Carbon dioxide concentration control device, program, carbon dioxide concentration control system, and carbon dioxide concentration control method

Abstract

This will more effectively reduce the excess carbon dioxide supplied to plants. [Solution] The carbon dioxide concentration control device (10) comprises an information acquisition unit (11) that acquires humidity change information (INF) and carbon dioxide concentration information, and a concentration control unit (12) that controls a carbon dioxide application device (200) that applies carbon dioxide to the cultivation environment based on a prediction index (PI) obtained by inputting the humidity indicated by the humidity change information (INF), the humidity change rate of the humidity change, and the humidity change acceleration of the humidity change into a decision tree (20), and carbon dioxide concentration information.

Description

Technical Field

[0001] The present invention relates to a carbon dioxide concentration control device, a program, a carbon dioxide concentration control system, and a carbon dioxide concentration control method.

Background Art

[0002] Techniques for supplying carbon dioxide to plants to promote growth are known. For example, Non-Patent Document 1 discloses a technique for applying carbon dioxide to promote the growth of strawberries.

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In order to reduce the environmental load, a technique for reducing the surplus of carbon dioxide supplied to plants is required. In the technique disclosed in Non-Patent Document 1, since the absorption state of carbon dioxide in plants is not considered, there is a problem that a surplus of carbon dioxide is likely to occur in an environment where plants are difficult to absorb carbon dioxide.

[0005] One aspect of the present invention aims to more effectively reduce the surplus of carbon dioxide supplied to plants.

Means for Solving the Problems

[0006] To solve the above problems, a carbon dioxide concentration control device according to one aspect of the present invention is a carbon dioxide concentration control device for controlling the carbon dioxide concentration in a cultivation environment in which plants are cultivated, comprising: an information acquisition unit that acquires humidity change information indicating the measurement result of humidity change in the cultivation environment and carbon dioxide concentration information indicating the measurement result of carbon dioxide concentration in the cultivation environment; and a concentration control unit that controls a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired by the information acquisition unit, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired by the information acquisition unit.

[0007] To solve the above problems, a carbon dioxide concentration control system according to one aspect of the present invention is a carbon dioxide concentration control system for controlling the carbon dioxide concentration in a cultivation environment in which plants are cultivated, comprising: an information acquisition unit that acquires humidity change information indicating the measurement result of humidity change in the cultivation environment and carbon dioxide concentration information indicating the measurement result of carbon dioxide concentration in the cultivation environment; and a concentration control unit that controls a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired by the information acquisition unit, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired by the information acquisition unit.

[0008] To solve the above problems, a carbon dioxide concentration control method according to one aspect of the present invention is a carbon dioxide concentration control method for controlling the carbon dioxide concentration in a cultivation environment in which plants are cultivated, which is performed by one or more computers, and includes: an information acquisition step of acquiring humidity change information indicating the measurement result of humidity change in the cultivation environment and carbon dioxide concentration information indicating the measurement result of carbon dioxide concentration in the cultivation environment; and a concentration control step of controlling a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired in the information acquisition step, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired in the information acquisition step. [Effects of the Invention]

[0009] According to one aspect of the present invention, the excess carbon dioxide supplied to plants can be reduced more effectively. [Brief explanation of the drawing]

[0010] [Figure 1] This is a schematic diagram showing an overview of the conditions inside the facility. [Figure 2] This is a block diagram showing the functional configuration of a carbon dioxide concentration control system according to one embodiment of the present invention. [Figure 3] This is a flowchart showing a method for controlling carbon dioxide concentration according to one embodiment of the present invention. [Figure 4] This figure shows an example of a decision tree used in a carbon dioxide concentration control device according to one embodiment of the present invention. [Figure 5] This graph shows an example of the change in fruit stalk diameter over time and an example of the change in the rate of change in fruit stalk diameter over time. [Figure 6] This figure shows an example of an algorithm generated by a carbon dioxide concentration control device according to one embodiment of the present invention. [Figure 7] This graph shows the measurement results of the change in stomatal opening in relation to the vapor pressure deficit. [Figure 8]This graph shows the measurement results of the change in photosynthetic rate in relation to vapor pressure deficit. [Figure 9] This graph shows the measurement results of the change in net assimilation rate with respect to vapor pressure deficit. [Modes for carrying out the invention]

[0011] [1. Configuration of the carbon dioxide concentration control system] Referring to Figures 1 and 2, the configuration of a carbon dioxide concentration control system 500 according to one embodiment of the present invention will be described. The carbon dioxide concentration control system 500 is a system that controls the carbon dioxide concentration in the cultivation environment in which a plant P is being cultivated.

[0012] For example, if plant P is cultivated inside a facility, the cultivation environment in which plant P is cultivated may be the interior of that facility. Also, for example, if plant P is cultivated outdoors, the cultivation environment in which plant P is cultivated may be the surrounding area (outdoor environment) of plant P.

[0013] The facility F shown in FIG. 1 is an example of the above-described facility, and the plant P is cultivated therein. The facility F may be, for example, a closed facility or an open facility. A closed facility refers to a cultivation facility that can provide an internal environment of the facility that is partially or completely separated from the outdoor environment. However, it is more preferable that the facility F to which the present invention is applied is a closed facility. Here, as an example of the facility F, a closed facility may include, for example, openable windows, or may not include the above-described windows. Further, for example, as shown in FIG. 1, the facility F may be configured by stretching or attaching a translucent member to a frame. Specific examples of the facility F include a vinyl house, a cultivation facility configured by attaching other hard plastics or glass, etc. to a frame, etc. A plant factory, including a closed-type plant factory and a sunlight-type plant factory, etc., is also an example of a cultivation facility. When the plant P is cultivated in a cultivation facility, for example, forms such as being directly planted in a field in the facility, being planted in a cultivation container (such as a pot or a container for elevated cultivation), etc. may be mentioned, but it is not particularly limited thereto. The cultivation container is filled with, for example, soil or a soil substitute.

[0014] Further, for example, when the plant P is cultivated outdoors, that is, when the cultivation environment in which the plant P is cultivated is around the plant P, the installation locations of the sensor 400 described later and the carbon dioxide outlet from the gas supply pipe 202, which are installed in the vicinity of the plant P, are included in the range of the cultivation environment.

[0015] It is more preferable that the plant P is cultivated in a facility. That is, the cultivation environment in which the plant P is cultivated may refer to the gas inside the facility F, for example, or may refer to the gas around the plant P when the plant P is cultivated outdoors.

[0016] In the present embodiment, as a specific example, the case where the cultivation environment in which the plant P is cultivated is inside the facility F will be described.

[0017] In this embodiment, it is assumed that the plant P is a strawberry. However, the plant P is not limited to strawberries, and any plant may be used as long as it can grasp an index that affects the carbon dioxide absorption efficiency in the plant.

[0018] As an example of an index that affects the carbon dioxide absorption efficiency in the plant, stress caused by water deficiency in the plant (hereinafter, "water stress") can be cited. Specifically, in the plant, as the water stress increases, the carbon dioxide absorption efficiency decreases, and as the water stress decreases, the carbon dioxide absorption efficiency tends to increase. Therefore, as an example, it is desirable that the plant P can grasp the state of water stress.

[0019] Specific examples of the plant P include herbs, and more specifically, plants cultivated for the purpose of obtaining vegetables, flowers, or fruit trees.

[0020] As shown in FIG. 2, the carbon dioxide concentration control system 500 includes a user terminal 100, a carbon dioxide application device 200, a server 300, and a sensor 400.

[0021] <1-1. User Terminal> The user terminal 100 is an information processing device capable of executing various information processes, and is used by staff in charge of carbon dioxide concentration management in the facility F. In this embodiment, it is assumed that the user terminal 100 is an industrial computer installed in the facility F as shown in FIG. 1. However, the user terminal 100 may be, for example, a desktop personal computer or a tablet terminal, or a smartphone. As shown in FIG. 2, the user terminal 100 includes an input unit 101, a display unit 102, a storage unit 103, a communication unit 104, and a control unit 105.

[0022] The input unit 101 is an interface that accepts various operations. In this embodiment, the keyboard and mouse serve as the input unit 101. The display unit 102 is an output unit that displays various information. In this embodiment, the monitor serves as the display unit 102. If the user terminal 100 is a smartphone or tablet terminal, the user terminal 100 may have a touch panel that integrates the input unit 101 and the display unit 102.

[0023] The memory unit 103 stores various information used by the user terminal 100. Examples of memory units 103 and the memory unit 301 described later include RAM (Random Access Memory), flash memory, and hard disk. The communication unit 104 is for the user terminal 100 to send and receive various information with other devices that constitute the carbon dioxide concentration control system 500. The control unit 105 is a programmable data logger (for example, CR1000X, Campbell sci) and controls all parts of the user terminal 100. The control unit 105 also executes processing to realize the various functions of the user terminal 100. The control unit 105 may be a microcomputer such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), or programmable logic controller. As shown in Figure 2, the control unit 105 is equipped with a carbon dioxide concentration control device 10.

[0024] The carbon dioxide concentration control device 10 is a device that controls the carbon dioxide concentration in the cultivation environment in which plants are grown, and in this embodiment it is built into the control unit 105. However, the carbon dioxide concentration control device 10 may be provided outside the control unit 105, for example, within the user terminal 100, or it may be provided in the carbon dioxide application device 200 or the server 300. As shown in Figure 2, the carbon dioxide concentration control device 10 includes an information acquisition unit 11 and a concentration control unit 12.

[0025] The information acquisition unit 11 acquires humidity change information INF and carbon dioxide concentration information as shown in Figure 2. Humidity change information INF is information that shows the measurement results of humidity changes in the cultivation environment. This humidity change may be, for example, at least one of vapor pressure deficit, relative humidity, and absolute humidity. Carbon dioxide concentration information is information that shows the measurement results of carbon dioxide concentration in the cultivation environment. This carbon dioxide concentration information may be acquired together with the humidity change information INF, for example.

[0026] The concentration control unit 12 controls the carbon dioxide application device 200 based on the output result obtained by inputting the humidity, the rate of change of said humidity, and the acceleration of said humidity change, which are indicated by the humidity change information INF acquired by the information acquisition unit 11, into a prediction model, as well as the carbon dioxide concentration information acquired by the information acquisition unit 11. The rate of change of humidity is the first derivative of the time value of the humidity change indicated by the humidity change information INF acquired by the information acquisition unit 11. The acceleration of the humidity change is the second derivative of the time value of the humidity change indicated by the humidity change information INF acquired by the information acquisition unit 11.

[0027] In this embodiment, the output from the prediction model is a prediction index PI (see Figures 4 and 5; details will be described later) that indicates the predicted water stress state of plant P at the time when the humidity change in the cultivation environment is measured. However, the output is not limited to the prediction index PI, and any output is acceptable as long as it indicates the predicted water stress state of plant P at the time when the humidity change in the cultivation environment is measured.

[0028] In this embodiment, the prediction model is the decision tree 20 shown in Figure 2, which is constructed by machine learning using training data 40. The training data 40 uses the humidity, the rate of change of said humidity, and the acceleration of the change of said humidity as example data, and a reference index (details described later) that indicates the state of water stress of plant P at a predetermined time as the ground truth data.

[0029] "A predetermined time point" refers to a time point after a predetermined period has elapsed from the start of the first period. In this embodiment, the start of the first period is the time when the sensor 400 begins measuring the temperature and humidity inside facility F, prior to the point in time when the concentration control unit 12 controls the carbon dioxide application device 200 based on the prediction index PI. However, the start of the first period may be defined as the time after a certain period of time has elapsed since the sensor 400 began measuring the temperature and humidity inside facility F. In this case, the start of the first period may be defined as the time immediately before the water stress state of the plant P changes rapidly after the sensor 400 begins measuring.

[0030] Furthermore, in this embodiment, the end of the first period is defined as the point in time when the sensor 400 finishes measuring the temperature and humidity inside facility F, prior to the point in time when the concentration control unit 12 controls the carbon dioxide application device 200 based on the prediction index PI. However, if the start of the first period is defined as the point in time when a certain amount of time has elapsed since the sensor 400 started measuring the temperature and humidity inside facility F, then the end of the first period will also be pushed back by a certain amount of time.

[0031] The training data 40 consists of multiple datasets, each set at a different predetermined time, with the example data and the correct answer data used as datasets. In this embodiment, training data 40 is generated for each of the following patterns: the first pattern "Season: Autumn, Condition: No humidification during the day", the second pattern "Season: Autumn, Condition: Humidification from 8 am to 12 pm", the third and fourth patterns "Season: Spring, Condition: No humidification during the day", the fifth pattern "Season: Spring, Condition: Humidification from 8 am to 12 pm", and the sixth pattern "Season: Spring, Condition: Humidification from 10 am to 2 pm".

[0032] The prediction model is not limited to decision trees, but may also be any known learning model such as support vector machines, logistic regression, neural networks, linear discriminant analysis, nonlinear discriminant analysis, Bayesian discriminant analysis, gradient boosting, k-nearest neighbors, or k-means algorithm. Furthermore, the prediction model may be constructed using unsupervised learning.

[0033] In this embodiment, the concentration control unit 12 generates an algorithm 30 based on the first factor information INF-1, the second factor information INF-2, the prediction index PI recorded in the leaf node 23 of the decision tree 20, and a predetermined carbon dioxide concentration setting value of the cultivation environment (see Figures 5 and 6). Here, the algorithm 30 is the algorithm that determines whether or not to allow the concentration control unit 12 to apply carbon dioxide to the carbon dioxide application device 200. The concentration control unit 12 also controls the carbon dioxide application device 200 according to the algorithm 30. The first factor information INF-1 is information indicating the prediction factors of the prediction index PI, recorded in the root node 21 of the decision tree 20 (see Figure 5). The second factor information INF-2 is information indicating the prediction factors of the prediction index PI, recorded in the intermediate node 22 of the decision tree 20 (see Figure 5), and is different from the first factor information INF-1. Details of the algorithm 30 will be described later.

[0034] Furthermore, in plants, photosynthesis is promoted and therefore growth is promoted in cultivation environments with high solar radiation, such as during the day, especially on sunny days. Therefore, for example, the information acquisition unit 11 may further acquire solar radiation information. Solar radiation information is information that shows the measurement results of solar radiation in the cultivation environment. Here, the concentration control unit 12 may further control the carbon dioxide application device 200 based on the solar radiation information acquired by the information acquisition unit 11.

[0035] <1-2. Carbon Dioxide Application Equipment> The carbon dioxide application device 200 is a device that applies carbon dioxide to the cultivation environment, that is, in this embodiment, it supplies carbon dioxide into facility F. The carbon dioxide application device 200 is an example of a carbon dioxide application device according to the present invention, and carbon dioxide may be applied into facility F using a carbon dioxide application device other than the carbon dioxide application device 200. As shown in Figure 2, the carbon dioxide application device 200 comprises a carbon dioxide supply unit 201, a gas supply pipe 202, a communication unit 203, and a control unit 204. The carbon dioxide supply unit 201 and the gas supply pipe 202 are installed in facility F in the manner shown in Figure 1.

[0036] The carbon dioxide supplier 201 is a source of carbon dioxide supplied to facility F. The carbon dioxide supplied by the carbon dioxide supplier 201 may be generated by, for example, the combustion of a carbon-containing substance, or by other means. Specific examples of "carbon-containing substances" include, but are not limited to, fossil fuels such as kerosene and hydrocarbon gases. The carbon dioxide supplier 201 is also equipped with a switch (not shown) to switch the start or stop of carbon dioxide supply. The carbon dioxide supplied from the carbon dioxide supplier 201 is sent to the gas supply pipe 202 and diffused into facility F from the outlet of the gas supply pipe 202. As an example, the position of the carbon dioxide outlet may be adjusted using a duct connected to the gas supply pipe 202. The communication unit 203 is a unit for the carbon dioxide application device 200 to send and receive various information with the user terminal 100. The control unit 204 controls the aforementioned switch on the carbon dioxide supplier 201 based on the operation command transmitted from the concentration control unit 12, switching the start or stop of carbon dioxide supply. The control unit 204 may be a control panel or it may be built into the carbon dioxide supply unit 201.

[0037] <1-3. Server> The server 300 is a communication device capable of sending and receiving data with the user terminal 100. The server 300 may be wirelessly connected to the user terminal 100 or connected via a wired connection. As shown in Figure 2, the server 300 comprises a storage unit 301, a communication unit 302, and a control unit 303.

[0038] The memory unit 301 stores various information used by the user terminal 100 and the server 300, respectively. The memory unit 301 also stores the decision tree 20, the algorithm 30, and the training data 40. However, some or all of the decision tree 20, the algorithm 30, and the training data 40 may be stored in, for example, the memory unit 103. The communication unit 302 is the unit for the server 300 to send and receive various information with the user terminal 100. The control unit 303 controls all parts of the server 300. The control unit 303 also executes processing to realize the various functions of the server 300. As shown in Figure 2, the control unit 303 includes a model building unit 313.

[0039] The model building unit 313 constructs the decision tree 20 using machine learning with the training data 40. Specifically, the model building unit 313 reads the training data 40 and the decision tree 20 before it is trained from the storage unit 301 and trains the decision tree 20 using machine learning. Once the machine learning is complete, the model building unit 313 stores the trained decision tree 20 in the storage unit 301. The model building unit 313 may be located outside the control unit 303 within the server 300, or it may be located on the user terminal 100.

[0040] <1-4. Sensors> Sensor 400 is a sensor that has the functions of measuring temperature, humidity, and carbon dioxide concentration, and is used to measure humidity changes and carbon dioxide concentration within facility F. Here, sensor 400 may also have the function of measuring solar radiation, for example, and may be used to measure solar radiation within facility F. In this embodiment, as shown in Figure 1, sensor 400 is installed near the plant body P and is connected to the user terminal 100 by wire or wireless connection. Specifically, sensor 400 measures the temperature within facility F and transmits the temperature information to the user terminal 100. Sensor 400 also measures the humidity within facility F and transmits the humidity information to the user terminal 100. Sensor 400 also measures the carbon dioxide concentration within facility F and transmits the carbon dioxide concentration information to the user terminal 100. Sensor 400 also measures the solar radiation within facility F and transmits the solar radiation information to the user terminal 100.

[0041] It should be noted that the use of sensor 400 is not mandatory; any device or equipment capable of measuring the target data acquired by the information acquisition unit 11 from among temperature, humidity, carbon dioxide concentration, and solar radiation may be used. Alternatively, physically separated temperature sensors, humidity sensors, carbon dioxide concentration sensors, and solar radiation sensors may be used instead of sensor 400.

[0042] <1-5. Variations in the configuration of carbon dioxide concentration control systems> The configuration of the carbon dioxide concentration control system 500 according to this embodiment is merely an example, and various system configurations can be adopted. For example, if the decision tree 20, algorithm 30, and training data 40 are all stored in the storage unit 103, and the model construction unit 313 is built into the user terminal 100, the server 300 becomes unnecessary. Also, for example, if the carbon dioxide concentration control device 10 is built into the control unit 303, the user terminal 100 becomes unnecessary. Alternatively, either the information acquisition unit 11 or the concentration control unit 12 may be built into the carbon dioxide application device 200 or the server 300. In other words, the information acquisition unit 11 and the concentration control unit 12 only need to be provided in any of the devices or equipment that make up the carbon dioxide concentration control system 500.

[0043] [2. Processing flow of carbon dioxide concentration control device] The processing flow of the carbon dioxide concentration control device 10 will be explained with reference to Figures 3 to 6. Here, the flowchart in Figure 3 is an example of the processing flow of the carbon dioxide concentration control device 10 and is an example of the carbon dioxide concentration control method according to the present invention. Therefore, the processing flow of the carbon dioxide concentration control device 10 is not limited to the example flowchart shown in Figure 3.

[0044] First, in step S11 of the flowchart shown in Figure 3, the sensor 400 measures the temperature, humidity, and carbon dioxide concentration inside facility F. Specifically, the input unit 101 receives a measurement start operation from a staff member responsible for managing the carbon dioxide concentration inside facility F, and the carbon dioxide concentration control device 10 transmits a measurement command to the sensor 400. Upon receiving the measurement command, the sensor 400 measures the temperature, humidity, and carbon dioxide concentration inside facility F at regular time intervals until a predetermined time. The sensor 400 also transmits the temperature information, humidity information, and carbon dioxide concentration information obtained at regular time intervals to the carbon dioxide concentration control device 10. The measurement start time and measurement end time of the sensor 400, as well as the time intervals in which the sensor 400 measures the temperature, humidity, and carbon dioxide concentration inside facility F, are all arbitrary. In this embodiment, the sensor 400 measures the temperature, humidity, and carbon dioxide concentration inside facility F at 1-second intervals. Here, the sensor 400 may, for example, further measure the amount of solar radiation inside facility F, similar to the temperature, humidity, and carbon dioxide concentration inside facility F described above, and further transmit the resulting solar radiation information to the carbon dioxide concentration control device 10.

[0045] Furthermore, the sensor 400 is automated and may perform measurements regardless of whether or not the input unit 101 has received the measurement start operation. Also, it is not essential that the sensor 400 performs measurements in time intervals from the start of measurement until a predetermined time. For example, the sensor 400 may continue measuring endlessly after the start of measurement, and the carbon dioxide concentration control device 10 may send an information transmission command to the sensor 400 when carbon dioxide concentration control is required.

[0046] Next, in S12 (information acquisition step), the information acquisition unit 11 calculates the humidity change within facility F using the temperature and humidity information received from the sensor 400, thereby acquiring humidity change information INF. Specifically, as an example, if the humidity change indicated by humidity change information INF is a vapor pressure deficit, the information acquisition unit 11 subtracts the amount of humid air water vapor indicated by the humidity information from the amount of saturated water vapor calculated using the temperature information. The information acquisition unit 11 then acquires the obtained vapor pressure deficit as humidity change information INF. The information acquisition unit 11 repeats this acquisition process at 1-second intervals and transmits the humidity change information INF to the concentration control unit 12 each time it is acquired.

[0047] Next, in S13, the concentration control unit 12 calculates the humidity change rate and humidity change acceleration. Specifically, the concentration control unit 12 calculates the target humidity change rate (hereinafter, "target humidity change rate") by taking the first derivative of the humidity change (hereinafter, "target humidity change") indicated by the humidity change information INF acquired by the information acquisition unit 11 with respect to time. The concentration control unit 12 also calculates the target humidity change acceleration (hereinafter, "target humidity change acceleration") by taking the second derivative of the target humidity change with respect to time. The concentration control unit 12 repeats this calculation process at 1-second intervals.

[0048] Next, in S14 (concentration control step), the concentration control unit 12 obtains the prediction index PI. Specifically, before executing the process in S14, the concentration control unit 12 reads the decision tree 20 from the server 300. Then, the concentration control unit 12 inputs the target humidity, target humidity change rate, and target humidity change acceleration calculated in S14 into the decision tree 20, and obtains the prediction index PI output from the decision tree 20.

[0049] In this embodiment, the concentration control unit 12 uses the decision tree 20 shown in Figure 4. The decision tree 20 consists of a root node 21, a plurality of intermediate nodes 22, and a plurality of leaf nodes 23. The root node 21 records the first factor information INF-1. Below, as a specific example, we will explain the case where the humidity change indicated by the humidity change information INF is the vapor pressure deficit. In Figure 4, "VPD≧14.222" and "VPD<14.222" are predictive factors for the prediction index PI, and are recorded in the root node 21 as the first factor information INF-1. "VPD" indicates the vapor pressure deficit within facility F, i.e., the humidity change, and "VPD≧14.222" and "VPD<14.222" are numerical ranges of vapor pressure deficit used to determine which intermediate node 22 to proceed to, respectively.

[0050] Intermediate node 22 records the second factor information INF-2. In Figure 4, as an example, "ddVPD≧-0.016" and "ddVPD<-0.016" are predictive factors for the predictive index PI and are recorded as the second factor information INF-2 at a specific intermediate node 22. "ddVPD" represents the rate of change in the vapor pressure deficit within facility F, i.e., the rate of change in humidity, and "ddVPD≧-0.016" and "ddVPD<-0.016" are the numerical ranges of the vapor pressure deficit change rate used to determine which intermediate node 22 to proceed to. Here, "dVPD" in Figure 4 represents the rate of change in the vapor pressure deficit within facility F, i.e., the rate of change in humidity.

[0051] In Figure 4, each of the multiple leaf nodes 23 records the cumulative result of the prediction metric PI in the decision tree 20 at 1-second intervals from the start to the end of the second period (details described later). Specifically, the decision tree 20 counts the number of prediction metric PI "0" and "1" recorded in the predetermined leaf node 23 when all predetermined first factor information INF-1 and predetermined second factor information INF-2 are satisfied from the start to the end of the second period. Then, at the end of the second period, the decision tree 20 calculates the proportion of prediction metric PI "0" to the total number of prediction metric PI "0" and "1" recorded in the predetermined leaf node 23 by dividing the number of prediction metric PI "0" by the total number of prediction metric PI "0" and "1". Furthermore, at the end of the second period, the decision tree 20 calculates the proportion of the prediction indicator PI "1" recorded in a predetermined leaf node 23 to the total number of prediction indicators PI "0" and "1" by dividing the number of prediction indicators PI "1" by the total number of prediction indicators PI "0" and "1".

[0052] For example, decision tree 20 records the predicted indicators PI "0" and "1" at the leaf node 23 of "Node 4" when all of the first factor information INF-1 "VPD < 14.222" and the second factor information INF-2 "VPD < 6.99 and dVPD < 0.056" are satisfied. In the example in Figure 4, the total number of predicted indicators PI "0" and "1" recorded at the leaf node 23 of "Node 4" from the start to the end of the second period is n = 1365. Also, at the end of the second period, the proportion of predicted indicators PI "0" recorded at the leaf node 23 of "Node 4" to the total number of predicted indicators PI "0" and "1" is approximately 0.98. Furthermore, at the end of the second period, the proportion of predicted indicators PI "1" recorded at the leaf node 23 of "Node 4" to the total number of predicted indicators PI "0" and "1" is approximately 0.02. The concentration control unit 12 obtains the cumulative results of the prediction index PI from such leaf nodes 23.

[0053] Note that the conditions and numerical values ​​shown for the first factor information INF-1, the second factor information INF-2, and the prediction index PI in Figure 4 are merely examples and can be appropriately changed depending on the type of plant P, the conditions within facility F, the timing of carbon dioxide concentration control, etc. Also, the number of intermediate nodes 22 and the number of leaf nodes 23 in Figure 4 are merely examples and can be appropriately changed depending on the type of plant P, the conditions within facility F, the timing of carbon dioxide concentration control, etc.

[0054] Furthermore, in this embodiment, the reference index, which is the correct answer data of the training data 40, is an index based on the first time derivative of the standard change in fruit stalk diameter, obtained by subtracting the fruit stalk diameter of plant P at the start of the first period from the fruit stalk diameter of plant P at a predetermined point in time, and then dividing this value by the fruit stalk diameter of plant P at the start of the first period. Specifically, the index output during the daytime, before the fruit stalk diameter of plant P begins to shrink rapidly, in other words, the index output during the time period when both the standard change in fruit stalk diameter and the first time derivative of the standard change in fruit stalk diameter take positive values, is defined as reference index "0". On the other hand, the index output during the time period from when the first time derivative of the standard change in fruit stalk diameter begins to decrease and becomes negative until it reaches its lowest value is defined as reference index "1" (see Figure 5 for details).

[0055] Furthermore, in this embodiment, the prediction index PI is an index based on at least the first time derivative of the fruit stalk diameter change, obtained by subtracting the fruit stalk diameter of plant P at the start of the second period from the fruit stalk diameter of plant P at the time when the humidity change in facility F is measured, and then dividing this value by the fruit stalk diameter of plant P at the start of the second period. Specifically, as shown in the example in Figure 5, the prediction index PI "0" is defined as the index output during the time before the fruit stalk diameter of plant P rapidly shrinks during the day, in other words, the time period when both the fruit stalk diameter change and the first time derivative of the fruit stalk diameter change are positive. On the other hand, the prediction index PI "1" is defined as the index output during the time period from when the first time derivative of the fruit stalk diameter change begins to decrease until it reaches its lowest value. Here, the time period in Figure 5 when the first time derivative of the fruit stalk diameter change is increasing corresponds to the time period when plant P is not subjected to water stress. On the other hand, the period from when the first derivative of the change in fruit stalk diameter in the same figure begins to decrease and becomes negative until it reaches its lowest value corresponds to the period during which water stress is applied to the plant P.

[0056] The second period is a period that follows the first period. In this embodiment, when the concentration control unit 12 controls the carbon dioxide application device 200 based on the prediction index PI, the start of the second period is the time when the sensor 400 starts measuring the temperature and humidity inside the facility F (hereinafter referred to as the "measurement start time"). However, the start of the second period may be set as a time after a certain period of time has elapsed from the measurement start time. In this case, it is conceivable that the start of the second period be set as the time immediately before the water stress state of the plant P changes rapidly after the sensor 400 starts measuring.

[0057] Furthermore, in this embodiment, when the concentration control unit 12 controls the carbon dioxide application device 200 based on the prediction index PI, the end of the second period is determined to be the time when the sensor 400 finishes measuring the temperature and humidity inside the facility F (hereinafter referred to as the "measurement end time"). However, if the start of the second period is set to a time after a certain period of time has elapsed from the measurement end time, the end of the second period will also be pushed back by a certain amount of time.

[0058] In this embodiment, the start and end times of measurement are predetermined, and the concentration control unit 12 repeatedly controls the carbon dioxide application device 200 within this predetermined time period. For example, the start time of measurement may be any point in September, which is a common time for planting plants P (strawberries). In this case, the end time of measurement may be any point in May of the following year, which is a common time for finishing the cultivation of plants P (strawberries). However, the start and end times of measurement are not limited to the examples in this embodiment. For example, the concentration control unit 12 may adopt a control method in which, once one carbon dioxide supply by the carbon dioxide application device 200 is completed, one second period is completed and the system moves on to the next second period. If this control method is adopted, the start time of measurement may be the point after the minimum stop time of the carbon dioxide application device 200 has elapsed since the end of the previous control of the carbon dioxide application device 200, and the end time of measurement may be the point after the end of the current control of the carbon dioxide application device 200.

[0059] Furthermore, the decision tree 20 may output at least one of the baseline indicator and the forecast indicator PI using letters or symbols. In other words, there are no limitations on the output methods for the baseline indicator and the forecast indicator PI.

[0060] The concentration control unit 12 repeats the aforementioned series of processes at 1-second intervals from the start to the end of the second period. The concentration control unit 12 may also temporarily store the decision tree 20, once read, in the memory (not shown) of the carbon dioxide concentration control device 10 until all of the aforementioned series of processes are completed.

[0061] Next, in S15 (concentration control step), the concentration control unit 12 generates the algorithm 30 shown in Figure 6 based on the first factor information INF-1, the second factor information INF-2, the prediction index PI, and a predetermined carbon dioxide concentration setting value for the cultivation environment. Specifically, the concentration control unit 12 repeats the process of S14 from the start to the end of the second period to find a regularity in the relationship between (i) the prediction factors of the first factor information INF-1 where the target humidity change is satisfied, and the prediction factors of each second factor information INF-2 where the target humidity, target humidity change rate, and target humidity change acceleration are satisfied, and (ii) the cumulative result of the prediction index PI recorded in the leaf node 23. Then, the concentration control unit 12 generates the algorithm 30 in accordance with the regularity it has found.

[0062] The predetermined carbon dioxide concentration setting value for the cultivation environment is a value used to set the carbon dioxide concentration in that cultivation environment. In other words, the predetermined carbon dioxide concentration setting value for the cultivation environment is a value used to set a numerical target for the carbon dioxide concentration in that cultivation environment. The predetermined carbon dioxide concentration setting value includes at least one of a predetermined upper limit and a predetermined lower limit. The predetermined upper limit is the upper limit set for the numerical range to which the numerical target for the carbon dioxide concentration in the cultivation environment can take. For example, the predetermined upper limit is the upper limit of carbon dioxide concentration at which no surplus occurs even when the carbon dioxide absorption efficiency of the plant is at its maximum. The predetermined lower limit is the lower limit set for the numerical range to which the numerical target for the carbon dioxide concentration in the cultivation environment can take. For example, the predetermined lower limit is the lower limit of carbon dioxide concentration at which carbon dioxide deficiency does not occur in the plant. Here, the predetermined carbon dioxide concentration setting value can be arbitrarily set according to the type of plant P, various conditions of the cultivation environment, the timing of carbon dioxide concentration control, etc. The predetermined carbon dioxide concentration setting value is stored in advance in the memory of the concentration control unit 12, for example.

[0063] Furthermore, as mentioned above, in plants, the efficiency of carbon dioxide absorption tends to decrease as water stress increases, and to increase as water stress decreases. Therefore, the concentration control unit 12, in conjunction with the regularity discovered as described above, generates an algorithm 30 such that, when a decrease in water stress is predicted in the plant P, i.e., when an increase in carbon dioxide absorption efficiency is predicted, carbon dioxide is applied to an extent that does not exceed a predetermined set upper limit, and when an increase in water stress is predicted, i.e., when a decrease in carbon dioxide absorption efficiency is predicted, carbon dioxide is applied to a minimum extent that does not fall below a predetermined set lower limit.

[0064] In this embodiment, as shown in Figure 5, the predictive factor of the first factor information INF-1 is the numerical range of humidity change within facility F, and the predictive factors of the second factor information INF-2 are the numerical ranges of the humidity, the rate of change of the humidity, and the acceleration of the change of the humidity. Therefore, the algorithm 30 generated by the concentration control unit 12 includes the numerical range of humidity in the cultivation environment, the numerical range of the rate of change of the humidity, and the numerical range of the acceleration of the change of the humidity. Furthermore, the numerical range of carbon dioxide concentration in the cultivation environment is indicated by at least one of a predetermined upper setting value and a predetermined lower setting value included in a predetermined carbon dioxide concentration setting value. Therefore, the algorithm 30 further includes the numerical range of carbon dioxide concentration in the cultivation environment. These numerical ranges serve as conditions for the concentration control unit 12 to decide whether or not to apply carbon dioxide to the carbon dioxide application device 200. In Figure 6, for example, the first numerical range 31 represents the numerical range of the vapor pressure deficit within facility F, the second numerical range 32 represents the numerical range of the rate of change of the vapor pressure deficit, the third numerical range 33 represents the numerical range of the acceleration of change of the vapor pressure deficit, and the numerical range defined by the lower limit 34 and upper limit 35 of the carbon dioxide concentration represents the numerical range of the carbon dioxide concentration within facility F.

[0065] The concentration control unit 12 may, for example, generate the algorithm 30 shown in Figure 6 based on a predetermined solar radiation setting value of the cultivation environment. The predetermined solar radiation setting value is a value used to determine the magnitude of solar radiation in the cultivation environment. The predetermined solar radiation setting value can be arbitrarily set according to the type of plant P, various conditions of the cultivation environment, the timing of carbon dioxide concentration control, etc. The predetermined solar radiation setting value is stored in advance in the memory of the concentration control unit 12, for example. The concentration control unit 12 may, for example, generate the algorithm 30 so that carbon dioxide is applied only when the amount of solar radiation indicated by the solar radiation information acquired by the information acquisition unit 11 is equal to or greater than the predetermined solar radiation setting value.

[0066] Next, in S16 (concentration control step), the concentration control unit 12 controls the carbon dioxide application device 200 using algorithm 30. The control flow of the carbon dioxide application device 200 by algorithm 30 will be explained below with reference to Figure 6.

[0067] First, in S211, the concentration control unit 12 determines whether or not it is within the set time period. The set time period is arbitrary, but it may be set to the daytime hours when the plant body P is susceptible to water stress, for example, from 6:00 to 18:00. The set time period is pre-stored in the memory of the concentration control unit 12, for example. If the result in S211 is No, the concentration control unit 12 executes the process in S211 again. If the result in S211 is Yes, the concentration control unit 12 proceeds to the process in S212.

[0068] Next, in S212, the concentration control unit 12 obtains information regarding the operating status of the carbon dioxide supply unit 201 from the control unit 204 and determines whether the carbon dioxide application device 200 is stopped or not. If the result in S212 is No, the concentration control unit 12 executes the process in S211 again. If the result in S212 is Yes, the concentration control unit 12 proceeds to the process in S213.

[0069] Next, in S213, the concentration control unit 12 determines whether the amount of solar radiation indicated by the solar radiation information acquired by the information acquisition unit 11 is equal to or greater than a predetermined solar radiation value. If the result in S213 is No, the concentration control unit 12 executes the process in S211 again. If the result in S213 is Yes, the concentration control unit 12 proceeds to the process in S214.

[0070] Next, in S214, the concentration control unit 12 determines whether the carbon dioxide concentration indicated by the carbon dioxide concentration information acquired by the information acquisition unit 11 is equal to or greater than a predetermined set upper limit. If the result in S214 is No, the concentration control unit 12 executes the process in S211 again. If the result in S214 is Yes, the concentration control unit 12 proceeds to the process in S215.

[0071] Next, in S215, the concentration control unit 12 determines whether the target vapor pressure deficit falls within the numerical range of "VPD≧14.22". If the answer is Yes in S214, the concentration control unit 12 proceeds to the process in S216. On the other hand, if the answer is No in S215, the concentration control unit 12 proceeds to the process in S218.

[0072] Next, in S216, the concentration control unit 12 determines whether the carbon dioxide concentration indicated by the carbon dioxide concentration information acquired by the information acquisition unit 11 is below a predetermined set lower limit. If the answer in S216 is Yes, the concentration control unit 12 sends an operation command for the carbon dioxide supplyer 201 to the control unit 204, causing the carbon dioxide application device 200 to apply carbon dioxide until the carbon dioxide concentration reaches a predetermined set lower limit or until a set time has elapsed (S217). The set time can be arbitrarily set according to the type of plant P, the conditions inside the facility F, the timing of carbon dioxide concentration control, etc. The set time is stored in advance in the memory of the concentration control unit 12, for example. After the processing in S217 is completed, the concentration control unit 12 executes the processing in S211 again. On the other hand, if the answer in S216 is No, the concentration control unit 12 executes the processing in S211 again.

[0073] Next, in S218, the concentration control unit 12 determines whether the target vapor pressure deficit is included within the numerical range of "VPD < 6.99" and whether the rate of change of the target vapor pressure deficit is included within the numerical range of "dVPD ≥ 0.056". If the answer in S218 is Yes, the concentration control unit 12 proceeds to the process in S219. On the other hand, if the answer in S218 is No, the concentration control unit 12 proceeds to the process in S221.

[0074] Next, in S219, the concentration control unit 12 determines whether the carbon dioxide concentration indicated by the carbon dioxide concentration information acquired by the information acquisition unit 11 is below a predetermined set lower limit. If the answer in S219 is Yes, the concentration control unit 12 causes the carbon dioxide application device 200 to apply carbon dioxide until the carbon dioxide concentration reaches the predetermined set lower limit or until a set time has elapsed (S220). After the processing in S220 is completed, the concentration control unit 12 executes the process in S211 again. On the other hand, if the answer in S219 is No, the concentration control unit 12 executes the process in S211 again.

[0075] Next, in S221, the concentration control unit 12 determines whether the target vapor pressure deficit is included within the numerical range of "6.99 ≤ VPD < 10.56", the rate of change of the target vapor pressure deficit is included within the numerical range of "dVPD ≥ 0.018", and the acceleration of change of the target vapor pressure deficit is included within the numerical range of "ddVPD ≥ 0.001". If the answer in S221 is Yes, the concentration control unit 12 proceeds to the process in S222. On the other hand, if the answer in S221 is No, the concentration control unit 12 proceeds to the process in S224.

[0076] Next, in S222, the concentration control unit 12 determines whether the carbon dioxide concentration indicated by the carbon dioxide concentration information acquired by the information acquisition unit 11 is below a predetermined set lower limit. If the answer in S222 is Yes, the concentration control unit 12 causes the carbon dioxide application device 200 to apply carbon dioxide until the carbon dioxide concentration reaches the predetermined set lower limit or until a set time has elapsed (S223). After the processing in S223 is completed, the concentration control unit 12 executes the process in S211 again. On the other hand, if the answer in S222 is No, the concentration control unit 12 executes the process in S211 again.

[0077] Next, in S224, the concentration control unit 12 determines whether the target vapor pressure deficit is included within the numerical range of "10.56 ≤ VPD < 14.22", the rate of change of the target vapor pressure deficit is included within the numerical range of "dVPD ≥ 0.018", and the acceleration of change of the target vapor pressure deficit is included within the numerical range of "ddVPD ≥ -0.016". If the answer in S224 is Yes, the concentration control unit 12 proceeds to the process in S225. On the other hand, if the answer in S224 is No, the concentration control unit 12 causes the carbon dioxide application device 200 to apply carbon dioxide until the carbon dioxide concentration reaches a predetermined set upper limit or until a set time has elapsed (S227). After the process in S227 is completed, the concentration control unit 12 executes the process in S211 again.

[0078] Next, in S225, the concentration control unit 12 determines whether the carbon dioxide concentration indicated by the carbon dioxide concentration information acquired by the information acquisition unit 11 is below a predetermined set lower limit. If the answer in S225 is Yes, the concentration control unit 12 causes the carbon dioxide application device 200 to apply carbon dioxide until the carbon dioxide concentration reaches the predetermined set lower limit or until a set time has elapsed (S226). After the processing in S226 is completed, the concentration control unit 12 executes the process in S211 again. On the other hand, if the answer in S225 is No, the concentration control unit 12 executes the process in S211 again.

[0079] If the control of the carbon dioxide application device 200 by algorithm 30 is to be continued, the concentration control unit 12 repeatedly executes the series of processes from S211 to S227. On the other hand, if the control of the carbon dioxide application device 200 by algorithm 30 is terminated, the concentration control unit 12 sends the generated algorithm 30 to the server 300 and stores it in the storage unit 301. Note that the conditions and numerical values ​​shown in the flowchart of Figure 6 are merely examples and can be appropriately changed depending on the type of plant P, the conditions inside the facility F, the timing of carbon dioxide concentration control, etc.

[0080] The concentration control unit 12 terminates its control of the carbon dioxide application device 200 according to algorithm 30, and the process in S16 ends. With the completion of the process in S16, the series of processes executed by the carbon dioxide concentration control device 10 ends.

[0081] [3. Variant Example] First, various indicators can be used as the reference indicator and the predictive indicator PI. For example, the indicator output during the time period when the reference rate of change in fruit stalk diameter is positive may be designated as the reference indicator "0", and the indicator output during the time period when the reference rate of change in fruit stalk diameter is negative may be designated as the reference indicator "1". In this case, the indicator output during the time period when the rate of change in fruit stalk diameter is positive will be the predictive indicator PI "0", and the indicator output during the time period when the rate of change in fruit stalk diameter is negative will be the predictive indicator PI "1".

[0082] Alternatively, for example, the index output during the time period when the reference rate of change in pedicel diameter is positive may be designated as reference index "0," and the index output during the time period when both the reference rate of change in pedicel diameter and the reference acceleration of change in pedicel diameter are negative may be designated as reference index "1." In this case, the index output during the time period when the rate of change in pedicel diameter is positive will be the prediction index PI "0," and the index output during the time period when both the rate of change in pedicel diameter and the acceleration of change in pedicel diameter are negative will be the prediction index PI "1."

[0083] Alternatively, for example, the index output during the time period when both the standard change in stem diameter and the standard rate of change in stem diameter are positive values ​​may be designated as the standard index "0", and the index output during the time period from when both the standard change in stem diameter and the standard rate of change in stem diameter begin to decrease until the standard rate of change in stem diameter goes from less than 0 to its lowest value may be designated as the standard index "1". In this case, the index output during the time period when both the change in stem diameter and the standard rate of change in stem diameter are positive values ​​may be designated as the prediction index PI "0", and the index output during the time period from when both the change in stem diameter and the standard rate of change in stem diameter begin to decrease until the standard rate of change in stem diameter goes from less than 0 to its lowest value may be designated as the prediction index PI "1".

[0084] Furthermore, the features of plant P used as the basis for outputting the reference index and the prediction index PI are not limited to the fruit stalk diameter of plant P. For example, one or more of the following diameters of herbaceous plants (stem diameter, petiole diameter, leaf thickness, fruit diameter) or tree trunk circumference, branch diameter, and fruit diameter, or stomatal opening may be used as the aforementioned features. Any features that can capture the water stress state of plant P may be used as the aforementioned features.

[0085] To summarize the above, the reference index should be based on at least the first time derivative of the arbitrary diameter reference change, obtained by subtracting the diameter of any part of plant P at the start of the first period from the diameter of any part of plant P at a predetermined point in time, and then dividing that value by the diameter of any part of plant P at the start of the first period. The prediction index PI should be based on at least the first time derivative of the arbitrary diameter change, obtained by subtracting the diameter of any part of plant P at the start of the second period from the diameter of any part of plant P at the time when the humidity change in the cultivation environment is measured, and then dividing that value by the diameter of any part of plant P at the start of the second period.

[0086] Furthermore, the state of water stress in plant P may be ascertained, for example, by the stomatal opening of any part of the herbaceous plant. That is, for example, the stomatal opening of any part of plant P may be used as the aforementioned characteristic quantity. Note that as water stress in plant P intensifies, stomatal opening tends to decrease, and carbon dioxide uptake by plant P decreases. In other words, the stomatal opening at which water stress in plant P intensifies is the stomatal opening that limits carbon dioxide uptake by plant P.

[0087] To summarize the above, the reference index should be based on at least the first time derivative of the arbitrary stomatal opening standard change, obtained by subtracting the stomatal opening of an arbitrary part of plant P at the start of the first period from the stomatal opening of an arbitrary part of plant P at a predetermined time point, and then dividing this value by the stomatal opening of an arbitrary part of plant P at the start of the first period. The prediction index PI should be based on at least the first time derivative of the arbitrary stomatal opening change, obtained by subtracting the stomatal opening of an arbitrary part of plant P at the start of the second period from the stomatal opening of an arbitrary part of plant P at the time when the humidity change in the cultivation environment is measured, and then dividing this value by the stomatal opening of an arbitrary part of plant P at the start of the second period.

[0088] Next, the concentration control unit 12 may control the carbon dioxide application device 200 without generating the algorithm 30. For example, a threshold value may be set in advance as a comparison target for the prediction index PI, and when the prediction index PI exceeds the threshold value, the concentration control unit 12 may send an operation command for the carbon dioxide supply device 201 to the control unit 204, thereby causing the concentration control unit 12 to apply carbon dioxide to the carbon dioxide application device 200.

[0089] Furthermore, the state of water stress in plant P may be determined, for example, by the amount of photosynthesis in any part of the herbaceous plant. That is, for example, the amount of photosynthesis in any part of plant P may be used as the aforementioned characteristic quantity. Note that the amount of photosynthesis in any part of plant P refers to the amount of carbon dioxide taken up by that part of plant P.

[0090] To summarize the above, the reference index should be based on at least the first time derivative of the arbitrary photosynthesis reference change, obtained by subtracting the amount of photosynthesis of any part of plant P at the start of the first period from the amount of photosynthesis of any part of plant P at a predetermined time point, and then dividing that value by the amount of photosynthesis of any part of plant P at the start of the first period. The prediction index PI should be based on at least the first time derivative of the arbitrary photosynthesis change, obtained by subtracting the amount of photosynthesis of any part of plant P at the start of the second period from the amount of photosynthesis of any part of plant P at the time when the humidity change of the cultivation environment is measured, and then dividing that value by the amount of photosynthesis of any part of plant P at the start of the second period.

[0091] Furthermore, the state of water stress in plant P may be assessed, for example, by the amount of net assimilation in any part of a herbaceous plant. That is, for example, the amount of net assimilation in any part of plant P may be used as the aforementioned characteristic quantity. Alternatively, for example, the amount of net assimilation per unit area in the aforementioned arbitrary part, i.e., the net assimilation rate in the aforementioned arbitrary part, may be used as the amount of net assimilation. Note that as water stress in a plant intensifies, carbon dioxide uptake by plant P decreases, and the amount of net assimilation tends to decline.

[0092] To summarize the above, the reference index should be based on at least the first time derivative of the arbitrary pure assimilation standard change, obtained by subtracting the amount of net assimilation of any part of plant P at the start of the first period from the amount of net assimilation of any part of plant P at a predetermined point in time, and then dividing that value by the amount of net assimilation of any part of plant P at the start of the first period. The prediction index PI should be based on at least the first time derivative of the arbitrary pure assimilation change, obtained by subtracting the amount of net assimilation of any part of plant P at the start of the second period from the amount of net assimilation of any part of plant P at the time when the humidity change of the cultivation environment is measured, and then dividing that value by the amount of net assimilation of any part of plant P at the start of the second period.

[0093] [4. Examples of implementation using software] The function of the carbon dioxide concentration control device 10 (hereinafter referred to as "the device") can be realized by a program that causes one or more computers to function as the device, and by a program that causes one or more computers to function as each control block of the device (particularly each part included in the control unit 105).

[0094] In this case, the device includes a computer having at least one carbon dioxide concentration control device (e.g., a processor) and at least one storage device (e.g., memory) as hardware for executing the program. By executing the program using this carbon dioxide concentration control device and storage device, the functions described in the above embodiment and the above modifications are realized.

[0095] The program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the device. In the latter case, the program may be supplied to the device via any wired or wireless transmission medium.

[0096] Furthermore, some or all of the functions of each control block can also be implemented by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the control blocks are formed is also included in the scope of the present invention. In addition, it is also possible to implement the functions of each control block using, for example, a quantum computer.

[0097] [5. Additional Notes] The present invention is not limited to the embodiments and modifications described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention.

[0098] A carbon dioxide concentration control device according to Embodiment 1 of the present invention is a carbon dioxide concentration control device for controlling the carbon dioxide concentration in a cultivation environment in which plants are being cultivated, comprising: an information acquisition unit that acquires humidity change information indicating the measurement result of humidity change in the cultivation environment and carbon dioxide concentration information indicating the measurement result of carbon dioxide concentration in the cultivation environment; and a concentration control unit that controls a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired by the information acquisition unit, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired by the information acquisition unit.

[0099] The three parameters—humidity, the first time derivative of the humidity change, and the second time derivative of the humidity change—interact with each other and influence the carbon dioxide absorption efficiency of the plant. According to the above configuration, the carbon dioxide application device is controlled based on the output results obtained from the prediction model and the carbon dioxide concentration in the cultivation environment. This output result has higher prediction accuracy compared, for example, to the output results obtained from a rule-based algorithm and the results of comparing it with predetermined set values. Therefore, the concentration control unit can control the carbon dioxide application device to apply only the amount of carbon dioxide needed in the cultivation environment, according to the carbon dioxide absorption efficiency of the plant. This allows for a more effective reduction of excess carbon dioxide supplied to the plant.

[0100] In the carbon dioxide concentration control device according to aspect 2 of the present invention, in aspect 1 above, the prediction model is constructed by machine learning using training data that uses the humidity indicated by the humidity change information at a predetermined time, the first time derivative of the humidity change, and the second time derivative of the humidity change as example data, and a reference index indicating the state of water stress of the plant at the predetermined time as the correct answer data, and the output result is a prediction index indicating the predicted state of water stress of the plant at the time when the humidity change of the cultivation environment is measured.

[0101] As mentioned above, the state of water stress in plants affects the efficiency of carbon dioxide absorption in plants. With the above configuration, the concentration control unit can control the carbon dioxide application device according to the state of water stress in plants, i.e., the carbon dioxide absorption efficiency in plants, by using a predictive index as a control parameter for the carbon dioxide application device. This makes it possible to more effectively reduce the excess carbon dioxide supplied to plants.

[0102] The carbon dioxide concentration control device according to embodiment 3 of the present invention, in embodiment 2 above, is an index based on at least the first time derivative of the arbitrary diameter standard change, obtained by subtracting the diameter of the arbitrary part of the plant at the start of the first period from the diameter of the arbitrary part of the plant at a predetermined time, and dividing the value obtained by dividing the value obtained by dividing the value obtained by dividing the value obtained by dividing the value obtained by dividing the value obtained by dividing the value of the arbitrary part of the plant at the start of the first period by the diameter of the arbitrary part of the plant at the start of the second period, which is a period later than the first period, by the diameter of the arbitrary part of the plant at the start of the second period, and is an index based on at least the first time derivative of the arbitrary diameter change.

[0103] According to the above configuration, the concentration control unit can control the carbon dioxide application device according to the carbon dioxide absorption efficiency of the plant by using a predictive index based on at least the first time derivative of the change in an arbitrary diameter as a control parameter for the carbon dioxide application device. This makes it possible to more effectively reduce the excess carbon dioxide supplied to the plant.

[0104] In the carbon dioxide concentration control device according to embodiment 4 of the present invention, in embodiment 2 or 3 above, the prediction model is a decision tree.

[0105] With the above configuration, the concentration control unit can control the carbon dioxide application device according to the carbon dioxide absorption efficiency of the plant by using the prediction index output from the decision tree as a control parameter for the carbon dioxide application device. This makes it possible to more effectively reduce the excess carbon dioxide supplied to the plant.

[0106] A carbon dioxide concentration control device according to aspect 5 of the present invention, in aspect 4 above, the concentration control unit generates an algorithm for determining whether or not to allow the carbon dioxide application device to apply carbon dioxide based on first factor information indicating the predictive factors of the predictive index recorded in the root node of the decision tree, second factor information different from the first factor information recorded in the intermediate node of the decision tree, the predictive index recorded in the leaf node of the decision tree, and a predetermined carbon dioxide concentration setting value of the cultivation environment, and controls the carbon dioxide application device by the algorithm, wherein the algorithm includes, as conditions for the concentration control unit to determine whether or not to allow the carbon dioxide application device to apply carbon dioxide, a numerical range for the humidity of the cultivation environment, a numerical range for the first time derivative of the change in humidity of the cultivation environment, a numerical range for the second time derivative of the change in humidity of the cultivation environment, and a numerical range for the carbon dioxide concentration of the cultivation environment.

[0107] With the above configuration, the concentration control unit not only predicts the carbon dioxide absorption efficiency of the plant at the time the measurement unit measures humidity changes using a decision tree, but also controls the carbon dioxide application device using an algorithm based on various information held by each node of the decision tree. As a result, the carbon dioxide concentration in the cultivation environment can be controlled more effectively than when the control of the carbon dioxide application device is based solely on predictive indicators, and consequently, the excess carbon dioxide supplied to the plant can be reduced more effectively using the decision tree.

[0108] In the carbon dioxide concentration control device according to embodiment 6 of the present invention, in any of embodiments 1 to 5 above, the humidity change is at least one of vapor pressure deficit, relative humidity, and absolute humidity.

[0109] According to the above configuration, the concentration control unit can control the carbon dioxide application device according to the carbon dioxide absorption efficiency of the plant by using a predictive index based on at least one of vapor pressure deficit, relative humidity, and absolute humidity as a control parameter for the carbon dioxide application device. This makes it possible to more effectively reduce the excess carbon dioxide supplied to the plant.

[0110] In the carbon dioxide concentration control device according to embodiment 7 of the present invention, in any of embodiments 1 to 6 above, the information acquisition unit further acquires solar radiation information indicating the measurement result of the solar radiation of the cultivation environment, and the concentration control unit controls the carbon dioxide application device based on the solar radiation information acquired by the information acquisition unit.

[0111] According to the above configuration, by applying carbon dioxide only in cultivation environments with sufficient sunlight to promote plant growth, the excess carbon dioxide supplied to plants can be reduced more effectively.

[0112] A program according to aspect 8 of the present invention is a program for causing a computer to function as a carbon dioxide concentration control device according to any of aspects 1 to 7 above, wherein the computer functions as the information acquisition unit and the concentration control unit.

[0113] The above configuration produces the same effects as those in embodiments 1 to 7.

[0114] A carbon dioxide concentration control system according to aspect 9 of the present invention is a carbon dioxide concentration control system for controlling the carbon dioxide concentration in a cultivation environment in which plants are being cultivated, comprising: an information acquisition unit that acquires humidity change information indicating the measurement result of humidity change in the cultivation environment and carbon dioxide concentration information indicating the measurement result of carbon dioxide concentration in the cultivation environment; and a concentration control unit that controls a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired by the information acquisition unit, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired by the information acquisition unit.

[0115] The above configuration produces the same effects as in the first embodiment.

[0116] A carbon dioxide concentration control method according to embodiment 10 of the present invention is a carbon dioxide concentration control method for controlling the carbon dioxide concentration of a cultivation environment in which plants are being cultivated, which is performed by one or more computers, and includes: an information acquisition step of acquiring humidity change information indicating the measurement result of humidity change in the cultivation environment and carbon dioxide concentration information indicating the measurement result of carbon dioxide concentration in the cultivation environment; and a concentration control step of controlling a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired in the information acquisition step, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired in the information acquisition step.

[0117] The above configuration produces the same effects as in the first embodiment. [Examples]

[0118] One embodiment of the present invention is described below.

[0119] In this embodiment, among the water stress conditions exemplified in the embodiment, the changes in stomatal opening and photosynthetic rate characteristics in response to changes in humidity were measured.

[0120] In this example, the vapor pressure deficit (VPD) was used as an example of humidity change.

[0121] Furthermore, in this embodiment, stomatal conductance was measured as an example of a characteristic quantity related to stomatal opening. In other words, in this embodiment, the change in stomatal conductance with respect to VPD was measured.

[0122] Furthermore, in this embodiment, the rate of photosynthesis of individual leaves was measured as an example of a characteristic quantity related to the amount of photosynthesis. In other words, in this embodiment, the change in the rate of photosynthesis of individual leaves in relation to VPD was measured.

[0123] In this embodiment, the changes in stomatal conductance and individual leaf photosynthesis rate were measured under two different carbon dioxide concentrations. Specifically, the carbon dioxide concentration around the leaves was measured under two conditions: 800 μmol / mol and 400 μmol / mol.

[0124] In this example, strawberries (a single-season variety called "Koiminori") were used as the subject of measurement.

[0125] Figure 7 is a graph showing the results of measuring the change in stomatal conductance of strawberries in relation to VPD. Figure 8 is a graph showing the results of measuring the change in individual leaf photosynthetic rate of strawberries in relation to VPD. In Figures 7 and 8, the horizontal axis represents the VPD value. In Figure 7, the vertical axis represents the stomatal conductance value. In Figure 8, the vertical axis represents the individual leaf photosynthetic rate. In Figures 7 and 8, the carbon dioxide concentration around the leaves is shown as a white circle plot for measurements at 800 μmol / mol and as a black triangle plot for measurements at 400 μmol / mol. The measurement results for each plot show the mean value under a sample size of n=3, and the range of the standard deviation corresponding to each plot is also shown.

[0126] As shown in Figure 7, stomatal conductance generally decreased as VPD increased under all carbon dioxide concentrations. Furthermore, as shown in Figure 8, photosynthetic rate decreased as VPD increased under all carbon dioxide concentrations. [Examples]

[0127] Other embodiments of the present invention are described below.

[0128] In this embodiment, among the water stress conditions exemplified in the embodiment, the change in the feature quantity related to the amount of net assimilation in response to changes in humidity was measured.

[0129] In this example, the daily average vapor pressure deficit (daily average VPD) was used as an example of humidity change.

[0130] Furthermore, in this embodiment, the pure assimilation rate was measured as an example of a feature related to the amount of pure assimilation. In other words, in this embodiment, the change in the pure assimilation rate with respect to VPD was measured.

[0131] In this example, strawberries (a single-season variety called "Kaorino") were used as the subject of measurement, and the measurement period was from September 12th to September 25th, 2024. Furthermore, in this example, measurements were taken of strawberries grown under both a greenhouse without humidification and a greenhouse with humidification during the aforementioned measurement period.

[0132] Figure 9 shows the results of measuring the change in the net assimilation rate of strawberries relative to VPD. Note that Figure 9 uses two graphs, Graph A and Graph B, to show the measurement results.

[0133] Graph A on the left in Figure 9 shows the measurement results of the daily average vapor pressure deficit (daily average VPD) during the measurement period for both greenhouses without humidification and greenhouses with humidification. The horizontal axis in Graph A represents the date during the measurement period. The vertical axis in Graph A represents the daily average VPD.

[0134] Graph B on the right in Figure 9 shows the measured net assimilation rate of strawberries grown in greenhouses with and without humidification during the measurement period. Graph B shows the net assimilation rate for each of the greenhouses with and without humidification. The net assimilation rates shown in Graph B are all average values ​​for a sample size of n=6.

[0135] As shown in Graph A of Figure 9, the daily average VPD value in the humidified greenhouse was less than or equal to the daily average VPD value in the unhumidified greenhouse for every day during the measurement period. Furthermore, as shown in Graph B of Figure 9, the net assimilation rate was higher in the humidified greenhouse than in the unhumidified greenhouse. From the above, it can be concluded that the net assimilation rate was higher in the humidified greenhouse, where the daily average VPD value was lower. [Explanation of Symbols]

[0136] 10. Carbon dioxide concentration control device 11 Information acquisition department 12 Concentration control unit 20. Decision Tree (Predictive Model) 21 Root Nodes 22 intermediate nodes 23 Leaf Nodes 30 Algorithms 31. First numerical range (numerical range of vapor pressure deficit within the facility) 32. Second numerical range (Numerical range of the first time derivative of the vapor pressure deficit within the facility) 33. Third numerical range (Numerical range of the second derivative of the time value of the vapor pressure deficit within the facility) 34. