JP7518413B2 - Pollen forecasting device, air treatment system, pollen forecasting method, and program - Google Patents

Description

This disclosure relates to a pollen forecasting device, an air treatment system, a pollen forecasting method, and a program.

Patent Document 1 discloses a pollen sensor that measures suspended pollen particles. In this pollen sensor, light with a predetermined polarization direction is irradiated into the air containing suspended particles, and the particle diameter of the suspended particles is measured based on the intensity of the scattered light from the suspended particles and the intensity of the orthogonal scattered light. This allows the pollen particles to be identified.

JP 3850418 A

A pollen sensor like that in Patent Document 1 can detect pollen particles near the pollen sensor, but cannot detect pollen particles floating in locations far from the pollen sensor. Therefore, even if a pollen sensor like that in Patent Document 1 is installed in a target space where people are staying, there are cases where it is not possible to grasp the floating status of pollen in the entire target space.

The purpose of this disclosure is to predict the pollen dispersion conditions throughout the entire target space.

The first aspect is directed to a pollen prediction device that predicts the floating state of pollen in a target space (S). The pollen prediction device (100) includes a first sensor (51) that detects the amount of dust floating in the target space (S), and a control unit (63) that obtains a statistical distribution that is generated in advance based on the amount of dust detected indoors, and predicts the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51).

In the first aspect, the control unit (63) predicts the possibility of pollen being suspended in the target space (S) based on a statistical distribution generated in advance and a detection value of a first sensor (51) that detects the amount of dust suspended in the target space (S). This makes it possible to predict the pollen suspension situation throughout the target space (S).

In the second aspect, in the first aspect, the control unit (63) acquires the statistical distribution that is generated in advance based on the amount of dust and the amount of pollen detected indoors, and predicts the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51).

In the second aspect, the control unit (63) predicts the possibility of pollen being suspended using a statistical distribution generated based on the amount of pollen detected in the room in addition to the amount of dust detected in the room, thereby making it possible to predict the possibility with greater accuracy.

In the third aspect, in the first or second aspect, the control unit (63) predicts the possibility using the degree of outlier, where the degree of outlier is the distance from the center of the statistical distribution.

In the third aspect, the possibility of pollen being suspended in the target space (S) is predicted using the outlier degree, which is the distance from the center of the statistical distribution. Therefore, it is possible to grasp the difference between the suspended dust state in the acquired statistical distribution and the suspended dust state in the target space (S). This allows for more accurate prediction of the possibility of pollen being suspended in the entire target space (S).

A fourth aspect is any one of the first to third aspects, in which the statistical distribution is generated based on the amount of dust when no pollen is detected in the room.

In the fourth aspect, the possibility of pollen being suspended in the target space (S) is predicted using a statistical distribution generated based on the amount of dust when no pollen is detected indoors. According to this, when the detection value of the first sensor (51) deviates significantly from the acquired statistical distribution, it becomes easier to determine that there is a high possibility that pollen is suspended in the target space (S). Therefore, it is possible to more accurately predict the possibility of pollen being suspended in the target space (S).

In a fifth aspect, in any one of the first to fourth aspects, the control unit (63) obtains a relational equation in which the amount of dust and the amount of pollen in the room are variables, and predicts the amount of pollen floating in the target space (S) by inputting the detection value of the first sensor (51) into the relational equation.

In the fifth aspect, the control unit (63) can use the acquired relational expression to predict the amount of pollen floating in the target space (S) from the amount of dust floating in the target space (S), which is the detection value of the first sensor (51).

A sixth aspect is any one of the first to fifth aspects, in which the statistical distribution is generated by machine learning.

In the sixth aspect, the possibility that pollen is floating in the target space (S) is predicted based on a statistical distribution generated by machine learning, so that the possibility can be predicted with greater accuracy.

The seventh aspect is an air treatment system comprising a pollen prediction device according to any one of the first to sixth aspects, and an air treatment device (20) that treats the air in the target space (S) based on the prediction results of the pollen prediction device.

In the seventh aspect, the air in the target space (S) is treated based on the prediction results of the pollen prediction device, so that pollen can be removed early before people staying in the target space (S) inhale the pollen.

The eighth aspect is directed to a pollen prediction method for predicting the floating state of pollen in a target space (S). The pollen prediction method includes a step of acquiring the amount of dust floating in the target space (S) detected by a first sensor (51) of a pollen prediction device (100), and a step of acquiring a statistical distribution generated in advance based on the amount of dust detected indoors, and predicting the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51).

In the eighth aspect, the possibility that pollen is floating in the target space (S) is predicted based on the acquired statistical distribution and the amount of dust floating in the target space (S). This makes it possible to predict the pollen floating situation throughout the target space (S).

The ninth aspect is directed to a program for predicting the floating state of pollen in a target space (S). The program causes a computer to execute a process of acquiring the amount of dust floating in the target space (S) detected by a first sensor (51) of a pollen prediction device (100), and a process of acquiring a statistical distribution generated in advance based on the amount of dust detected indoors, and predicting the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51).

In the ninth aspect, the program is executed to predict the possibility that pollen is floating in the target space (S) based on the acquired statistical distribution and the amount of dust floating in the target space (S). This makes it possible to predict the pollen floating situation throughout the target space (S).

FIG. 1 is a block diagram of an air treatment system including a pollen prediction device according to an embodiment. FIG. 2 is a schematic diagram of a target space to which the pollen prediction device is applied. FIG. 3 is an explanatory diagram showing the selection of data used in the statistical distribution generating unit. FIG. 4 is an explanatory diagram showing the degree of deviation. FIG. 5 is an explanatory diagram for explaining the threshold value. FIG. 6 is a flowchart showing the statistical distribution generation operation. FIG. 7 is a flowchart showing the pollen airborne prediction operation. FIG. 8 is a diagram corresponding to FIG. 1 according to the first modified example of the embodiment. FIG. 9 is a diagram corresponding to FIG. 1 according to the third modified example of the embodiment. FIG. 10 is a flowchart showing a relational expression generating operation according to the third modification of the embodiment. FIG. 11 is a flowchart showing a pollen amount prediction operation according to the third modification of the embodiment. FIG. 12 is an explanatory diagram showing a method for calculating the rate of possibility of pollen floating in the air in the eighth embodiment.

Embodiments of the present disclosure will be described in detail below with reference to the drawings. Note that the present disclosure is not limited to the embodiments shown below, and various modifications are possible without departing from the technical spirit of the present disclosure. Each drawing is intended to conceptually explain the present disclosure, and therefore dimensions, ratios, or numbers may be exaggerated or simplified as necessary for ease of understanding.

<<Embodiment>>
An air treatment system (1) according to an embodiment will be described.

(1) Overall Configuration of the Air Treatment System As shown in Fig. 1, the air treatment system (1) of this embodiment includes a pollen prediction device (100), an air purifier (20) as an air treatment device, and a communication terminal (30). The pollen prediction device (100) includes a pollen sensor unit (40), a dust sensor unit (50), and a server device (60).

The dust sensor unit (50) and the pollen sensor unit (40) are connected to a server device (60) via the Internet (N). The server device (60) is also connected to an air purifier (20) and a communication terminal (30) via the Internet (N).

As shown in FIG. 2, the air treatment system (1) is used in a target space (S) in which a person (H) is present. The target space (S) is an indoor space such as an ordinary house or an office. The pollen prediction device (100) of the air treatment system (1) predicts the pollen floating condition in the target space (S). Specifically, the pollen prediction device (100) predicts the possibility that pollen is floating in the target space (S) (hereinafter referred to as pollen floating possibility). Details of the pollen prediction device (100) will be described later.

The air treatment device (20) of this embodiment is an air purifier. The air purifier (20) is placed in the target space (S). The air purifier (20) has a communication unit (21) and an air purification control unit (22). The communication unit (21) includes a receiving unit that receives the prediction result of the possibility of pollen floating transmitted from the pollen prediction device (100). The air purification control unit (22) controls the fan of the air purifier (20) to treat (purify) the air in the target space (S) based on the received prediction result.

The communication terminal (30) is, for example, a personal computer, a smartphone, a tablet terminal, a mobile phone, etc. The communication terminal (30) has a communication unit (31) and a notification unit (32). The communication unit (31) includes a receiving unit that receives the prediction result of the possibility of pollen floating transmitted from the pollen prediction device (100). The notification unit (32) notifies the received prediction result. The notification unit (32) includes a display unit that displays the prediction result and a sound output unit that outputs the prediction result as sound. The display unit is, for example, configured with a liquid crystal monitor. The sound output unit is, for example, configured with a speaker. Note that, if the air purifier (20) has a notification unit, the notification unit of the air purifier (20) may notify the prediction result of the pollen prediction device (100).

(2) Details of the Pollen Prediction Device As described above, the pollen prediction device (100) has a pollen sensor unit (40), a dust sensor unit (50), and a server device (60). As shown in Fig. 2, the pollen sensor unit (40) and the dust sensor unit (50) are disposed in the target space (S). The pollen prediction device (100) of this embodiment has one pollen sensor unit (40) and one dust sensor unit (50).

(2-1) Pollen Sensor Unit The pollen sensor unit (40) includes a pollen sensor (41), a data processing unit (42), and a communication unit (43). The pollen sensor (41) detects the amount of pollen floating in the air in the room. In addition, the pollen sensor (41) of this embodiment detects the type of pollen floating in the air in the room. The pollen sensor (41) of this embodiment is an imaging device (camera).

The data processing unit (42) outputs information about pollen floating in the target space (S) by analyzing image data captured by the pollen sensor (41). In this embodiment, the information about pollen is the number of pollen particles and the type of pollen. If no pollen is detected in the target space (S), the data processing unit (42) outputs that the number of pollen particles is 0.

The communication unit (43) is connected to the Internet (N) via a wired or wireless communication line. The communication unit (43) includes a transmission unit that transmits the pollen-related information output by the data processing unit (42) to the server device (60) via the Internet (N). Here, the processing in the data processing unit (42) requires a certain amount of time to output significant information related to pollen. The time required for this processing is, for example, 20 minutes. Therefore, the communication unit (43) transmits the pollen-related information to the server device (60) approximately every 20 minutes.

(2-2) Dust Sensor Unit The dust sensor unit (50) includes a dust sensor (51), a data processing unit (52), and a communication unit (53). The dust sensor (51) detects the amount of dust floating in the room. The dust sensor (51) corresponds to the first sensor of the present disclosure.

