than the amplitude threshold THab1 and less than the amplitude threshold THat1". If the path identification information of the target path is "P002", the processor 12 identifies the filter information "within the time window Wt2 and the amplitude window Wa2".
  • the processor 12 Based on the identified filter information, the processor 12 extracts the component of the ultrasonic beam traveling along the target path from among the components included in the received waveform data.
  • the measuring device 10 specifies the propagation time (S1114).
  • the processor 12 refers to the "coordinates" field of the sensor data table (FIG. 9), and specifies the coordinates of the sound wave transmitter 20 and the sound wave receiver 30 that constitute the sensor pair for each sensor pair. .
  • the processor 12 calculates the path length Ds of the target path based on the specified combination of the coordinates of the acoustic wave transmitter 20 and the coordinates of the acoustic wave receiver 30 .
  • the path length corresponds to the distance between the sound wave transmitting device 20 and the sound wave receiving device 30 .
  • the path length Ds can be calculated with reference to the coordinates and attitudes of the sound wave transmitting device 20 and the sound wave receiving device 30, as well as the spatial shape information.
  • the path length Ds may be stored in advance in the path data table (FIG. 10).
  • the processor 12 identifies the time (hereinafter referred to as “propagation time”) t corresponding to the peak value of the component extracted in step S1113.
  • the propagation time t is the time required from when the ultrasonic beam is transmitted by the ultrasonic wave transmitting device 20 until the ultrasonic beam traveling along the target path reaches the ultrasonic wave receiving device 30 (that is, from the start point to the end point of the target path). time for the ultrasonic beam to propagate up to ).
  • step S1114 If step S1114 has not been completed for all target routes (S1115-NO), the measuring device 10 executes step S1111.
  • step S1114 When step S1114 is completed for all target routes Pi (S1115-YES), the measuring device 10 terminates acquisition of propagation time information (S1101) in FIG.
  • the measuring device 10 executes model input generation (S1102). Specifically, the processor 12 generates model inputs by referring to the spatial shape information obtained in step S1100 and the propagation time information obtained in step S1101. As an example, processor 12 generates model inputs by vectorizing values contained in spatial shape information and propagation time information.
  • the measuring device 10 estimates the physical quantity distribution (S1103).
  • the processor 12 gives the model input generated in step S1102 to the learned inference model (eg, inference model MD10).
  • the inference model makes an inference about the distribution of physical quantities over the target space SP based on the model input, and returns an estimation result.
  • the processor 12 determines and outputs distribution information (for example, distribution information DI11).
  • the processor 12 uses the output distribution information, for example, to control the air conditioner 40 or to present information (for example, for the user of the target space or the administrator of the target space to grasp the environment in the target space). screen generation and image output to the display unit).
  • the measuring device 10 acquires spatial shape information and propagation time information.
  • the measuring device 10 generates a model input with reference to the spatial shape information and the propagation time information.
  • the measuring device 10 estimates the distribution of physical quantities over the target space SP by giving model inputs to the learned inference model.
  • the distribution of physical quantities over the target space can be estimated in a short period of time compared to normal CFD simulation.
  • the measuring apparatus 10 does not require boundary conditions regarding the target space SP, it is possible to estimate the distribution of physical quantities over the target space SP even if such information is unknown. That is, it is possible to avoid deterioration of convergence (analysis divergence) that occurs when CFD analysis is performed using boundary conditions with low accuracy, thereby improving calculation results.
  • FIG. 16 is an explanatory diagram of machine learning for constructing an inference model used in the second embodiment.
  • FIG. 17 is an explanatory diagram of the outline of the second embodiment.
  • the measurement device 10 uses an inference model constructed by machine learning to determine the distribution information of the target space.
  • the inference model MD20 used by the measuring device 10 is constructed by machine learning (supervised learning) using learning data including learning input and correct data.
  • Machine learning may be performed by the measuring device 10 or may be performed by an external device (for example, a learning server).
  • the learning input is generated based on the boundary condition BC20, the spatial shape information SS10, and the propagation time information PT10.
  • Correct data is generated based on the distribution information DI10.
  • the boundary condition BC20 is, for example, a condition relating to the inflow or outflow of fluid into the space SP10 (not shown) for collecting learning data.
  • Boundary condition BC20 is obtained, for example, in response to a user instruction to an input device.
  • the boundary condition BC20 may be data at one point in time or time-series data.
  • Boundary condition BC20, or other boundary conditions described herein, may include, for example, at least one of the following information. ⁇ Air volume or temperature at air inlet, outlet, external opening, air conditioner 40, or fan (e.g., electric fan) ⁇ Wall surface temperature ⁇ Wall heat transfer ⁇ Wall heat capacity ⁇ Temperature of outer surface of wall position and calorific value
  • the boundary condition BC20 may be referenced in the CFD simulation for estimating the distribution information DI10.
  • the inference model MD20 can generate physical quantities over a given space from the spatial shape and boundary conditions of a given space, and the propagation time of sound waves along paths in the space. distribution can be estimated with reasonable accuracy.
  • the measuring device 10 generates a model input by referring to the boundary condition BC21, spatial shape information SS11, and propagation time information PT11.
  • the measurement device 10 provides model inputs to the inference model MD20.
  • the inference model MD20 makes inferences based on model inputs and returns distribution information DI21.
  • the boundary condition BC21 is, for example, a condition regarding the inflow or outflow of fluid into the target space SP.
  • the boundary condition BC21 may be data at one point in time or time-series data.
  • the boundary condition BC21 is acquired, for example, in response to a user's instruction to the input device.
  • the distribution information DI21 is the result of estimating the distribution of physical quantities over the target space SP by the inference model MD20.
  • the measuring device 10 of the second embodiment performs inference using model inputs including boundary conditions. Therefore, according to the measurement apparatus 10, inference is performed in consideration of boundary conditions, so that the accuracy of estimation regarding the distribution of physical quantities over the target space SP can be improved compared to the first embodiment.
  • FIG. 18 is an explanatory diagram of the outline of the third embodiment.
  • FIG. 19 is a diagram showing another example of FIG.
  • the measuring device 10 of the third embodiment uses a combination of an inference model constructed by machine learning and a CFD simulation to determine distribution information of the target space SP.
  • CFD simulations use continuity equations, Navier-Stokes equations, or energy equations given initial conditions (e.g., initial distribution of temperature/wind speed (including wind direction)), spatial geometry, and boundary conditions. It is a method to obtain the temperature/wind speed distribution in the space.
  • initial conditions e.g., initial distribution of temperature/wind speed (including wind direction)
  • spatial geometry e.g., initial distribution of temperature/wind speed (including wind direction)
  • boundary conditions e.g., initial distribution of temperature/wind speed (including wind direction)
  • boundary conditions e.g., initial distribution of temperature/wind speed (including wind direction)
  • the measurement device 10 can use the inference model MD10 described in the first embodiment. As shown in FIG. 18, the measuring device 10 generates a model input with reference to spatial shape information SS11 and propagation time information PT11. The measurement device 10 provides model inputs to the inference model MD10. The inference model MD10 makes inferences based on model inputs and returns distribution information DI11.
  • the measuring device 10 performs a CFD simulation with reference to the spatial shape information SS11, the boundary conditions BC21, and the distribution information DI11.
  • the distribution information DI11 corresponds to initial conditions in the CFD simulation.
  • the measuring device 10 determines the distribution information DI32 by performing a CFD simulation.
  • the distribution information DI32 is information on the distribution of physical quantities over the target space SP.
  • the measuring device 10 can use the inference model MD20 described in the second embodiment. As shown in FIG. 19, the measuring device 10 generates a model input with reference to the boundary condition BC21, the spatial shape information SS11, and the propagation time information PT11. The measurement device 10 provides model inputs to the inference model MD20. The inference model MD20 makes inferences based on model inputs and returns distribution information DI21.
  • the measuring device 10 performs a CFD simulation with reference to the spatial shape information SS11, the boundary conditions BC21, and the distribution information DI21.
  • the distribution information DI21 corresponds to initial conditions in the CFD simulation.
  • the measuring device 10 determines the distribution information DI42 by performing a CFD simulation.
  • the distribution information DI42 is information on the distribution of physical quantities over the target space SP.
  • the measurement device 10 of the third embodiment uses a combination of an inference model (eg, inference model MD10 or inference model MD20) and CFD simulation. Therefore, according to this measuring device 10, since it is possible to set reasonable initial conditions obtained by inference, the CFD simulation can be converged in a short time. Moreover, according to this measuring device 10, by using CFD simulation, highly accurate distribution information can be obtained compared with the case where an inference model is used alone.
