Detection device, detection method, and program

Abstract

The invention provides a technology for identifying or quantifying allergens of pollen, mould, mites and the like. The detection device uses a plurality of semiconductor sensors. The plurality of semiconductor sensors exhibit relatively different response characteristics to the same gas composition. The detection device acquires the signal patterns detected by the plurality of semiconductor sensors, identifies or quantifies the allergen from the plurality of signal patterns, and outputs an identification or quantification result.

Description

Technical Field

[0001] This invention relates to a detection device, a detection method, and a procedure. Background Technology

[0002] Allergies caused by airborne allergens plague many people. Typical allergens include pollen, mold (fungi), and indoor dust (dust mites), among which hay fever is a common allergy in Japan. Because the amount of pollen dispersed is correlated with the incidence and severity of hay fever, pollen dispersal has been measured in the past. Mold (fungi) and indoor dust (dust mites) have also been measured using morphological and immunological methods in the past.

[0003] Techniques for detecting tiny particles using lasers have been considered (see, for example, Patent Document 1). Patent Document 1 discloses a technique for detecting particles of a desired size using a simple light-receiving method by detecting light diffracted by a laser by particles having a predetermined particle size.

[0004] Existing technical documents Patent documents Patent Document 1: Japanese Patent Application Publication No. 07-294415 Summary of the Invention [The problem the invention aims to solve] However, in existing technologies, even particles of similar size to the allergen may be detected, making it difficult to identify the allergen. Therefore, quantifying each allergen is also challenging.

[0005] This invention provides a technique for identifying or quantifying allergens such as pollen, mold, and mites.

[0006] Methods for solving problems The detection device in the first aspect of the present invention is a detection device for identifying or quantifying allergens in a target space. Control Department Acquire signal patterns detected by multiple semiconductor sensors that exhibit relatively different response characteristics to the same gas composition. Based on the multiple signal patterns acquired from the multiple semiconductor sensors, allergens are identified or quantified. Output the results of identification or quantification.

[0007] According to the first aspect of the present invention, allergens such as pollen, mold, and mites can be identified or quantified.

[0008] The detection device in the second aspect of the present invention is the detection device according to the first aspect, wherein the plurality of semiconductor sensors change the signal pattern by adsorbing gas components.

[0009] The detection device in the third aspect of the present invention is the detection device according to the second aspect, wherein the signal pattern includes a pattern of current, voltage, conductivity or resistance that changes due to the adsorption of gas components on the semiconductor sensor.

[0010] The detection device in the fourth aspect of the present invention is the detection device according to the third aspect, wherein the signal pattern includes a pattern of current, voltage, conductivity or resistance that changes due to the detachment of gas components adsorbed on the semiconductor sensor.

[0011] The detection device in the fifth aspect of the present invention is the detection device according to the first to fourth aspects, wherein the gas component is introduced into the semiconductor sensor without being pressurized or liquefied.

[0012] The detection device in the sixth aspect of the present invention is the detection device according to the first to fourth aspects. The control unit inputs the multiple signal patterns into a model generated by machine learning, which relates the multiple signal patterns of the multiple semiconductor sensors to the type or amount of allergens, thereby identifying or quantifying allergens in the target space.

[0013] The detection device in the seventh aspect of the present invention is the detection device according to the first to sixth aspects, wherein the plurality of semiconductor sensors include two or more types of semiconductor sensors, and the two or more types of semiconductor sensors are composed of different metal oxides.

[0014] The detection device in the eighth aspect of the present invention is the detection device according to the fourth aspect, wherein the conductivity is repeated in a cycle of decrease and recovery. The control unit extracts one waveform between the maximum and minimum values ​​of the conductivity as the signal pattern, and repeatedly uses the signal pattern to identify or quantify allergens.

[0015] The detection device in the ninth aspect of the present invention is the detection device according to the eighth aspect, wherein the control unit controls the environmental equipment to reduce or prevent the amount of allergen from increasing when the quantitative amount of allergen exceeds a threshold or when the slope of the amount of allergen relative to time exceeds a threshold.

[0016] The detection device in the 10th aspect of the present invention is the detection device according to the 1st to 9th aspects, wherein the control unit displays the change in the amount of allergen over time according to the type of allergen.

[0017] The detection device in the 11th aspect of the present invention is the detection device according to the 1st to 10th aspects, wherein the allergen is pollen, mold or dust mites.

[0018] The detection device in the 12th aspect of the present invention is the detection device according to the 1st to 11th aspects, wherein the plurality of semiconductor sensors simultaneously output signals regarding the plurality of gas components. The control unit acquires the signal pattern, which includes the signals simultaneously output by the plurality of semiconductor sensors.

[0019] The detection method in the 13th aspect of the present invention is a detection method for identifying or quantifying allergens in a target space using a detection device. Control Department Acquire signal patterns detected by multiple semiconductor sensors that exhibit relatively different response characteristics to the same gas composition. Based on the multiple signal patterns acquired from the multiple semiconductor sensors, allergens are identified or quantified. Output the results of identification or quantification.

[0020] According to the 13th method of the present invention, allergens such as pollen, mold, and mites can be identified or quantified.

[0021] The program in the 14th aspect of the present invention causes the computer to function as the detection device described in any one of the 1st to 12th aspects.

[0022] According to the 14th method of the present invention, allergens such as pollen, mold, and mites can be identified or quantified.

[0023] [The effects of the invention] This invention provides a technique for identifying or quantifying allergens such as pollen, mold, and mites. Attached Figure Description

[0024]

【 Figure 1 This diagram illustrates the system configuration of an allergen detection system and provides an overview of the allergen detection method.

[0025]

【 Figure 2 The diagram shows a variation of the system configuration of an allergen detection system.

[0026]

【 Figure 3 The diagram shows an example of the system configuration of an allergen detection system.

[0027]

【 Figure 4 The diagram shows an example of the hardware configuration of an information processing device.

[0028]

【 Figure 5 This is an example of a functional block diagram that describes the functions of an information processing device and environmental equipment (or sensor unit) in modules.

[0029]

【 Figure 6[Image showing a TIC chromatogram of volatile compounds obtained from 10 mg of pollen from cedar, cypress, red pine, and chestnut trees.]

[0030]

【 Figure 7 A schematic diagram illustrating the detection principle of a semiconductor sensor.

[0031]

【 Figure 8 The image shows an example of a signal pattern detected by a semiconductor sensor.

[0032]

【 Figure 9 The diagram above represents an example of a neural network used by the model generation department for learning.

[0033]

【 Figure 10 [A diagram illustrating the learning process of a neural network.]

[0034]

【 Figure 11 The diagram above represents an example of a neural network used by the model generation unit to learn an allergen identification model.

[0035]

【 Figure 12 The figure shows an example of a sampling method for the signal pattern and resistance value detected by a semiconductor sensor.

[0036]

【 Figure 13 A patterned representation of the training data.

[0037]

【 Figure 14A An example of a timing signal waveform.

[0038]

【 Figure 14B An example of the circumscribed rectangle of a signal pattern between two maxima (or two minima).

[0039]

【 Figure 15 The diagram shows an example of a model structure when using a neural network corresponding to image data.

[0040]

【 Figure 16 The image shows an example of an allergen detection result screen displayed on a user terminal or remote control.

[0041]

【 Figure 17 The image shown is an example of an outdoor pollen detection result displayed on a user terminal.

[0042]

【 Figure 18 This is an example of a sequence diagram illustrating how an allergen detection system displays allergen detection results and controls the processing of environmental equipment. Detailed Implementation

[0043] The following describes an allergen detection system and an allergen detection method performed by the allergen detection system as an example of how the present invention is implemented.

[0044] <Supplementary Information on Allergen Testing Methods> Although 47.8% of Japanese people feel they have "hay fever" (see Non-Patent Literature 1 below), besides pollen, other known typical allergens include mold (fungi) and indoor dust (dust mites). Methods for evaluating environmental allergen pollution are broadly divided into morphological and immunological methods. The Durham method (see Non-Patent Literature 2) is a well-known standard method for morphological pollen analysis. The Durham method involves exposing a glass slide coated with Vaseline outdoors for 24 hours, then observing and determining under a microscope whether the attached particles are pollen. Pollen dispersal is measured in units of 1 cm³. 2 The number of sediments is represented by ].

[0045] In addition, the pollen collector developed by the British company Burkard is also well known. This pollen collector is based on Hirst's scheme (see Non-Patent Document 3), in which tape is attached to a rotating roller in a chamber that introduces atmosphere, and a person counts the pollen attached to the tape using a microscope.

[0046] In recent years, simpler automated analysis devices have become widely used as a substitute for microscopic observation. The pollen observation system provided by the Ministry of the Environment (see Non-Patent Document 4) calculates the pollen concentration (pollen particles / m³) based on the count of laser scattering caused by pollen. 3 Although this pollen observation system is automated and simple, it has the drawback of measuring particles that are similar in size to pollen grains as pollen grains (see Non-Patent Literature 5).

[0047] Therefore, some new analytical methods have also been explored. For example, the use of Coulter counting (see Non-Patent Document 6), Fourier transform infrared absorption (see Non-Patent Document 7), or Raman spectroscopy (Non-Patent Document 8) based on the change in current generated when particles pass between two electrodes is encouraged for pollen identification. In addition, attempts are being made to determine the concentration of Cry j1 (Ubisch body, i.e., Ubestite), an allergen in cedar pollen extracts, by surface plasmon resonance (Non-Patent Document 9). In addition, attempts are being made to infer the particle size based on the scattered light from ultraviolet irradiation and to identify the pollen species using the blue / red fluorescence intensity ratio (Non-Patent Document 10). Except for Cry j1, all of these are physical methods for detecting pollen.

[0048] In addition, as an immunological method, quantification of single allergens (major allergens) is being carried out using ELISA (Enzyme-linked Immunosorbent Assay).

[0049] However, standard morphological methods (Durham method, rotary drum method) require microscopic observation and are not suitable for automation. Automated analyzers that improve upon the standard method may also detect particles of similar size as allergens, and accumulated allergens must be periodically removed from the analyzer, making continuous testing impossible and resulting in low specificity.

[0050] While single-allergen quantification using the immunological method ELISA is highly specific, it cannot be performed continuously. Gas analysis methods require devices to prevent new contamination from generated gases, making them unsuitable for miniaturization. Particularly concerning mites, unlike pollen, morphological methods cannot be used, and automated analysis techniques are lacking.

[0051] <<Non-Patent Literature>> Non-patent literature 1 Novartis Pharmaceuticals Co., Ltd. "Occupational Reality Survey on Hay Fever" (Survey period: March-April 2020).

[0052] Non-patent literature 2O.C. Durham, “The volumetric incidence of atmospheric allergens; a proposed standard method of gravity sampling, counting, and volumetric interpolation of results,” J. Allergy, vol. 79, p. 79, 1946 Non-patent literature 3 J.M. Hirst, “AN AUTOMATIC VOLUMETRIC SPORE TRAP”, Annals of Applied Biology, vol. 39, p. 257, 1952 Non-patent literature 4 http: / / kafun.taiki.go.jp / Non-patent literature 5: Yoda Shigetoshi et al., “On the amount of cedar pollen intrusion into various indoor locations,” Journal of the Japanese Society of Otorhinolaryngology, Vol. 108, p. 801, 2005. Non-patent literature 6. Zhang et al., “An electronic pollen detection method using Coulter counting principle”, Atmospheric Environment, vol. 39, p. 5446, 2005 Non-patent literature 7 C.S. Pappas et al., “New Method for Pollen Identification by FT-IR Spectroscopy”, Applied Spectroscopy, vol. 57, p. 23, 2003 Non-patent literature 8F. Schulte et al., “Chemical Characterization and Classification of Pollen”, Analytical Chemistry, vol. 80, p. 9551, 2008 Non-patent literature 9Q. Wang et al., “Diurnal and Nocturnal Behaviour of Airborne Cryptomeria japonica Pollen Grains and the Allergenic Species in Urban Atmosphere of Saitama, Japan”, Asian Journal of Atmospheric Environment, vol.7, p.65, 2013 Non-patent literature 10 Mitsumoto et al., “Development of a novel real-time pollen-sorting counter using species-specific pollen autofluorescence”, Aerobiologia, vol. 26, p. 99, 2010 <Identification and quantification of allergens using the semiconductor sensor of the present invention> Therefore, one of the objectives of this invention is to continuously identify or quantify allergens such as pollen, mold, and mites suspended in the air. To this end, this invention uses two or more semiconductor sensors 11 with different metal oxide compositions to detect allergens. The semiconductor sensors 11 react to low concentrations of VOCs (Volatile Organic Compounds) contained in unpressurized or liquefied gases, causing a change in conductivity (the reciprocal of resistance). Therefore, the detection device can be miniaturized for use in homes and other similar locations. Furthermore, the semiconductor sensors 11 measure VOCs in a cycle of approximately several minutes and require no cleaning, thus enabling continuous allergen measurement. Measurement can be performed automatically without human intervention. Moreover, in this invention, since allergens are identified and quantified using a model generated by learning the signal patterns of the semiconductor sensors 11, pollen, mold, and mites can be identified and quantified with high specificity.

[0053] <About Terminology> A target space is a space with enough air to sustain human life, such as an indoor living space. A target space can also be outdoors.

[0054] The gaseous components refer to nitrogen, oxygen, water vapor, and trace gases that make up air, and trace gases include VOCs. Although pollen and the like are not gases, in this embodiment, allergens are sometimes included as one of the gaseous components. This is because VOCs contain components derived from allergens.

[0055] Allergens are substances suspended in a target space that may trigger allergic reactions. Given that allergens contain VOCs, semiconductor sensors detect the VOCs.

[0056] Mites include mite corpses, feces, molted skins, and other mite-derived materials. Molds can also include yeasts and mushrooms. Furthermore, the detected molds are primarily mycelia, but spores may also be detected.

[0057] <Overview of Allergen Detection Methods> Figure 1 This is a diagram illustrating an example of the system configuration and allergen detection method of an allergen detection system 100. The environmental device 10, sensor unit 8, user terminal 70, and information processing device 60 are communicatively connected via network N.

[0058] Sensor units 8 are preferably installed in each room of the building where resident 9 resides. Alternatively, only one sensor unit 8 may be installed in the building. Sensor unit 8 preferably has an air intake device, in which multiple semiconductor sensors 11 are built into the airflow path. These semiconductor sensors 11 react to allergens in the target space 7 where resident 9 resides, causing a change in their conductivity.

[0059] Environmental equipment 10 is a device that controls the environment related to air quality, such as an air conditioner, ventilation device, or air purifier. Environmental equipment 10 can have multiple functions, such as an air conditioner, ventilation device, or air purifier, and each function of environmental equipment 10 can also exist independently.

[0060] User terminal 70 is a terminal device used by resident 9. A web browser or native application runs on user terminal 70, which receives information from information processing device 60 via network N for display on a screen. Resident 9 confirms the type and amount of allergens currently detected and operates environmental device 10 to reduce the amount of allergens or prevent their increase. User terminal 70 can be carried by resident 9 and does not need to be located in the same space as the space where sensor unit 8 is located.