The dust sensor (51) of this embodiment is a sensor capable of detecting fine particles having a particle diameter of 2.5 μm or less (a so-called PM2.5 sensor). The dust sensor (51) may be a sensor capable of detecting fine particles having a particle diameter of 10 μm or less (a so-called PM10 sensor), or may be a sensor capable of detecting both fine particles having a particle diameter of 2.5 μm or less and fine particles having a particle diameter of 10 μm or less. The dust referred to here includes dust, dirt, yellow sand, dead mites, human skin flakes, etc.

The data processing unit (52) processes the output of the dust sensor (51) to output the amount of dust as a dust concentration in the target space (S). The data processing unit (52) outputs an integrated value (integrated trend) of the output of the dust sensor (51) per unit time.

The communication unit (53) is connected to the Internet (N) via a wired or wireless communication line. The communication unit (53) includes a transmission unit that transmits the dust concentration output by the data processing unit (52) to the server device (60) via the Internet (N). Here, the processing in the data processing unit (52) requires a predetermined time. The time required for this processing is, for example, one minute. Therefore, the communication unit (53) transmits the dust concentration to the server device (60) approximately every minute.

Here, pollen is composed of pollen particles and pollen microparticles that adhere to the surface of the pollen particles. The pollen microparticles include microparticles called orbicules and pollen pieces that are generated when pollen particles are crushed. The particle diameter of pollen particles is about 30 μm. On the other hand, the particle diameter of pollen microparticles is about 0.3 μm to about 12 μm, which is much smaller than the pollen particles.

Therefore, relatively heavy pollen particles tend to fall in a relatively short time after entering an indoor space. In contrast, light pollen particles have the characteristic of being more likely to remain suspended in the indoor space for a long time and to be dispersed and suspended throughout the indoor space. In addition, pollen particles have the characteristic of being suspended in greater amounts than pollen particles, since many pollen particles are attached to one pollen particle.

The dust sensor (51) can detect these pollen particles in addition to dust and dirt. Therefore, the output of the dust sensor (51) can be used to predict the possibility that pollen particles will be suspended in the target space (S).

Here, it is conceivable to adopt only the pollen sensor (41) as the sensor to be equipped in the air treatment system (1). However, since it takes about 20 minutes for the pollen sensor unit (40) to output information regarding pollen, pollen may spread within the target space (S) while waiting for the output result of the pollen sensor unit (40), which may worsen the hay fever of the person (H) staying in the target space (S).

In contrast, the air treatment system (1) of this embodiment is equipped with a dust sensor unit (50) that can output detection results in a relatively short time, making it possible to predict the pollen floating condition in the target space (S) at an early stage. This makes it possible to prevent the worsening of hay fever in people (H) in the target space (S).

(2-3) Server Device The server device (60) is provided on a cloud on the Internet (N) and includes a control unit (63), a storage unit (61), and a communication unit (62).

The communication unit (62) includes a receiving unit that receives data transmitted from the pollen sensor unit (40) and the dust sensor unit (50). The communication unit (62) includes a transmitting unit that transmits the prediction result of the possibility of pollen floating, which will be described later, to the air purifier (20) and the communication terminal (30).

The memory unit (61) stores data received by the communication unit (62). The memory unit (61) also stores a statistical distribution, which will be described later. The memory unit (61) includes at least one of a hard disk drive (HDD), a random access memory (RAM), and a solid state drive (SSD).

The control unit (63) includes a microcomputer and a memory device. The memory device stores software for operating the microcomputer. The control unit (63) has, as functional components, a statistical distribution generation unit (64) and a probability prediction unit (65). In other words, the control unit (63) functions as the statistical distribution generation unit (64) and the probability prediction unit (65) by executing a program stored in the memory device.

The program stored in the memory device of the control unit (63) corresponds to the program of the present disclosure. This program causes the control unit (63) as a computer to execute a first process and a second process. In the first process, the amount of dust suspended in the target space (S) detected by the dust sensor (51) is acquired. In the second process, a statistical distribution generated in advance based on the amount of dust detected in the target space (S) is acquired, and the possibility that pollen is suspended in the target space (S) is predicted based on the acquired statistical distribution and the detection value of the dust sensor (51).

(2-3-1) Statistical Distribution Generator Because the amount of dust detected in the target space (S) changes each time, the dust suspension state in the target space (S) can be expressed as a distribution. Therefore, the statistical distribution generator (64) generates a statistical distribution based on the amount of dust previously detected in the target space (S). In detail, the statistical distribution generator (64) generates a statistical distribution based on the detection value of the dust sensor (51) stored in the memory unit (61).

The statistical distribution in this embodiment is a probability density function generated based on past detection values of the dust sensor (51). This probability density function is a probability density function in which the horizontal axis represents the detection value of the dust sensor (51) (specifically, the dust concentration) and the vertical axis represents the occurrence frequency or occurrence probability. Note that the statistical distribution generated by the statistical distribution generating unit (64) may be a histogram of past detection values of the dust sensor (51).