  • an inference model eg, inference model MD10 or inference model MD20
  • FIG. 20 is a detailed flowchart of analysis processing according to the third embodiment.
  • the analysis process starts upon completion of the physical quantity estimation process (Fig. 13).
  • the measuring device 10 sets initial conditions (S1140). Specifically, the processor 12 acquires distribution information (for example, distribution information DI11 or distribution information DI21) obtained by the physical quantity estimation process (FIG. 13). The processor 12 sets the initial conditions of the target space SP by referring to the acquired distribution information.
  • distribution information for example, distribution information DI11 or distribution information DI21
  • the measuring device 10 performs spatial configuration setting (S1141). Specifically, the processor 12 acquires spatial shape information (for example, spatial shape information SS11) of the target space SP used in the physical quantity estimation process (FIG. 13). The processor 12 refers to the acquired spatial shape information and sets the spatial shape of the target space SP.
  • spatial shape information for example, spatial shape information SS11
  • the measuring device 10 sets boundary conditions (S1142). Specifically, the processor 12 acquires a boundary condition (for example, a boundary condition BC21) regarding the target space SP. Processor 12 sets the obtained boundary conditions. As an example, the processor 12 may set the wind velocity in the direction perpendicular to the wall to 0 at the position of the wall in the target space SP.
  • a boundary condition for example, a boundary condition BC21
  • the processor 12 may set the wind velocity in the direction perpendicular to the wall to 0 at the position of the wall in the target space SP.
  • the measuring device 10 analyzes the physical quantity distribution.
  • the processor 12 refers to the settings in steps S1140 to S1142 and performs the CFD simulation. That is, the processor 12 performs numerical calculations regarding the distribution of physical quantities over the target space SP. Accordingly, processor 12 determines and outputs distribution information (eg, distribution information DI32 or distribution information DI42). The processor 12 uses the output distribution information, for example, to control the air conditioner 40 or to present information (for example, for the user of the target space or the administrator of the target space to grasp the environment in the target space). screen generation and image output to the display unit). With the end of step S1143, the analysis processing (FIG. 20) ends.
  • the measurement device 10 of the third embodiment uses a combination of an inference model (eg, inference model MD10 or inference model MD20) and CFD simulation.
  • an inference model eg, inference model MD10 or inference model MD20
  • CFD simulation highly accurate distribution information can be obtained as compared with the case of using an inference model alone.
  • FIG. 21 is an explanatory diagram of machine learning for constructing an inference model used in the fourth embodiment.
  • FIG. 22 is an explanatory diagram of the outline of the fourth embodiment.
  • the measurement device 10 uses a combination of an inference model constructed by machine learning and a CFD simulation to determine distribution information of the target space SP.
  • the inference model MD50 used by the measuring device 10 is constructed by machine learning (supervised learning) using learning data including learning input and correct data.
  • Machine learning may be performed by the measuring device 10 or may be performed by an external device (for example, a learning server).
  • the learning input is generated based on the spatial shape information SS10 and the propagation time information PT10.
  • the correct data is generated based on the boundary condition BC20 and the distribution information DI10.
  • the inference model MD50 can obtain the distribution of physical quantities over the given space and the It is possible to infer spatial boundary conditions with reasonable accuracy.
  • the measuring device 10 generates a model input by referring to spatial shape information SS11 and propagation time information PT11.
  • the measurement device 10 provides model inputs to the inference model MD50.
  • the inference model MD50 performs inference based on the model input and returns boundary conditions BC51 and distribution information DI51.
  • the boundary condition BC51 is the estimation result of the boundary condition regarding the target space SP by the inference model MD50.
  • the distribution information DI51 is the result of estimating the distribution of physical quantities over the target space SP by the inference model MD50.
  • the measuring device 10 performs a CFD simulation with reference to the spatial shape information SS11, the boundary conditions BC51, and the distribution information DI51.
  • the distribution information DI51 corresponds to initial conditions in the CFD simulation.
  • the measuring device 10 determines the distribution information DI52 by performing a CFD simulation.
  • the distribution information DI52 is information on the distribution of physical quantities over the target space SP.