[0061] The information processing device 60 (an example of a detection device) is a server device that performs various information processing in this invention. The information processing device 60 generates an allergen quantification model and an allergen identification model, which will be described later, and identifies and quantifies allergens by inputting the signal pattern detected by the semiconductor sensor 11 into the allergen quantification model and the allergen identification model.

[0062] Assuming the learning phase for generating the allergen quantification model and the allergen identification model has been completed, the process for identifying and quantifying allergens is described below. As will be described later, the allergen quantification model and the allergen identification model output the type and quantity of allergens based on the signal pattern detected by the semiconductor sensor 11.

[0063] (1) The sensor unit 8 sends the signal data (instantaneous value) detected by the semiconductor sensor 11 to the information processing device 60.

[0064] (2) The information processing device 60 accumulates signal data. When a signal pattern with a waveform quantity is obtained, it is input into the generated allergen quantitative model and allergen identification model, so as to continuously and automatically identify and quantify allergens.

[0065] (3) The information processing device 60 sends the type and amount of allergens obtained from the signal pattern to the user terminal 70. Based on the amount of allergens, the information processing device 60 sends a message suggesting that the environmental equipment 10 should be operated to reduce or prevent the amount of allergens from increasing.

[0066] (4) The user terminal 70 displays the type and amount of allergens, as well as suggestions on how the environmental device 10 should be operated. By inputting the operation of the environmental device 10, the user terminal 70 sends the control request of the environmental device 10 to the information processing device 60.

[0067] (5) The information processing device 60 converts the control request of the environmental device 10 into control information for the environmental device 10 and sends it to the environmental device 10. Thus, the environmental device 10 can be controlled to reduce the amount of allergens or prevent them from increasing.

[0068] Thus, by using a semiconductor sensor 11, the present invention can continuously and automatically identify and quantify allergens such as pollen, mold, and mites without increasing the size of the environmental equipment 10.

[0069] <<Examples of System Composition>> Sensor unit 8 is not independent, such as Figure 2 As shown, the semiconductor sensor 11 can also be built into the indoor unit 10b. Figure 2 This describes a modified example of the system configuration of the allergen detection system 100. The allergen detection system 100 mainly includes an outdoor unit 10a as a heat source unit, one or more indoor units 10b as utilization units, and a remote control device (hereinafter referred to as "remote control 15") as an input device for inputting commands related to various settings.

[0070] Outdoor unit 10a and indoor unit 10b are referred to as air conditioners. Outdoor unit 10a and indoor unit 10b are connected via refrigerant connecting pipes (gas connecting pipes GP), forming a refrigerant circuit. Furthermore, in the allergen detection system 100, multiple communication networks (network NW1, network NW2) are constructed to function as signal transmission paths between the indoor and outdoor units. Network NW2 can be wired or wireless.

[0071] The remote control 15 is a user interface that accepts settings such as temperature and humidity. The information processing device 60 can function in conjunction with... Figure 1 same.

[0072] (1) An embedded semiconductor sensor 11 is built into the air inlet of the indoor unit 10b and other locations. The indoor unit 10b sends signal data to the outdoor unit 10a.

[0073] (2) The outdoor unit 10a sends signal data to the information processing device 60.

[0074] (3) The information processing device 60 accumulates signal data. When a signal pattern with a waveform quantity is obtained, it is input into the generated allergen quantitative model and allergen identification model, so as to continuously and automatically identify and quantify allergens.

[0075] (4) The information processing device 60 sends the type and amount of allergens obtained from the signal pattern to the remote control 15 via the outdoor unit 10a and the indoor unit 10b.

[0076] (5) The remote control 15 displays the type and amount of allergens. Based on the amount of allergens, the remote control 15 displays suggestions for operating the environmental equipment 10 to reduce or prevent the allergen level from increasing. The user inputs the operation instructions for running the environmental equipment 10, and the remote control 15 sends the setting information to the indoor unit 10b.

[0077] (6) The indoor unit 10b controls itself according to the setting information. As a result, the environmental equipment 10 can be controlled to reduce the amount of allergens or prevent them from increasing.

[0078] <System Composition of an Allergen Detection System> Next, refer to Figure 3 Explain the system composition of the allergen detection system 100. Figure 3 This is a diagram illustrating an example of the system configuration of an allergen detection system 100.

[0079] The allergen detection system 100 provides various IoT-enabled services to everyone from managers to general users by enabling communication between various environmental devices 10, such as air conditioners, ventilation systems, and air purifiers, and a cloud-based information processing device 60 via a network N. The edge device 80, environmental devices 10, sensor switches 53, and user terminals 70 are located on the customer side, while the information processing device 60 is located in a data center or the cloud, such as the internet. Furthermore, since the edge device 80 is only a device for centrally managing the environmental devices 10 and sensor switches 53, it may not be necessary to have an edge device 80.

[0080] As described above, the environmental device 10 is preferably a device that comes into contact with the air in the living space, such as an air conditioning unit, a ventilation device, or a purifier. The living space is not limited to indoors but can also be outdoors. Generally, the edge device 80 is sometimes connected to anti-theft devices, heat source devices, fire alarms, AHUs (air handling units), electricity meters, lighting, etc.

[0081] The environmental device 10 and the sensor switch class 53 are controlled by the edge device 80. In other words, the edge device 80 applies the required operations to the environmental device 10 and the sensor switch class 53 to achieve the desired purpose. The content of the control varies depending on the type of environmental device 10 and sensor switch class 53. For example, if the environmental device 10 is an air conditioner, it may include all controls related to the functions of the air conditioner, such as the typically settable heating / cooling mode, set temperature, airflow, humidity, and airflow direction. The sensor switch class 53 may include a semiconductor sensor 11.

[0082] Environmental device 10 collects corresponding operating data and mainly sends it to edge device 80 periodically. The periodicity is, for example, once per minute, once per 10 minutes, once per 60 minutes, etc., and can also be set by the user or information processing device 60. The operating data varies depending on the environmental device 10. For example, in the case of an air conditioner, there are various data such as high and low refrigerant pressure, refrigerant temperature, fan speed, and CPU temperature of the microcomputer.

[0083] Edge device 80 is a controller for controlling environmental equipment 10 and sensor switches 53. In the absence of edge device 80, information processing device 60 controls environmental equipment 10 and sensor switches 53.

[0084] The information processing device 60 may be one or more server devices. Figure 3 The diagram shows one information processing device 60, but multiple information processing devices 60 can be configured according to their functions. Furthermore, the functions of the information processing device 60 can also be centralized in a single server device. Additionally, multiple devices with the same functions can be prepared as the information processing device 60, allowing multiple information processing devices 60 to communicate and process data simultaneously, like a server cluster.

[0085] The information processing device 60 inputs signal patterns into the allergen quantification model and the allergen identification model to identify and quantify allergens, and provides the information to the user terminal 70, etc. The information processing device 60 can identify and quantify allergens not only for individual residences or buildings, but also by region, and can share this information through public broadcasting, etc.

[0086] in addition, Figure 3 Although not explicitly described, an information processing device that generates quantitative allergen models and allergen identification models, different from the information processing device 60, may exist. In this case, the quantitative allergen models and allergen identification models generated by other information processing devices are imported into the information processing device 60. In this invention, for ease of explanation, it is assumed that the quantitative allergen models and allergen identification models are generated by the information processing device 60.

[0087] The information processing device 60 can also function as a web server. The web server responds to requests from client software (web clients) such as web browsers operated by the user, providing the client with screen information described in HTML files, XML, CSS files, JavaScript (registered trademark), etc. Applications that use web mechanisms in this way are called web applications.

[0088] Furthermore, the information processing device 60 preferably supports cloud computing. Cloud computing refers to a mode of use that utilizes resources on the network without being aware of specific hardware resources.

[0089] User terminal 70 is a client terminal that displays various screens provided by information processing device 60. User terminal 70 can be used by an administrator or by a general user (resident 9 in this invention). When environmental equipment 10 is located in a typical home, the administrator can be a family member of the resident, or the resident 9 can also act as the administrator. When environmental equipment 10 is located in a building managed by a company, the administrator can be, for example, the facility manager.

[0090] User terminal 70 can be, for example, a PC (Personal Computer), smartphone, tablet, PDA (Personal Digital Assistant), wearable PC (sunglasses type, watch type, etc.). Here, it only needs to have communication capabilities and be able to run a web browser. Alternatively, the user terminal 70 can run a native application specific to the allergen detection system 100 instead of a web browser.

[0091] <Hardware Composition of Information Processing Devices> Reference Figure 4 This describes the hardware configuration of the information processing device 60. Figure 4 This diagram illustrates an example of the hardware configuration of the information processing device 60. (For example...) Figure 4 As shown, the information processing device 60 includes a processor 221, a memory 222, an auxiliary storage device 223, an I / F (Interface) device 224, a communication device 225, and a drive device 226. Furthermore, the various hardware components of the information processing device 60 are interconnected via a bus 207.

[0092] The processor 221 includes various computing devices such as a CPU (Central Processing Unit). The processor 221 loads various programs into the memory 222 and executes them. The processor 221 controls the entire information processing device 60.

[0093] The memory 222 includes main storage devices such as ROM (Read Only Memory) and RAM (Random Access Memory). The processor 221 and the memory 222 together form a computer, and the processor 221 executes various programs read from the memory 222.

[0094] Auxiliary storage device 223 stores various programs and various data used by the processor 221 when the various programs are executed.

[0095] I / F device 224 is a connection device that connects display device 230, operation device 240 (an example of an external device), and information processing device 60. Display device 230 displays the internal status of information processing device 60. Operation device 240 is used when the administrator of information processing device 60 inputs various instructions to information processing device 60.

[0096] The communication device 225 is a communication device used to communicate with the edge device 80 and the user terminal 70 via the network N.

[0097] The drive unit 226 is a device used to set the recording medium 250. The recording medium 250 mentioned here includes media that record information optically, electrically, or magnetically, such as CD-ROMs, floppy disks, and magneto-optical disks. Furthermore, the recording medium 250 may also include semiconductor memories that record information electrically, such as ROMs and flash memory.

[0098] Furthermore, various programs installed in the auxiliary storage device 223 can be installed, for example, by placing the distributed recording medium 250 in the drive device 226 and having the drive device 226 read the various programs recorded in the recording medium 250. Alternatively, various programs installed in the auxiliary storage device 223 can also be installed by downloading them from the network N via the communication device 225.

[0099] <About Features> Next, refer to Figure 5 This section details the functional components of the allergen detection system 100. Figure 5 This is an example of a functional block diagram that explains the functions of the information processing device 60 and the environmental device 10 (or sensor unit 8) by dividing them into individual modules. Additionally, Figure 5 For ease of explanation, the functions of the learning phase and the reasoning phase are shown simultaneously.

[0100] The environmental device 10 or sensor unit 8 includes multiple semiconductor sensors 11, a sensor control unit 12, and a transmitter 13. Multiple semiconductor sensors 11 are arranged at the air inlet, etc., of the environmental device 10 or sensor unit 8. Since the semiconductor material used for the gas semiconductor sensors 11 is frequently used under conditions of high temperature in the air, thermally and chemically stable metal oxides are used. The conductivity (i.e., resistance value) of the semiconductor sensors 11 changes according to the adsorption and desorption of VOCs in the air. Multiple semiconductor sensors 11 are configured to be suitable for detecting different VOCs, and the composition of the metal oxide of each semiconductor sensor 11 varies depending on the semiconductor sensor 11. The multiple semiconductor sensors 11 have different characteristics regarding how their conductivity, etc., changes for the same gas composition.

[0101] An example of the composition of the metal oxide in the semiconductor sensor 11 is shown below. Examples of metal oxides include zinc oxide (ZnO), tin oxide (SnO2), iron oxide (Fe2O3), tungsten oxide (WO3), indium oxide (In2O3), etc., but are not limited to these.

[0102] The sensor control unit 12 controls these semiconductor sensors 11. For example, the sensor control unit 12 generates a driving voltage and supplies it to the semiconductor sensors 11, and performs the following processes: issuing a VOC detection instruction to the semiconductor sensors 11, acquiring the signal from the semiconductor sensors 11, noise removal, filtering, and outlier removal, etc.

[0103] The transmitting unit 13 transmits the signal data detected by the semiconductor sensor 11 to the information processing device 60 in real time or after accumulating a certain amount via the network N. Since the information processing device 60 receives the signal data in a time sequence, it can regenerate the signal patterns of each semiconductor sensor 11.

[0104] The information processing device 60 includes an acquisition unit 21, a data unit 22, a model generation unit 23, a parsing unit 24, an output unit 25, and an environmental device control unit 26. These units of the information processing device 60 are called the control unit 20, which is a function or means implemented by the processor 221 of the information processing device 60 by executing instructions of a program expanded in the memory 222.

[0105] The acquisition unit 21 acquires signal data detected by the semiconductor sensor 11 from the environmental device 10. The acquisition unit 21 may request signal data from the environmental device 10 and receive the signal data in response, or it may receive the data directly without requesting it. Furthermore, the acquisition unit 21 may accumulate the signal data in the data unit 22. If additional training data is available, the model generation unit 23 may use the commercially available environmental device 10 or sensor unit 8 to reconstruct the model based on the signal data accumulated in the data unit 22.

[0106] During the learning phase, the data unit 22 stores training data. In this embodiment, the training data includes, for example, "signal patterns and types of allergens" and "signal patterns and amounts of allergens." Details of the training data will be provided later. Figure 13 The text is described in the middle.

[0107] The model generation unit 23 learns and trains data to generate an allergen identification model from "signal patterns and types of allergens" and an allergen quantitative model from "signal patterns and quantities of allergens". The analysis unit 24 contains the generated allergen identification model and allergen quantitative model.

[0108] During the inference phase, the analysis unit 24 accumulates the signal data detected in real time by the semiconductor sensor 11 to generate a signal pattern, which is then input into the allergen identification model and the allergen quantification model. The allergen identification model outputs the quantity of allergens, and the allergen quantification model outputs the type of allergen. The analysis unit 24 transmits the quantity and type of allergens to the output unit 25 and the environmental device control unit 26.

[0109] The output unit 25 sends the type or quantity of detected allergens to the user terminal 70 or the remote control 15. The output unit 25 can generate screen information such as HTML that can be displayed through a web application or web page. Since the user terminal 70 or the remote control 15 displays the type or quantity of allergens, the user can operate the environmental device 10, for example, to activate the air purification function.

[0110] When the amount of allergen exceeds a threshold, or the rate of increase of allergen exceeds a threshold, the environmental equipment control unit 26 sends control information to the environmental equipment 10 to activate the air purification function of the environmental equipment 10. Thus, the environmental equipment 10 can begin operation to reduce the amount of allergen or prevent its increase.

[0111] exist Figure 5 In this device, the environmental equipment 10 has a semiconductor sensor 11 and a sensor control unit 12, while the information processing device 60 has a control unit 20. However, the semiconductor sensor 11 and the control unit 20 can also be configured in a single device. For example, in household air conditioners, air purifiers, and ventilation systems, a single device can include the semiconductor sensor 11, the sensor control unit 12, and the control unit 20. Furthermore, in the case of an air conditioner with indoor and outdoor units, the indoor unit can have the semiconductor sensor 11 and the sensor control unit 12, and the outdoor unit can have the control unit 20; alternatively, either the indoor or outdoor unit can have the semiconductor sensor 11, the sensor control unit 12, and the control unit 20. In the case of a ventilation-type air conditioner, the semiconductor sensor 11 is preferably capable of detecting allergens on both the indoor and outdoor sides. Furthermore, if the semiconductor sensor 11 is shared on both the indoor and outdoor sides, cost increases can be suppressed.

[0112] <Methods for the identification and quantification of allergens> Reference Figure 6 This illustrates the inconveniences of existing GC-MS (gas chromatography-mass spectrometry). Figure 6 The TIC (Total Ion Chromatogram) chromatograms of volatile compounds obtained from 10 mg of pollen from cedar, cypress, red pine, and chestnut trees are shown. Figure 6As shown, in the method of identifying pollen species from VOCs using GC-MS, the species are determined based on the difference in retention time from sample injection to the appearance of the peak value. Therefore, the maximum retention time required for the VOC peak is not suitable for continuous measurement. Furthermore, the need for devices for heating and discharging the heated gas makes it difficult to reduce equipment costs or achieve miniaturization.

[0113] Therefore, as Figure 7 As shown, in this embodiment, allergens are detected by semiconductor sensor 11. Figure 7 This is a schematic diagram illustrating the detection principle of the semiconductor sensor 11. (For example...) Figure 7 As shown, the semiconductor sensor 11 detects gas components by utilizing the reaction between oxygen adsorbed on the surface of a metal oxide and reducing gases (CH3, H2, CO, CO2, H2O, etc.). Specifically, the reducing gases correspond to VOCs contained in allergens, and the conductivity changes through the adsorption and desorption of these reducing gases. Although there are multiple semiconductor sensors 11, each simultaneously outputs signals related to multiple gas components.

[0114] Figure 8 This represents an example of a signal pattern 110 detected by the semiconductor sensor 11. Figure 8 The vertical axis represents resistance, and the horizontal axis represents time. The semiconductor sensor 11 uses the unadsorbed VOC state (initial state) as the reference value RD for the resistance. The resistance decreases due to the depletion of oxygen in the sensor's electron depletion layer (adsorption of reducing gases), and is restored by the supply of oxygen from the air, thus repeating the cycle. The length of one cycle varies depending on the semiconductor sensor 11 and the type of VOC, for example, several minutes. The semiconductor sensor 11 can measure a sufficient number of resistance values ​​to reproduce the waveform in one cycle.

[0115] Thus, since the resistance value can be continuously measured, VOCs originating from allergens can be continuously measured. Furthermore, the semiconductor sensor 11 exhibits high sensitivity at low concentrations, enabling the detection of VOCs originating from allergens even without pressurizing or liquefying the air. In addition, the semiconductor sensor 11 is characterized by high cost reduction during mass production and good responsiveness. Therefore, the environmental device 10 can be miniaturized. By using the semiconductor sensor 11, compared to GC-MS, a cheaper and smaller environmental device 10 and sensor unit 8 can be achieved.

[0116] in addition, Figure 8 The signal pattern showing the resistance value is displayed, but the signal pattern can also be current, voltage, conductivity, or resistance (impedance). The control unit 20 identifies or quantifies allergens based on the signal pattern of current, voltage, conductivity, or resistance (impedance).

[0117] <An example of an allergen> As described above, when VOCs originating from allergens are adsorbed onto the semiconductor sensor 11, the resistance value varies depending on the material of the semiconductor sensor 11 and the type of allergen (VOC). That is, each semiconductor sensor 11 detects a different signal pattern depending on the type of allergen. Furthermore, the signal pattern also varies depending on the amount of allergen. Therefore, by performing machine learning on the signal patterns of multiple semiconductor sensors 11, it is possible to identify and quantify allergens.

[0118] Pollen, mites, and mold are known examples of allergens. Furthermore, in the case of pollen, the information processing device 60 can identify pollen from cedar, cypress, rice, miscanthus, ryegrass, timothy, yellow cogon grass, orchardgrass, red pine, black pine, artemisia, Canada goldenrod, ragweed, hops, birch, and ginkgo. The information processing device 60 of this embodiment can also identify the type of pollen, as well as the types of mites and molds.

[0119] <Allergen Identification and Quantification Using Neural Networks> In this embodiment, an allergen identification model and an allergen quantification model are generated using machine learning. Machine learning is a technology that enables computers to acquire human-like learning abilities. It refers to the technology where computers autonomously generate algorithms for data identification and other judgments from pre-ingested training data and apply them to predict new data. The learning methods used for machine learning can be any of the following: supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, and deep learning, or a combination of these methods. The learning method used for machine learning is not limited. As an example, an allergen identification model and an allergen quantification model using neural networks are illustrated.

[0120] Reference Figure 9 This is an example illustrating the method for creating a quantitative allergen model. Figure 9 This represents an example of a neural network used by the model generation unit 23 for learning. Figure 9 The neural network is a regression (continuous value prediction) neural network that takes multiple data inputs to the input layer 41 and outputs a single output value.

[0121] Figure 9This is a neural network with L fully connected layers from input layer 41 to output layer 43 (assuming the number of weighted intermediate layers 42 and output layer 43 = 2 layers). A deep neural network is called a DNN (Deep Neural Network). The layers between input layer 41 and output layer 43 are called intermediate layers 42 (or hidden layers). The number of intermediate layers 42 and the number of nodes 35a-35c in each layer are simplified for illustrative purposes and are only one example. The number of nodes 35a in input layer 41 can be the number of elements in the vector used as input data. In the case of continuous value prediction, the number of nodes 35c in output layer 43 is usually 1. Therefore, an allergen quantification model is created according to the number of allergen types.

[0122] exist Figure 9 In the neural network, one node 35b of the intermediate layer 42 is connected to all nodes 35a of the input layer 41, and one node 35c of the output layer 43 is connected to all nodes 35b of the intermediate layer 42 (fully connected). The product of the output z of node 35a of the input layer 41 and the connection weight w is input to node 35b of the intermediate layer 42, and the product of the output z of node 35b of the intermediate layer 42 and the connection weight w is input to node 35c of the output layer 43. Equation (1) represents the method for calculating the output signal of node 35b.

[0123]

Number 1

[0124]

Number 2

[0125] Node 35c of output layer 43 uses the activation function typically used in output layers. The activation function for output layer 43 of a regression model is generally the identity function (y=x).

[0126] Figure 10 This is a diagram illustrating the learning process of a neural network. The neural network processes the data input to input layer 41 and outputs a value from output layer 43. For example, given input data such as "0.1" and "0.3", node 35c of output layer 43 outputs "5". Assume that the training data pre-sets a training value of "7" corresponding to "0.1" and "0.3". The training data for the allergen quantification model is, for example, the quantity of allergens (per unit area or volume, and the number per unit time) counted using a method described in the aforementioned non-patent literature.

[0127] During the learning phase, a loss function is used to evaluate the error between the training data and the output value, and the weights w and b are adjusted to make the output value closer to the training data. The loss function for the regression model can be a function that calculates the squared error. The value of the loss function is propagated to the nodes of the input layer 41 through a calculation method called error backpropagation. During the propagation process, the weights w and b between nodes are learned.

[0128] Next, refer to Figure 11 This illustrates the neural network used to identify allergens. Figure 11 This illustrates an example of a neural network used by the model generation unit 23 for learning an allergen identification model. Additionally, Figure 11 The explanation mainly focuses on the relationship with Figure 9 The difference. Figure 11 The neural network is a classification neural network that takes multiple data inputs to the input layer 41 and outputs the probability of each data belonging to each allergen by the output layer 43.

[0129] The configuration of input layer 41 and intermediate layer 42 can be similar to... Figure 9 , Figure 10 The same applies. In the case of classification, the softmax function is typically used as the activation function for node 35c of output layer 43. Through the softmax function, the inputs to each node 35c of output layer 43 are converted into probabilities of 0 to 1.

[0130] For example, given input data like "0.1" and "0.3", nodes 35 of output layer 43 output "0.15", "0.80", and "0.05". Assume the training data pre-sets "0 1 0" corresponding to "0.1" and "0.3". The training data for classification can be a one-hot vector. "0 1 0" corresponds to "pollen, mites, and mold" respectively. When the input data is pollen detection, the training data is "1 0 0"; when it's mite detection, the training data is "0 1 0"; and when it's mold detection, the training data is "0 0 1". Therefore, Figure 11 The “01 0” indicates that the training data is mites.

[0131] During the learning phase, the error between the training data and the output value is evaluated using a loss function, and the weights w and b are adjusted to make the output value closer to the training data. The loss function of the classification model can be a function that calculates the cross-entropy. The value of the loss function is propagated to the nodes of the input layer 41 through a calculation method called error backpropagation. During the propagation process, the weights w and b between nodes are learned.

[0132] <Preprocessing of waveform data> In this embodiment, the input data input to the input layer 41 is the waveform data detected by the semiconductor sensor 11. Figure 12 This represents an example of a signal pattern 110 detected by a semiconductor sensor 11. As described above, the signal pattern 110 repeatedly cycles of decreasing and recovering. It is believed that the waveform of one cycle contains characteristics of how the semiconductor sensor 11 reacts to an allergen. Therefore, the analysis unit 24 uses this one cycle (one signal pattern 110) as input data for the identification and quantification of the allergen. However, half a cycle of the signal pattern 110 can also be used. This allows for the identification and quantification of the allergen in a shorter time.

[0133] like Figure 8 As shown, the analysis unit 24 extracts the signal pattern 110 between maxima from the continuously acquired signal waveform. It can also extract the signal pattern between minima. Alternatively, it can extract the signal pattern between maxima and minima. Then, as... Figure 12 As shown, the analysis unit 24 samples P values ​​from the signal pattern 110 using the sampling window 120. The sampling window 120 samples a range (time length) of signal data from the signal pattern 110. For example, one sampling window 120 acquires P resistance values. In order to acquire the waveform of the signal pattern 110 without omission, the analysis unit 24 shifts the sampling window 120 relative to time overlap. Let the number of sampling windows 120 covering one signal pattern 110 be Q, then the amount of data N input to one signal pattern 110 is as follows.

[0134] N = P × Q The training data is a combination of the input data and the training data. As mentioned above, the training data is the quantity of allergens in regression and a one-hot vector representing the types of allergens in classification.

[0135] Figure 13 The training data is represented in a pattern. Figure 13 The row shown (one record) represents one training data point. Training data consists of input data and the quantity of allergens, or input data and the types of allergens. Input data is preferably normalized to 0-1 and standardized. The person in charge needs to prepare a large amount of input data, including the quantity and types of allergens, in advance.

[0136] The model generation unit 23 obtains training data of batch size from the data unit 22, repeats the process for each batch for a specified number of epochs, thereby generating an allergen identification model and an allergen quantification model.

[0137] Let N be the amount of data input for a signal pattern 110, and M be the number of semiconductor sensors 11. Then the number of nodes 35a in the input layer 41 is N×M.

[0138] <<Methods for treating waveform data as image data>> exist Figure 12 , Figure 13 The text describes using the values ​​that constitute the signal pattern 110 as input data, but the input data can also be an image.

[0139] Figure 14A , Figure 14B This diagram illustrates the conversion of signal pattern 110 into image data. As a method for converting to image data, the information processing device 60 performs preprocessing, such as... Figure 14A As shown, the timing signal is plotted as a graph, etc. The information processing device 60 preferably performs interpolation between data points. Furthermore, as... Figure 14B As shown, the information processing device 60 extracts the circumscribed rectangle of the signal pattern 110 between two maxima (or two minima). The analysis unit 24 preferably shapes the image data into square image data with a defined number of pixels in both directions. Furthermore, in this embodiment, since the colors of the image data do not contain information about the quantity or type of allergens, the analysis unit 24 binarizes the image data.

[0140] Neural networks that take image data as input can perform convolution and pooling. Convolutional layers function similarly to filters, extracting various features and patterns within an image, such as line shapes. Furthermore, pooling reduces the amount of data with minimal information loss and achieves a degree of invariance to positional biases.

[0141] Figure 15 This represents an example of a model structure when using a neural network corresponding to image data. Figure 15 The neural network 30 is a neural network called CNN (Convolutional Neural Network). The image data 31 of the signal pattern 110 of each semiconductor sensor 11 is input into the neural network 30, and the neural network 30 outputs the type 32 or quantity 33 of the allergen. Figure 15 For ease of explanation, one neural network 30 outputs the type 32 or quantity 33 of allergens, but separate neural networks 30 can be constructed to output the type 32 or quantity 33 of allergens. Furthermore, regarding the quantity 33 of allergens, a separate neural network 30 can be constructed for each allergen.

[0142] Neural network 30 has convolutional layers 1 and 2, pooling layers 1 and 2, activation functions 1-3, and fully connected layers 1 and 2. Convolutional layers 1 and 2 convert the numerical data into a single value by calculating the sum of the element-wise products of a grid of numerical data called a kernel (or filter) and a portion of the image (called a window) of the same size as the kernel. Convolutional layers 1 and 2 perform this conversion by gradually moving the window, thus transforming it into small grid-like numerical data (i.e., tensors).

[0143] Pooling layers 1 and 2 process data to generate a single value from the numerical data in the window. Examples include max pooling of the maximum value in the selected window and average pooling of the average value in the selected window. Convolutional layers 1 and 2 extract features from the captured image, while pooling layers 1 and 2 blur the object's position. Activation functions 1-3 are functions that perform non-linear transformations (activations) on the input (e.g., ReLU, tanh, Sigmoid). Fully connected layer 1 is equivalent to input layer 41, which summarizes the output of activation function 2, and fully connected layer 2 is equivalent to output layer 43. The summed values ​​from fully connected layer 1 are passed to fully connected layer 2 via activation function 3. Fully connected layer 2 has one (quantitative) or multiple (identification) output nodes, which output the type or quantity of allergens.

[0144] Furthermore, the number of convolutional layers 1 and 2, pooling layers 1 and 2, activation functions 1 to 3, and fully connected layers 1 and 2 shown in the diagram are only one example, and the processing order is also only one example.

[0145] Figure 15 The training data for the neural network in case 30 is... Figure 13 The N×M input data are replaced with M image data. Assuming the resolution of the image data converted from one signal pattern 110 is, for example, 100 pixels × 100 pixels, and the number of semiconductor sensors 11 is M, then the number of nodes 35 in the input layer 41 is 100 × 100 × M. Figure 15 In the case of neural network 30, the content of the filter and the coefficients between nodes of fully connected layers 1 and 2 are adjusted by learning.

[0146] <<Other Machine Learning Methods>> Besides neural networks (perceptrons, deep learning), machine learning methods also include support vector machines (SVMs), logistic regression, decision trees, random forests, and others, not limited to the methods described in this implementation. For example, SVMs are mainly used for classification models. A SVM is a method that maximizes the margin hyperplane between positive and negative example vectors in the training data. Kernel methods that map data to a high-dimensional feature space can also classify non-linear data.

[0147] Boosting decision trees are a method that independently learns multiple weak classifiers, such as decision trees, and then integrates their predictions using methods like majority voting, outputting the overall prediction (of the strong classifier). While boosting decision trees are used for regression, boosting regression trees, which employ the same method, can also be applied to regression.

[0148] In addition, there are various methods suitable for regression and classification, and the methods for generating the allergen identification model and allergen quantification model in this implementation can be any method.

[0149] <Examples of allergen identification and quantification> Figure 16 This is the allergen detection result screen 300 displayed on the user terminal 70 or remote control 15. The horizontal axis of the allergen detection result screen 300 is the current time, and the vertical axis is the amount of allergen. The change in the amount of allergen over time is represented by line graphs 301 to 303. Figure 16 As an example, pollen, mold, and dust mites are shown as allergens, but the allergen test result screen 300 can also display the number of other types of allergens. The resident 9 can operate the user terminal 70 or the remote control 15 to indicate which allergen to display. In the allergen test result screen 300, if the user specifies pollen as the allergen, the number of different types of pollen can be displayed; if dust mites are specified, the number of different types of dust mites can be displayed; if mold is specified, the number of different types of mold can be displayed. Furthermore, the resident 9 can also specify the time range for displaying line graphs 301-303. The allergen test result screen 300 can also display the maximum and minimum values ​​of each allergen over a certain period of time.

[0150] exist Figure 16 In the allergen test results screen 300, pollen counts increased sharply after 12:45. In this situation, output unit 25 preferably issues a warning to resident 9. Figure 16The system displays a message 304 stating "Pollen levels are increasing. Do you want to turn on the air purification function?", a "Yes" button 305, and a "No" button 306. When the resident 9 presses the "Yes" button 305, the information processing device 60 sends control information, such as activating the air purification function, to the environmental device 10. Alternatively, the environmental device control unit 26 can appropriately control the environmental device 10 based on the allergen detection results, without requiring the output unit 25 to query the user.

[0151] Furthermore, the output unit 25 can also highlight (flashing, thick lines, etc.) a line graph showing an increasing trend. Additionally, if the amount of allergen is reduced through the control of the environmental device 10, the output unit 25 can also display this on the allergen detection result screen 300. The resident 9 can confirm that the allergen has been reduced through the control of the environmental device 10.

[0152] Because the information processing device 60 of this embodiment can continuously detect allergens, it can alert the resident 9 in real time and control the environmental equipment 10 without delay. It can also detect in real time when allergens are reduced through the control of the environmental equipment 10, and inform the resident 9 of the effectiveness of the air purification function.

[0153] Furthermore, when the environmental device 10 is installed outdoors, or when it can inhale outdoor air, the information processing device 60 can provide a map showing the distribution of pollen based on the detection results of allergens (mainly pollen) and the location information of the environmental device 10. The location information of the environmental device 10 can be determined, for example, by IP address, or by the environmental device 10 sending a postal code or address to the information processing device 60.

[0154] Figure 17 The user terminal 70 displays the outdoor pollen detection results screen 320. The information processing device 60 can receive pollen levels from various locations from the environmental equipment 10. The information processing device 60 divides the map into a grid and calculates the average pollen level within each grid. The information processing device 60 uses color coding to differentiate the grids based on pollen levels, thus visually displaying the pollen levels at each location. For example, residents 9 can check pollen dispersal levels when traveling to distant places. Additionally, the information processing device 60 can also provide pollen levels for various locations to a weather information service.

[0155] <Overall Processing Flow> Next, refer to Figure 18 This describes the overall processing flow of the allergen detection system 100. Figure 18 This is a sequence diagram illustrating how the allergen detection system 100 displays the allergen detection results and controls the environmental equipment 10's processing.

[0156] S101: The acquisition unit 21 of the information processing device 60 repeatedly acquires signal data measured by the semiconductor sensor 11 from the environmental device 10 or the sensor unit 8, for example, at regular intervals or at regular intervals.

[0157] S102: The analysis unit 24 of the information processing device 60 extracts a signal pattern 110 of a waveform quantity from the time-series signal data, and inputs the signal pattern 110 into the allergen identification model and the allergen quantification model to identify and quantify the allergen.

[0158] S103: Output unit 25 compares the measured amount of each allergen with the threshold to determine whether to recommend the operation of environmental equipment 10. In addition, output unit 25 calculates the slope of the allergen amount relative to time based on the quantitative results of allergen measurements in the past, compares the slope with the threshold, and determines whether to recommend the operation of environmental equipment 10.

[0159] S104, S105: When resident 9 inputs an operation to make user terminal 70 display allergen test result screen 300, user terminal 70 sends a request for allergen test result screen 300 to information processing device 60.

[0160] S106: In response to a request from the user terminal 70, the output unit 25 of the information processing device 60 creates an allergen detection result screen 300. The output unit 25 sends the screen information of the allergen detection result screen 300 to the user terminal 70. Alternatively, the allergen detection result screen 300 can be displayed by the remote control 15 instead of the user terminal 70. In this case, even if the resident 9 does not perform the operation of displaying the allergen detection result screen 300, the allergen detection result screen 300 can still be displayed on the remote control 15.

[0161] S107: User terminal 70 receives the screen information of allergen test result screen 300 and displays allergen test result screen 300.

[0162] S108, S109: When resident 9 selects the "Yes" button 305 on the allergen detection result screen 300, user terminal 70 sends a control request for environmental device 10 to information processing device 60. If remote control 15, instead of user terminal 70, displays the allergen detection result screen 300, it can send setting information to environmental device 10.

[0163] S110: In response to a control request from the environmental device 10 of the user terminal 70, the environmental device control unit 26 of the information processing device 60 creates control information. The environmental device control unit 26 may create control information such as activating the air purification function or activating the environmental device 10 if it is stopped. The environmental device control unit 26 can activate the environmental device 10 and further activate the air purification function.

[0164] S111: The environmental equipment control unit 26 sends control information to the environmental equipment 10. The environmental equipment 10 converts the control information into setting information and controls itself, so that the amount of allergens can be reduced or prevented from increasing in the living space of the occupant 9.

[0165] In addition, Figure 18 In the process, when the resident 9 presses the "Yes" button 305, the environmental device 10 is controlled. However, based on the judgment result of step S103, the environmental device control unit 26 can directly control the environmental device 10 without asking the resident 9.

[0166] <Main Effects> As described above, in this invention, two or more semiconductor sensors 11 with different metal oxide compositions are used to detect VOCs contained in allergens. The semiconductor sensors 11 react to low concentrations of VOCs contained in unpressurized or liquefied gases, causing a change in conductivity. Therefore, the detection device can be miniaturized for use in homes and similar locations. Furthermore, the measurement cycle for VOCs by the semiconductor sensors 11 is approximately several minutes, allowing for continuous measurement of allergens. Measurement can be performed automatically without human intervention. Moreover, in this invention, since a model generated by machine learning of the signal patterns of the semiconductor sensors 11 identifies allergens, pollen, mold, and mites can be identified and quantified with high specificity.

[0167] <Other Application Examples> The above examples illustrate the optimal form for carrying out the present invention, but the present invention is not limited to these examples at all, and various modifications and substitutions can be made without departing from the spirit of the present invention.

[0168] For example, in this embodiment, a semiconductor sensor 11 is mainly used to detect allergens, but other gas sensors can also be used to detect allergens. For example, there are optical (NDIR) gas sensors that utilize the infrared absorption characteristics of gas molecules, fuel cell gas sensors that utilize the oxidation reaction generated by the detection electrode, or contact combustion gas sensors that utilize the combustion heat on the surface of the detection element.

[0169] also, Figure 5 The structural example is divided according to main functions to facilitate understanding of the processing of the information processing device 60. This invention is not limited to the method or name of the division of processing units. The processing of the information processing device 60 can also be divided into more processing units according to the processing content. Furthermore, it is also possible to divide the processing so that one processing unit contains more processing.

[0170] Furthermore, the group of devices described in the embodiments represents only one of a plurality of computing environments used to implement the embodiments disclosed in this specification. In one embodiment, the information processing apparatus 60 includes multiple computing devices such as a server cluster. The multiple computing devices are configured to communicate with each other via any type of communication link including a network or shared memory, and to implement the processing disclosed in this specification.

[0171] The functions of the present invention described above can be implemented not only by software processing executed by a program, but also by one or more processing circuits. Here, "processing circuit" in this specification includes processors that are programmed to execute the functions via software, such as processors implemented by electronic circuits, or devices such as ASICs (Application Specific Integrated Circuits), DSPs (Digital Signal Processors), FPGAs (Field Programmable Gate Arrays), and conventional circuit modules designed to perform the functions described above.

[0172] <Reasons for the effect> In the first aspect of the invention, "signal patterns detected by multiple semiconductor sensors exhibiting relatively different response characteristics to the same gas component are acquired, and allergens are identified or quantified based on the multiple signal patterns acquired from the multiple semiconductor sensors." Therefore, different signal patterns can be obtained for the same gas component, and these different signal patterns contain a wealth of information about the allergen. The detection device analyzes these patterns, thus enabling the identification or quantification of the allergen. The semiconductor sensors do not require cleaning, thus allowing for continuous identification or quantification. The analysis of the signal patterns requires no human intervention, enabling automatic identification or quantification.

[0173] In the second aspect of the present invention, "multiple semiconductor sensors change the signal pattern by adsorbing gas components," thus obtaining a signal pattern that reflects the adsorption of VOCs originating from allergens contained in the gas components.

[0174] In the third aspect of the present invention, "the signal pattern includes a pattern of current, voltage, conductivity or resistance that changes due to the adsorption of gas components on the semiconductor sensor", thus a signal pattern reflecting the adsorption of VOCs originating from allergens and representing changes in current, voltage, conductivity or resistance can be obtained.

[0175] In the fourth aspect of the present invention, "a pattern of current, voltage, conductivity or resistance that changes due to the detachment of gas components adsorbed on the semiconductor sensor" can be obtained, thus obtaining a signal pattern that reflects the detachment of VOCs originating from allergens and represents changes in current, voltage, conductivity or resistance.

[0176] In the fifth aspect of the present invention, "the gas component is introduced into the semiconductor sensor without pressurization or liquefaction," thus eliminating the need for pressurization or liquefaction equipment and providing a small detection device that can be used in places such as homes.

[0177] In the sixth aspect of the present invention, "by inputting the plurality of signal patterns into a model generated by machine learning the relationship between the plurality of signal patterns of the plurality of semiconductor sensors and the type or amount of allergens, allergens in the target space can be identified or quantified," thus enabling the identification and quantification of pollen, mold and mites with high specificity.

[0178] In the seventh aspect of the present invention, "the plurality of semiconductor sensors include two or more types of semiconductor sensors, the two or more types of semiconductor sensors having different metal oxide compositions", so the plurality of semiconductor sensors can display relatively different response characteristics for the same gas composition.

[0179] In the eighth aspect of the present invention, "the conductivity is repeated in a cycle of decreasing and recovering, and the control unit extracts a waveform between two maxima or between two minima of the conductivity as the signal pattern, and uses the signal pattern to repeatedly identify or quantify allergens." Therefore, if there is enough time for one cycle of the signal pattern, allergens in the target space can be identified or quantified. Furthermore, since the signal pattern is repeated cyclically, allergens can be measured continuously.

[0180] In the ninth aspect of the present invention, "when the quantitative amount of allergen exceeds a threshold, or when the slope of the allergen amount relative to time exceeds a threshold, the environmental equipment is controlled to reduce the amount of allergen or prevent it from increasing," so the environmental equipment can be automatically controlled according to the amount of allergen and the way it increases.

[0181] In the tenth aspect of the present invention, "the change in the amount of allergen relative to time is displayed according to the type of allergen", thus visualizing the amount of allergen and the way it increases to the user, prompting them to activate the environmental device.

[0182] In the eleventh aspect of the present invention, the "allergen" that is identified or quantified is pollen, mold or mites, thus pollen, mold or mites can be identified.

[0183] In the 12th aspect of the present invention, "multiple semiconductor sensors simultaneously output signals about multiple gas components" to be identified or quantified, thus eliminating the need to wait for separation time as in chromatography to identify or quantify allergens.

[0184] This application claims priority based on Japan Patent Application No. 2023-172258, filed on October 3, 2023, and incorporates the entire contents of Japan Patent Application No. 2023-172258 into this application.

[0185] [Attached image labels] 10 Environmental Equipment 11 Semiconductor Sensors 60 Information Processing Device 70 user terminals 100 Allergen Detection System

Claims

1. A detection device for identifying or quantifying allergens in a target space, wherein, Control Department Acquire signal patterns detected by multiple semiconductor sensors that exhibit relatively different response characteristics to the same gas composition. Based on the multiple signal patterns acquired from the multiple semiconductor sensors, allergens are identified or quantified. Output the results of identification or quantification.

2. The detection device according to claim 1, wherein the plurality of semiconductor sensors change the signal pattern according to the adsorption of gas components.

3. The detection device according to claim 2, wherein the signal pattern comprises a pattern of current, voltage, conductivity or resistance that varies on the semiconductor sensor according to the adsorption of gas components.

4. The detection apparatus according to claim 3, wherein the signal pattern comprises a pattern of current, voltage, conductivity, or resistance that varies according to the detachment of gas components adsorbed on the semiconductor sensor.

5. The detection device according to any one of claims 1 to 4, wherein the gas component is introduced into the semiconductor sensor without being pressurized or liquefied.

6. The detection device according to any one of claims 1 to 4, wherein the control unit inputs the plurality of signal patterns into a model generated by machine learning of the relationship between the plurality of signal patterns of the plurality of semiconductor sensors and the type or amount of allergens, thereby identifying or quantifying allergens in the target space.

7. The detection device according to claim 1, wherein the plurality of semiconductor sensors comprises two or more types of semiconductor sensors, and the two or more types of semiconductor sensors are composed of different metal oxides.

8. The detection device according to claim 4, wherein the conductivity is repeated in a cycle of decrease and recovery. The control unit extracts one waveform between the maximum and minimum values ​​of the conductivity as the signal pattern, and repeatedly uses the signal pattern to identify or quantify allergens.

9. The detection device according to claim 8, wherein when the amount of a quantitative allergen exceeds a threshold, or when the slope of the amount of allergen relative to time exceeds a threshold, the control unit controls the environmental equipment to reduce or prevent the amount of allergen from increasing.

10. The detection device according to claim 1, wherein the control unit displays the change in the amount of allergen over time according to the type of allergen.

11. The detection device according to claim 1, wherein the allergen is pollen, mold, or mites.

12. The detection device according to claim 1, wherein the plurality of semiconductor sensors simultaneously output signals regarding the plurality of gas components, The control unit acquires the signal pattern, which includes the signals simultaneously output by the plurality of semiconductor sensors.

13. A detection method, which is a detection device for identifying or quantifying allergens in a target space. Control Department Acquire signal patterns detected by multiple semiconductor sensors that exhibit relatively different response characteristics to the same gas composition. Based on the multiple signal patterns acquired from the multiple semiconductor sensors, allergens are identified or quantified. Output the results of identification or quantification.

14. A program for enabling a computer to function as the detection device of claim 1.

Images