In this embodiment, the data used to generate the statistical distribution is the amount of dust when no pollen is detected in the target space (S). In detail, when the pollen amount data transmitted from the pollen sensor unit (40) is zero, the statistical distribution is generated using the dust amount data (the amount of dust when no pollen is detected) transmitted from the dust sensor unit (50) to the server device (60) during the period when the pollen amount is zero. In this manner, the statistical distribution generation unit (64) generates the statistical distribution based on the detection values of the pollen sensor (41) and the dust sensor (51).

The selection of data used to generate the statistical distribution will be described in more detail with reference to FIG. 3. As shown in FIG. 3, when looking at the detection value of the pollen sensor (41) during a period when pollen is relatively not dispersed outdoors (hereinafter referred to as outside the pollen season, for example, June to December), the pollen sensor (41) may occasionally detect pollen.

This may be because the pollen sensor (41) mistakenly detects non-pollen particles as pollen. Therefore, in this case, it is unclear whether the detected dust contains fine pollen particles, and the data detected by the dust sensor (51) is unreliable.

Therefore, in this embodiment, outside of the pollen season, only the detection values of the dust sensor (51) when pollen is not detected by the pollen sensor (41) (area F1 in FIG. 3) are selected, and a statistical distribution is generated using only the selected data on the amount of dust when pollen is not detected. This makes it possible to generate a statistical distribution of the detection values of the dust sensor (51) when no pollen is present in the target space (S). As a result, it is possible to accurately grasp the tendency of the amount of dust when pollen is not floating in the target space (S), which serves as a criterion for determining the possibility of pollen being floating in the air.

The statistical distribution in this embodiment is generated by machine learning. As described above, learning is performed using data on the amount of dust when pollen is not detected, so a highly accurate statistical distribution can be generated.

(2-3-2) Possibility Prediction Unit The possibility prediction unit (65) acquires a statistical distribution previously generated in the statistical distribution generation unit (64), and predicts the possibility that pollen is floating in the target space (S) (hereinafter referred to as pollen floating possibility) based on the acquired statistical distribution and the detection value C of the dust sensor (51) at the current time.

Specifically, the possibility prediction unit (65) compares the acquired statistical distribution with the detection value C of the dust sensor (51) at the current time, and predicts the possibility of pollen being airborne by evaluating the degree of deviation Z calculated by the comparison.

Here, the "degree of outlier Z" is the distance between the center of the statistical distribution and the detection value C of the dust sensor (51) at the current point in time in the statistical distribution, as shown in FIG. 4. In this embodiment, the "center of the statistical distribution" is the average value m. Note that the "center of the statistical distribution" may also be the median or the mode.

In this way, by using the degree of deviation Z to predict the possibility of pollen floating, it is possible to understand the difference between the dust floating state in the statistical distribution and the current dust floating state.

The statistical distribution in this embodiment is generated based on data on the amount of dust when no pollen is detected in the target space (S). Therefore, when the outlier Z is calculated in this statistical distribution, if the value of outlier Z is large (the amount of dust at the current time is far from the center of the statistical distribution), it can be inferred that the environment of the target space (S) at the current time is different from the environment of the target space (S) when no pollen is detected. This makes it possible to predict that there is a high possibility that pollen is floating in the target space (S) at the current time.

In this embodiment, the degree of deviation Z is evaluated based on whether the degree of deviation Z exceeds a preset threshold A. Specifically, if the degree of deviation Z is equal to or less than the threshold A, it is determined that there is no possibility of pollen being airborne (Z≦A). On the other hand, if the degree of deviation Z is greater than the threshold A, it is determined that there is a possibility of pollen being airborne (A<Z).

Here, the threshold A is, for example, as shown in FIG. 5, a value of the area where a statistical distribution P generated based on dust amount data acquired outside of the pollen season overlaps with a statistical distribution Q generated based on dust amount data acquired during the pollen season. Note that the pollen season refers to a period when pollen is relatively dispersed outdoors, for example, the period from January to May.

While the data on the amount of dust during pollen season includes dust and pollen, the data on the amount of dust outside of pollen season mainly includes dust and almost no pollen. Therefore, as shown in FIG. 5, the statistical distribution P outside of pollen season and the statistical distribution Q during pollen season tend to have different distributions. Therefore, in this embodiment, the value of the area where the statistical distribution P outside of pollen season and the statistical distribution Q during pollen season overlap is set as the threshold A, so that the presence or absence of the possibility of pollen floating in the target space (S) can be determined. Note that the threshold A may be, for example, the maximum value in the statistical distribution P outside of pollen season or the minimum value in the statistical distribution Q during pollen season. In addition, if the statistical distribution P outside of pollen season and the statistical distribution Q during pollen season do not overlap, any value between the two statistical distributions may be set as the threshold A.

(3) Operation of the Pollen Prediction Device The operation of the pollen prediction device (100) will be described. When the pollen prediction device (100) starts operating, the pollen sensor unit (40) and the dust sensor unit (50) start detecting pollen and dust. The pollen sensor unit (40) and the dust sensor unit (50) sequentially transmit the detection values detected by each unit to the server device (60) via the Internet (N). The server device (60) predicts the possibility of pollen airborne based on the received detection values, and outputs the prediction result.

Here, the operation of the server device (60) will be described in detail with reference to Figs. 6 and 7. The control unit (63) simultaneously performs the statistical distribution generation operation and the pollen airborne prediction operation.