  • the measuring device 10 of the fourth embodiment uses the inference model MD50 and the CFD simulation in combination. Therefore, according to this measuring device 10, since it is possible to set reasonable initial conditions obtained by inference, the CFD simulation can be converged in a short time. Moreover, according to this measuring apparatus 10, by using CFD simulation, highly accurate distribution information can be obtained as compared with the case where the inference model MD50 is used alone. Furthermore, according to this measurement apparatus 10, the boundary condition BC51 can be obtained by reasoning, so even if the boundary condition regarding the target space SP is unknown, the CFD simulation can be executed. Note that the inference model may estimate some of the boundary conditions used in the CFD simulation. In this case, the rest of the boundary conditions used for the CFD simulation are obtained, for example, in response to user instructions to the input device.
  • FIG. 23 is an explanatory diagram of machine learning for constructing an inference model used in the fifth embodiment.
  • FIG. 24 is an explanatory diagram of the outline of the fifth embodiment.
  • the measuring device 10 uses a combination of an inference model constructed by machine learning and a two-stage CFD simulation to determine distribution information of the target space SP.
  • the inference model MD60 used by the measuring device 10 is constructed by machine learning (supervised learning) using learning data including learning input and correct data.
  • Machine learning may be performed by the measuring device 10 or may be performed by an external device (for example, a learning server).
  • the learning input is generated based on the boundary condition BC20, the spatial shape information SS10, the propagation time information PT10, and the distribution information DI60a.
  • Correct data is generated based on the distribution information DI10.
  • the distribution information DI60a is information about the distribution of physical quantities over the space SP10. Specifically, the distribution information DI60a is obtained by roughly calculating the distribution of physical quantities over the space SP10 by CFD simulation. For example, the distribution information DI60a corresponds to a solution in the process of convergence of the CFD simulation regarding the physical quantity over the space SP10.
  • the inference model MD60 can determine the spatial shape of a given space, the propagation time of a sound wave on a path in that space, the boundary conditions for that space, and the From the distribution of physical quantities over the space (coarse accuracy), it is possible to estimate the distribution of physical quantities over the space with reasonable accuracy.
  • the measuring device 10 performs the first CFD simulation with reference to the initial condition IC61, the spatial shape information SS11, and the boundary condition BC21 to determine the distribution information DI61a.
  • the initial condition IC61 relates to the distribution of physical quantities over the target space SP.
  • the initial condition IC61 is acquired, for example, in response to a user's instruction to the input device. Alternatively, the initial condition IC61 may be randomly determined.
  • the distribution information DI61a is information on the distribution of physical quantities over the target space SP. Specifically, the distribution information DI61a is obtained by roughly calculating the distribution of physical quantities over the target space SP through a CFD simulation. For example, the distribution information DI61a corresponds to a solution in the middle of convergence of the CFD simulation regarding the physical quantity over the target space SP.
  • the measuring device 10 generates a model input by referring to the boundary condition BC21, spatial shape information SS11, propagation time information PT11, and distribution information DI61a.
  • the measurement device 10 provides model inputs to the inference model MD60.
  • the inference model MD60 makes inferences based on model inputs and returns distribution information DI61.
  • the distribution information DI61 is the estimation result of the physical quantity distribution over the target space SP by the inference model MD60.
  • the measuring device 10 performs the second CFD simulation with reference to the spatial shape information SS11, the boundary condition BC21, and the distribution information DI61.
  • the distribution information DI61 corresponds to initial conditions in the CFD simulation.
  • the measuring device 10 determines the distribution information DI62 by performing CFD simulation.
  • the distribution information DI62 is information on the distribution of physical quantities over the target space SP.
  • the measuring device 10 of the fifth embodiment uses a combination of the inference model MD60 and the two-stage CFD simulation. Therefore, according to this measuring device 10, the inference is performed in consideration of the result of numerical calculation (that is, the distribution information DI 61a) by rough CFD simulation, so the accuracy of the inference is improved. In addition, according to this measuring apparatus 10, it is possible to set reasonable initial conditions obtained by inference, so that the CFD simulation can be converged in a short period of time. Furthermore, according to the measuring apparatus 10, by using the CFD simulation, it is possible to obtain highly accurate distribution information as compared with the case of using the inference model MD60 alone.
  • the inference model MD60 may be modified such that the model inputs do not include boundary conditions, such as the inference model MD10.
  • inference model MD60 may be modified to estimate boundary conditions in addition to distribution information, such as inference model MD50.
  • FIG. 25 is an explanatory diagram of machine learning for constructing an inference model used in the sixth embodiment.
  • FIG. 26 is an explanatory diagram of the outline of the sixth embodiment.
  • the measuring device 10 uses a combination of an inference model constructed by machine learning, a CFD simulation (one stage or two stages), and a data assimilation model to determine distribution information of the target space SP.
  • the inference model MD70 used by the measuring device 10 is constructed by machine learning (supervised learning) using learning data including learning input and correct data.
  • Machine learning may be performed by the measuring device 10 or may be performed by an external device (for example, a learning server).
  • the learning input is generated based on the spatial shape information SS10 and the distribution information DI60a.
  • Correct data is generated based on the distribution information DI10.
  • the inference model MD70 can obtain the distribution of physical quantities over the space from the spatial shape of a given space and the distribution of physical quantities over the space (coarse accuracy). can be estimated with reasonable accuracy.
  • the measuring device 10 generates a model input by referring to the spatial shape information SS11 and the distribution information DI61a.
  • the distribution information DI61a is obtained by performing a CFD simulation as in FIG.
  • the measurement device 10 provides model inputs to the inference model MD70.
  • the inference model MD70 makes inferences based on model inputs and returns distribution information DI71.
  • the distribution information DI71 is the result of estimating the distribution of physical quantities over the target space SP by the inference model MD70.
  • the measuring device 10 performs a CFD simulation with reference to the spatial shape information SS11, the boundary conditions BC21, and the distribution information DI71.
  • the distribution information DI71 corresponds to initial conditions in the CFD simulation.
  • the measuring device 10 determines the distribution information DI72 by performing a CFD simulation.
  • the distribution information DI72 is information on the distribution of physical quantities over the target space SP.
  • the measuring device 10 performs data assimilation processing by giving the distribution information DI72 and the propagation time information PT11 to the data assimilation model.
  • Metrology device 10 may use a sequential Bayesian filter, a variational Bayesian filter, a Kalman filter, or other data assimilation models. Thereby, the measuring device 10 determines the distribution information DI73.
  • the distribution information DI73 is information on the distribution of physical quantities over the target space SP.
  • the measuring device 10 of the sixth embodiment uses the inference model MD70, the CFD simulation, and the data assimilation model in combination. Therefore, according to the measuring device 10, the numerical calculation result (that is, the distribution information DI72) by the CFD simulation is corrected according to the actual measurement result (that is, the propagation time information) regarding the distribution of the physical quantity in the target space SP. Highly accurate distribution information DI73 can be obtained.
  • the inference model MD70 may be modified to include boundary conditions in the model input, for example, like the inference model MD20. Alternatively, inference model MD70 may be modified to estimate boundary conditions in addition to distribution information, such as inference model MD50.
  • the inference model MD70 may be modified to include propagation time information PT11a in the model input instead of distribution information DI61a. In this case, the CFD simulation for obtaining the distribution information DI61a becomes unnecessary.
  • the storage device 11 may be connected to the measuring device 10 via the network NW.
  • NW network
  • the measuring device 10 can also be implemented by a client/server type system.
  • processor 12 performs inference, simulation, and data assimilation using a learning model.
  • at least one of inference using the learning model, simulation, and data assimilation may be performed by general-purpose or dedicated hardware other than processor 12 .
  • the example of acquiring air characteristic distribution information through CFD simulation was mainly explained, but the specific method of simulation is not limited.
  • a simulation using a ROM Reduced Order Model
  • the amount of simulation calculation can be reduced, and the calculation time can be shortened.
  • a simulation combining three-dimensional CFD and ROM may be performed.
  • the information about the heat source existing in the space can include information about the heat quantity (for example, heat generation quantity) of the heat source.
  • a heat source can be an object (eg, an air conditioner, a stove) or a living organism (eg, a person).
  • the inference model described above may be modified to include additional conditions in the model inputs. This makes it possible to improve the accuracy of inference.
  • the inference model may be modified to estimate additional conditions (or parts thereof) in addition to distribution information (and boundary conditions). As a result, even if the additional conditions are unknown, it is possible to perform a simulation considering the additional conditions.