As shown in FIG. 6, in the statistical distribution generation operation, the server device (60) receives information about pollen transmitted from the pollen sensor unit (40) (step ST11). Next, the server device (60) stores the received information about pollen in the memory unit (61) (step ST12).

The server device (60) receives the dust amount data transmitted from the dust sensor unit (50) at the same time as step ST11 (step ST13). The server device (60) then stores the received dust amount data in the memory unit (61) (step ST14).

Next, the statistical distribution generating unit (64) of the control unit (63) acquires dust amount data when pollen was not detected in the target space (S) during a predetermined period from the dust amount data stored in the memory unit (61), and generates a statistical distribution from the acquired dust amount data (step ST15). At this time, the statistical distribution generating unit (64) generates the statistical distribution by machine learning. Next, the server device (60) stores the generated statistical distribution in the memory unit (61) (step ST16).

This statistical distribution generation operation is repeated at predetermined time intervals. As a result, the statistical distribution stored in the memory unit (61) is periodically updated to the latest statistical distribution.

As shown in FIG. 7, in the pollen airborne prediction operation, the possibility prediction unit (65) of the control unit (63) acquires the latest statistical distribution stored in the memory unit (61) (step ST21). Next, the server device (60) receives data on the amount of dust, which is the detection value C of the dust sensor (51) at the current time (step ST22).

The likelihood prediction unit (65) calculates the degree of deviation Z based on the acquired statistical distribution and the amount of dust at the current time (step ST23). Specifically, the statistical distribution is compared with the amount of dust at the current time, and the degree of deviation Z in the statistical distribution is calculated.

Next, the possibility prediction unit (65) evaluates whether the calculated deviation degree Z exceeds a preset threshold value A (step ST24). Next, the possibility prediction unit (65) outputs the evaluation result of the deviation degree Z as a prediction result of the possibility of pollen being airborne (step ST25). Next, the communication unit (62) transmits the prediction result output from the possibility prediction unit (65) to the air purifier (20) and the communication terminal (30) (step ST26).

This pollen airborne prediction operation is repeated every predetermined time (e.g., one minute). This allows the latest prediction results to be output according to changes in the environment of the target space (S) (e.g., whether ventilation is present or not).

(4) Operation of the Air Treatment System The operation of the air treatment system (1) will be described below. When the air treatment system (1) starts operating, the pollen prediction device (100) and the air purifier (20) start operating.

In the communication terminal (30), when the communication unit (31) receives the prediction result of the possibility of pollen floating from the server device (60), it notifies the notification unit (32) of the received prediction result. This makes it possible to encourage people (H) staying in the target space (S) to take actions to prevent hay fever. Specifically, it is possible to encourage actions such as cleaning the target space (S), spraying a pollen bursting prevention agent, and maintaining an air purifier.

Furthermore, in the air purifier (20), when the communication unit (21) receives the prediction result, the air purification control unit (22) controls the fan to purify the air in the target space (S) based on the prediction result. Specifically, when the communication unit (21) receives a prediction result indicating that there is no possibility of pollen being airborne, the air purification control unit (22) does not change the control at the time of reception. On the other hand, when the communication unit (21) receives a prediction result indicating that there is a possibility of pollen being airborne, the air purification control unit (22) increases the rotation speed of the fan to promote purification of the air in the target space (S).

In this way, the air purification control unit (22) controls the air purifier (20) based on the prediction results received, so the air in the target space (S) can be purified early. This makes it possible to remove pollen before it is inhaled by people (H) staying in the target space (S). As a result, it is possible to prevent hay fever or reduce pollen symptoms.

(5) Features (5-1)
The control unit (63) acquires a statistical distribution that has been generated in advance based on the amount of dust detected in the target space (S) in the past, and predicts the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value C of the dust sensor (51) at the current time.

Here, the dust sensor (51) can detect fine pollen particles, which have a smaller particle diameter than pollen grains. Because the fine pollen particles are small and light, they remain suspended in the target space (S) for a long time and are dispersed and suspended throughout the target space (S). Therefore, by using the detection value of the dust sensor (51), it is possible to predict the possibility that pollen is suspended throughout the target space (S).

In addition, the dust sensor (51) can output detection results in a shorter time than the pollen sensor (41). Therefore, by using the detection value of the dust sensor (51) to predict the pollen suspension state in the target space (S), it is possible to output prediction results early.

(5-2)
The control unit (63) predicts the possibility that pollen is floating in the target space (S) based on a statistical distribution that has been generated in advance based on the amounts of dust and pollen previously detected in the target space (S) and on the detection value of the dust sensor (51).

The statistical distribution is generated based on the amount of pollen previously detected in the target space (S) in addition to the amount of dust, allowing for more accurate prediction of the possibility of pollen being airborne.

(5-3)
When the distance from the center in the statistical distribution is taken as the outlier degree Z, the control unit (63) predicts the possibility of pollen being suspended by using the outlier degree Z. This makes it possible to grasp the difference between the suspension state of dust in the acquired statistical distribution and the suspension state of dust in the target space (S), thereby making it possible to more accurately predict the possibility of pollen being suspended in the entire target space (S).

(5-4)
The statistical distribution is generated based on the amount of dust when pollen is not detected in the target space (S). According to this, if the detection value of the dust sensor (51) is significantly different from the acquired statistical distribution, it is easy to determine that there is a high possibility that pollen is floating in the target space (S). Therefore, it is possible to more accurately predict the possibility of pollen being floating in the target space (S).

(5-5)
The statistical distribution is generated by machine learning, which allows for more accurate prediction of pollen airborne probability.

(5-6)
The air treatment system (1) includes a pollen prediction device (100) and an air purifier (20) that treats the air in a target space (S) based on the prediction result of the pollen prediction device (100). Since the air in the target space (S) is treated based on the prediction result of the pollen prediction device, pollen can be removed early before a person staying in the target space (S) inhales the pollen.

(5-7)
The pollen prediction method includes the steps of acquiring the amount of dust floating in a target space (S) detected by a dust sensor (51) of a pollen prediction device (100), and acquiring a statistical distribution generated in advance based on the amount of dust previously detected in the target space (S), and predicting the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the dust sensor (51).

Accordingly, by using the detection value of the dust sensor (51), it is possible to predict the possibility that pollen is floating throughout the target space (S).

In addition, the dust sensor (51) can output detection results in a shorter time than the pollen sensor (41). Therefore, by using the detection value of the dust sensor (51) to predict the pollen suspension state in the target space (S), it is possible to output prediction results early.

(5-8)
The program causes a computer to execute a process of acquiring the amount of dust floating in the target space (S) detected by a dust sensor (51) of the pollen prediction device (100), and a process of acquiring a statistical distribution generated in advance based on the amount of dust detected in the target space (S) in the past, and predicting the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the dust sensor (51).

Accordingly, by using the detection value of the dust sensor (51), it is possible to predict the possibility that pollen is floating throughout the target space (S).

In addition, the dust sensor (51) can output detection results in a shorter time than the pollen sensor (41). Therefore, by using the detection value of the dust sensor (51) to predict the pollen suspension state in the target space (S), it is possible to output prediction results early.

(6) Modifications The above embodiment may be modified as follows: In the following description, the differences from the above embodiment will be mainly described.

(6-1) Modification 1
As shown in FIG. 8, in the pollen prediction device (100) of this embodiment, the statistical distribution generated by the control unit (63) may be generated based on the amount of dust and the amount of pollen detected in an indoor space other than the target space (S).

Specifically, a sensor unit (SU) is placed in an indoor space other than the target space (S). This sensor unit (SU) is a combination of a pollen sensor unit (40) and a dust sensor unit (50). The sensor unit (SU) detects dust and pollen in the indoor space and transmits information on the amount of dust and pollen to a server device (60) via the Internet (N).

The control unit (63) of the server device (60) generates a statistical distribution based on the amount of dust and the amount of pollen in other indoor spaces, and predicts the possibility of pollen being suspended in the target space (S) based on the generated statistical distribution and the detection value of the dust sensor (51) placed in the target space (S).

In this modified example, the statistical distribution is generated based on data from indoor spaces other than the target space (S), so the statistical distribution is generated from a larger amount of data. This makes it possible to more accurately predict the possibility of pollen being airborne.

(6-2) Modification 2
In the above embodiment, the statistical distribution is generated based on the detection values of the dust sensor (51) and the pollen sensor (41), but it may be generated based only on the detection value of the dust sensor (51). In this case, for example, by generating the statistical distribution based on the detection value of the dust sensor (51) during a period when pollen is relatively low in the air, it is possible to generate a statistical distribution that serves as a basis for predicting the possibility of pollen floating even without information on the detection value of the pollen sensor (41).

(6-3) Modification 3
In the pollen prediction device (100) of this embodiment, the control unit (63) may predict the amount of pollen floating in the target space (S). Specifically, as shown in Fig. 9, the control unit (63) further includes, as functional components, a relational equation generation unit (66) and a pollen amount prediction unit (67).

The relational equation generating unit (66) generates a relational equation in which the amount of dust is an explanatory variable and the amount of pollen is a target variable. In detail, the relational equation generating unit (66) generates the relational equation based on past detection values of the dust sensor (51) and the pollen sensor (41) stored in the memory unit (61). The amount of pollen in this modified example is the number of pollen grains.

The pollen amount prediction unit (67) obtains the relational equation generated by the relational equation generation unit (66) and predicts the number of pollen particles floating in the target space (S) by inputting the current detection value C of the dust sensor (51) into the relational equation.

In the operation of the server device (60) of the pollen prediction device (100), the control unit (63) further simultaneously performs a relational equation generation operation and a pollen amount prediction operation.

As shown in FIG. 10, in the relational equation generation operation, the relational equation generation unit (66) acquires data on the amount of dust and the number of pollen grains stored in the memory unit (61) (step ST31). Next, the relational equation generation unit (66) generates a relational equation using the amount of dust and the amount of pollen as variables from the acquired data (step ST32). At this time, the relational equation generation unit (66) generates the relational equation by statistical processing or machine learning. Next, the relational equation generation unit (66) and the server device (60) store the generated relational equation in the memory unit (61) (step ST33).

This operation of generating the relational equation is repeated at predetermined time intervals. As a result, the relational equation stored in the memory unit (61) is periodically updated to the latest relational equation.

As shown in FIG. 11, in the pollen amount prediction operation, the pollen amount prediction unit (67) acquires the latest relational equation stored in the memory unit (61) (step ST41). Next, the server device (60) receives data on the amount of dust, which is the current detection value C of the dust sensor (51) (step ST42).

Next, the pollen amount prediction unit (67) calculates a predicted value of the pollen count based on the obtained relational equation and the current amount of dust (step ST43). Specifically, the current amount of dust is input to the obtained relational equation, and the predicted value of the pollen count is calculated and output.

Next, the communication unit (62) transmits the predicted value output from the pollen amount prediction unit (67) to the communication terminal (30) (step ST44).

This pollen amount prediction operation is repeated at a predetermined time interval (e.g., every minute). This allows the latest predicted value to be output in response to changes in the environment of the target space (S) (e.g., whether ventilation is present or not).

According to this modified example, the control unit (63) can use the acquired relational equation to predict the amount of pollen floating in the target space (S) from the amount of dust floating in the target space (S), which is the detection value of the dust sensor (51).

(6-4) Modification 4
The pollen prediction device (100) of this embodiment may have a plurality of dust sensors (51) with different detection ranges. For example, the dust sensors (51) may include a sensor (PM2.5 sensor) capable of detecting fine particles having a particle diameter of 2.5 μm or less and a sensor (PM10 sensor) capable of detecting fine particles having a particle diameter of 10 μm or less. In this case, a statistical distribution is generated based on the detection values of the dust sensors (51). In this case, a plurality of two-dimensional statistical distributions may be generated, or a three-dimensional statistical distribution may be generated.

In this case, in addition to the statistical distribution generated based on the detection values of each dust sensor (51), a statistical distribution may be generated based on the ratio of the output values of each dust sensor (51). Specifically, for example, the statistical distribution may be a probability density function with the horizontal axis representing the ratio of the PM2.5 sensor detection value to the PM10 sensor detection value (PM2.5/PM10).

(6-5) Modification 5
A plurality of dust sensors (51) of the pollen prediction device (100) of this embodiment may be disposed in the target space (S), thereby making it possible to more accurately predict the possibility of pollen being suspended in the entire target space (S).

(6-6) Modification 6
When the air purifier (20) is equipped with a dust sensor, the dust sensor of the air purifier (20) may be used as the dust sensor (51) of the pollen forecasting device (100) of the present embodiment.

(6-7) Modification 7
In the pollen prediction device (100) of this embodiment, the detection value C of the dust sensor (51) at the current time used by the possibility prediction unit (65) may be a distribution of outputs of the dust sensor (51) for a predetermined time including the current time. Specifically, the possibility prediction unit (65) predicts the possibility of pollen floating by comparing the acquired statistical distribution with a distribution of detection values of the dust sensor (51) for a predetermined period including the current time (hereinafter referred to as the output distribution at the current time).

In this case, the possibility of pollen being airborne is predicted by calculating the degree of outlier Z by comparing the acquired statistical distribution with the current output distribution, and evaluating the calculated degree of outlier Z. In this case, the degree of outlier Z is the distance between the center (e.g., mean, median, mode) of the statistical distribution and the center (e.g., mean, median, mode) of the current output distribution. When predicting the possibility of pollen being airborne using the degree of outlier Z, the possibility prediction unit (65) evaluates whether the calculated degree of outlier Z exceeds a preset threshold value, as in the above-described embodiment.

According to this modified example, the possibility prediction unit (65) predicts the possibility of pollen being airborne based on the output distribution at the current time, and the possibility of pollen being airborne is predicted by comparing the temporal characteristics of the past statistical distribution and the current output of the dust sensor (51). This makes it possible to predict the possibility of pollen being airborne with greater accuracy.

(6-8) Modification 8
In the pollen prediction device (100) of this embodiment, the prediction result of the possibility of pollen floating, output by the possibility prediction section (65), may be output as a ratio.

Specifically, a first threshold A1 and a second threshold A2 larger than the first threshold A1 are set in advance as thresholds for evaluating the possibility of pollen being airborne. Then, as shown in FIG. 12, the possibility prediction unit (65) outputs "Possibility of pollen being airborne is 0%" when the calculated degree of deviation Z is equal to or less than the first threshold A1, outputs "Possibility of pollen being airborne is 100%" when the calculated degree of deviation Z is equal to or greater than the second threshold A2, and outputs the percentage of the possibility of pollen being airborne according to the degree of deviation Z when the calculated degree of deviation Z is greater than the first threshold A1 and less than the second threshold A2.

Here, for example, as shown in FIG. 12, when a statistical distribution P outside of pollen season overlaps with a statistical distribution Q during pollen season, the start value of the overlap may be set to a first threshold value A1, and the end value of the overlap may be set to a second threshold value A2.

(6-9) Modification 9
In the pollen prediction device (100) of this embodiment, data used by the statistical distribution generating unit (64) when generating a statistical distribution may be the amount of dust when pollen is detected in the target space (S). In detail, when the pollen amount data transmitted from the pollen sensor unit (40) is greater than 0, the statistical distribution is generated using the dust amount data (the amount of dust when pollen is detected) transmitted from the dust sensor unit (50) to the server device (60) during a period when the amount of pollen is greater than 0.