  • the sound wave receiving device 30 may include a plurality of ultrasonic transducers 31 as in the sound wave transmitting device 20 .
  • the sound wave transmitter 20 may include a plurality of ultrasonic transducers 21 arranged in an array.
  • the plurality of ultrasonic transducers 21 may be controlled by individual drive control signals, may be controlled by the same drive control signal in group units, or may be controlled by the same drive control signal as a whole. good too.
  • the sound wave receiving device 30 may include a plurality of ultrasonic transducers 31 arranged in an array.
  • the plurality of ultrasonic transducers 31 may be controlled by individual drive control signals, may be controlled by the same drive control signal in group units, or may be controlled by the same drive control signal as a whole. good too.
  • one sound wave transmitting device 20 transmits ultrasonic beams along a plurality of paths
  • one sound wave receiving device 30 receives ultrasonic beams along a plurality of paths.
  • Each of the n (n is an integer equal to or greater than 2) sound wave transmitters 20 is an ultrasonic beam along one path (that is, the n sound wave transmitters 20 are ultrasonic beams along n paths)
  • each of the n acoustic wave receivers 30 may receive an ultrasound beam along each path (i.e., n acoustic wave receivers 30 may transmit ultrasound along n paths beam).
  • the sound wave transmitting device 20 may transmit an ultrasonic beam including an autocorrelation signal with relatively strong autocorrelation (for example, M-sequence signal, Gold code, etc.). Thereby, the S/N ratio of the measurement result of the temperature of the space can be further improved.
  • an autocorrelation signal with relatively strong autocorrelation for example, M-sequence signal, Gold code, etc.
  • the sound wave receiving device 30 may identify the sound wave transmitting device 20 that is the source of the ultrasonic beam by having the sound wave transmitting devices 20 individually transmit ultrasonic beams containing different autocorrelation signals. Further, by transmitting ultrasonic beams having different oscillation frequencies for each of the sound wave transmitting devices 20, the sound wave receiving device 30 may identify the sound wave transmitting device 20 that is the source of the ultrasonic beam.
  • the measuring device 10 calculates the following air characteristic distributions based on the propagation characteristics of sound waves (for example, propagation time, amplitude change, phase change, frequency change, etc.). Measurement is also possible.
  • ⁇ Concentration distribution of chemical substances (e.g. CO2) in the air ⁇ Humidity distribution ⁇ Odor distribution ⁇ Toxic gas distribution
  • one ultrasonic transducer may have a function of transmitting ultrasonic waves and a function of receiving ultrasonic waves.
  • the sound wave transmitting device 20 transmits an ultrasonic beam having directivity.
  • this embodiment is also applicable when the sound wave transmitter 20 transmits an audible sound beam (that is, a sound wave having a frequency different from that of the ultrasonic beam).
  • sensing is performed in the measurement target space using the sound wave transmitting device 20 and the sound wave receiving device 30 as sensors, and the propagation time information of the ultrasonic waves acquired as sensing information is used for model input and data assimilation.
  • An example of calculating distribution information of air properties in a target space has been described.
  • the information is not limited to this, and the air characteristic distribution information may be calculated using sensing information other than the ultrasonic wave propagation time information.
  • the measuring device 10 may use, as sensing information, temperature information acquired from a temperature sensor installed in the measurement target space, and wind speed information and wind direction information acquired from a wind speed sensor installed in the measurement target space.
  • the measuring device 10 may use at least one of temperature information, wind speed information, and wind direction information instead of the above-described propagation time information PT10 and PT11.
  • the measuring device 10 determines distribution information DI72 by performing a CFD simulation with reference to spatial shape information SS11, boundary conditions BC21, and distribution information DI71 as an initial condition. Then, the measuring device 10 gives the distribution information DI72 and the wind speed information acquired from the wind speed sensor to the data assimilation model and performs data assimilation processing, thereby determining the distribution information DI73.
  • the measuring device 10 may combine at least one of the temperature information, the wind speed information, and the wind direction information with the propagation time information of the ultrasonic wave and use it for model input or data assimilation. As a result, it is possible to measure the distribution of the air properties in the space with higher accuracy.
  • Reference Signs List 1 air conditioning system 10: measuring device 11: storage device 12: processor 13: input/output interface 14: communication interface 20: sound wave transmitting device 21: ultrasonic transducer 22: control circuit 30: sound wave receiving device 31: ultrasonic transducer 32: Control circuit 40: Air conditioner 50: Thermometer

Landscapes

  • Engineering & Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Mechanical Engineering (AREA)
  • Combustion & Propulsion (AREA)
  • Theoretical Computer Science (AREA)
  • Environmental & Geological Engineering (AREA)
  • Software Systems (AREA)
  • Mathematical Physics (AREA)
  • Signal Processing (AREA)
  • Tourism & Hospitality (AREA)
  • Human Resources & Organizations (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Economics (AREA)
  • General Health & Medical Sciences (AREA)
  • Environmental Sciences (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Ecology (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Biodiversity & Conservation Biology (AREA)
  • Computing Systems (AREA)
  • Atmospheric Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Fuzzy Systems (AREA)
  • Human Computer Interaction (AREA)
  • Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)

Abstract

Un dispositif de traitement d'informations selon un aspect de la présente divulgation comprend : un moyen d'acquisition d'informations de forme spatiale, concernant la forme spatiale d'un espace cible ; un moyen d'acquisition d'informations de détection à partir d'un capteur de l'espace cible ; un moyen de référence aux informations de forme spatiale et aux informations de détection, permettant de générer une entrée de modèle ; et un moyen permettant de donner l'entrée de modèle à un modèle instruit, afin de déduire une distribution de grandeur physique dans l'espace cible.
PCT/JP2022/019874 2021-05-26 2022-05-11 Dispositif et procédé de traitement d'informations et programme WO2022249882A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2021088271 2021-05-26
JP2021-088271 2021-05-26

Publications (1)

Publication Number Publication Date
WO2022249882A1 true WO2022249882A1 (fr) 2022-12-01

Family

ID=84229948

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2022/019874 WO2022249882A1 (fr) 2021-05-26 2022-05-11 Dispositif et procédé de traitement d'informations et programme

Country Status (1)

Country Link
WO (1) WO2022249882A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014130113A (ja) * 2012-12-28 2014-07-10 Panasonic Corp 超音波送信装置、超音波受信装置、超音波送信方法、超音波受信方法、温度測定のための伝搬時間計測システム、及び温度測定のための伝搬時間計測方法
JP2019158819A (ja) * 2018-03-16 2019-09-19 学校法人東京女子医科大学 計測支援システム、計測支援方法及び計測支援プログラム
JP2020106153A (ja) * 2018-12-26 2020-07-09 株式会社日立製作所 空調制御システム及び方法
JP2020118536A (ja) * 2019-01-23 2020-08-06 大成建設株式会社 風速分布推定装置及び風速分布推定方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2014130113A (ja) * 2012-12-28 2014-07-10 Panasonic Corp 超音波送信装置、超音波受信装置、超音波送信方法、超音波受信方法、温度測定のための伝搬時間計測システム、及び温度測定のための伝搬時間計測方法
JP2019158819A (ja) * 2018-03-16 2019-09-19 学校法人東京女子医科大学 計測支援システム、計測支援方法及び計測支援プログラム
JP2020106153A (ja) * 2018-12-26 2020-07-09 株式会社日立製作所 空調制御システム及び方法
JP2020118536A (ja) * 2019-01-23 2020-08-06 大成建設株式会社 風速分布推定装置及び風速分布推定方法

Similar Documents

Publication Publication Date Title
US8996292B2 (en) Apparatus and method generating a grid map
JP4269623B2 (ja) 血流可視化診断装置
Marichelvam et al. A bat algorithm for realistic hybrid flowshop scheduling problems to minimize makespan and mean flow time
Wrona et al. Shaping the acoustic radiation of a vibrating plate
JP7030368B2 (ja) 情報処理装置、情報処理方法、及び、プログラム
WO2022249882A1 (fr) Dispositif et procédé de traitement d'informations et programme
Pal et al. Human-like sensing for robotic remote inspection and analytics
JP4877744B2 (ja) 建物の音シミュレーションシステム
CN110390124B (zh) 确定物体形状的方法和设备、制造计算装置的方法
WO2020116022A1 (fr) Système, procédé de commande et dispositif de commande de capteur
WO2021199877A1 (fr) Système de commande, procédé de commande et programme
JP2007003989A (ja) 音環境解析シミュレーションシステム
JP2022124347A (ja) 計測装置、プログラム、および計測方法
JP2022110654A (ja) 計測装置、プログラム、および計測方法
JP2021165626A (ja) 計測装置、コントローラ、プログラム、および計測方法
JP2001174339A (ja) 音波を用いた温度測定時のセンサーの最適位置決定方法
JP7001289B2 (ja) 情報処理装置、情報処理方法、及び、プログラム
WO2021145440A1 (fr) Dispositif de mesure, procédé de mesure et programme
JP2021113808A (ja) 計測装置、計測方法、及び、プログラム
Magalhães et al. A model for three-dimensional simulation of acoustic emissions from rotating machine vibration
JP2021018368A (ja) 形状最適化システム、形状最適化方法、及びプログラム
JP2022125505A (ja) 計測装置、計測方法、およびプログラム
JP2022131003A (ja) 計測装置、計測方法、およびプログラム
Dogaru et al. An efficient sound propagation software simulator based on cellular automata
CN116755477B (zh) 流体通道的流量自动控制调节方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22811155

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22811155

Country of ref document: EP

Kind code of ref document: A1