As shown in FIG. 3, when looking at the detection value of the pollen sensor (41) during pollen season, there are cases where the pollen sensor (41) does not detect pollen. This may be because the pollen sensor (41) is unable to detect pollen floating in the target space (S). Therefore, it can be said that the reliability of the detection value data of the dust sensor (51) at this time is low.

Therefore, during the pollen season, by selecting only the detection values (area F2 in FIG. 3) of the dust sensor (51) when pollen is detected by the pollen sensor (41) and generating a statistical distribution using only the selected data on the amount of dust when pollen is detected, it is possible to generate a statistical distribution of the detection values of the dust sensor (51) in a state in which pollen is present in the target space (S). This makes it possible to accurately grasp the tendency of the amount of dust when pollen is floating in the target space (S), which serves as a criterion for determining the possibility of pollen being floating in the air.

In this modified example, the possibility prediction unit (65) predicts the possibility of pollen being airborne by evaluating the outlier Z, as in the above embodiment. The statistical distribution in this modified example is generated based on data on the amount of dust when pollen is detected in the target space (S). Therefore, when the outlier Z is calculated in the statistical distribution, if the value of the outlier Z is small (if the amount of dust at the current time is close to the center of the statistical distribution), it can be inferred that the environment of the target space (S) at the current time is similar to the environment of the target space (S) when pollen was detected. This makes it possible to infer that there is a high possibility that pollen is airborne in the target space (S) at the current time.

(6-10) Modification 10
The air treatment device of the present embodiment may be an air conditioner or a ventilator having an air purification function (e.g., an air purification filter). In this case, too, the control unit (63) controls a fan of the air conditioner or the ventilator based on the prediction result of the pollen prediction device (100) to activate the air purification function and purify the air in the target space (S).

Other Embodiments
The above embodiment may be configured as follows.

In the pollen prediction device (100) of the above embodiment, the control unit (63) and the memory unit (61) do not have to be provided in the server device (60). Specifically, the control unit (63) and the memory unit (61) of the pollen prediction device (100) may be provided in the air purifier (20), or may be provided in a unit separate from the air purifier (20) and placed in the target space (S).

In the pollen prediction device (100) of the above embodiment, the difference between the dust suspension state in the statistical distribution and the dust suspension state in the target space (S) is evaluated using a measure called the degree of outlier, but this is not limited to this, and the evaluation may be performed using, for example, a statistical testing method.

Although the embodiments and modifications have been described above, it will be understood that various modifications of form and details are possible without departing from the spirit and scope of the claims. Furthermore, elements of the above embodiments, modifications, and other embodiments may be combined or substituted as appropriate.

The descriptions "first," "second," "third," etc. mentioned above are used to distinguish the words to which these descriptions are attached, and do not limit the number or order of the words.

As described above, the present disclosure is useful for pollen forecasting devices, air treatment systems, pollen forecasting methods, and programs.

1. Air handling system
20 Air purifier (air treatment device)
51 Dust sensor (first sensor)
63 Control Unit
100 Pollen forecasting device
S target space

Claims (9)

A device for predicting the pollen dispersion state in a target space (S),
a first sensor (51) that detects an amount of dust suspended in the target space (S);
and a control unit (63) that acquires a statistical distribution that has been generated in advance based on the amount of dust and the amount of pollen detected indoors, and predicts the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51). The control unit (63) acquires the statistical distribution that has been generated in advance based on the amount of dust, the amount of pollen, and the type of pollen detected indoors, and predicts the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51).
The pollen forecasting device according to claim 1 . The pollen prediction device according to claim 1 or 2, wherein the control unit (63) predicts the possibility by using an outlier degree, the outlier degree being a distance from a center of the statistical distribution. The pollen prediction device according to claim 1 , wherein the statistical distribution is generated based on an amount of dust when no pollen is detected in the room. A device for predicting the pollen dispersion state in a target space (S),
a first sensor (51) that detects an amount of dust suspended in the target space (S);
a control unit (63) that acquires a statistical distribution that has been generated in advance based on an amount of dust detected indoors, and predicts a possibility that pollen is floating in the target space (S) based on the statistical distribution and a detection value of the first sensor (51),
The control unit (63) obtains a relational equation having variables of the amount of dust and the amount of pollen in the room, and predicts the amount of pollen floating in the target space (S) by inputting the detection value of the first sensor (51) into the relational equation.
Pollen forecasting device. The pollen forecasting device according to claim 1 or 5 , wherein the statistical distribution is generated by machine learning. The pollen forecasting device according to claim 1 or 5 ,
and an air treatment device (20) that treats the air in the target space (S) based on the prediction result of the pollen prediction device. A method for predicting pollen dispersion in a target space (S), comprising:
acquiring an amount of dust suspended in the target space (S) detected by a first sensor (51) of a pollen prediction device (100);
acquiring a statistical distribution generated in advance based on the amount of dust and the amount of pollen detected indoors, and predicting the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51). A program for predicting the pollen dispersion state in a target space (S),
A process of acquiring an amount of dust suspended in the target space (S) detected by a first sensor (51) of a pollen prediction device (100);
The program causes a computer to execute a process of acquiring a statistical distribution that has been generated in advance based on the amount of dust and the amount of pollen detected indoors, and predicting the possibility that pollen is floating in the target space (S) based on the statistical distribution and the detection value of the first sensor (51).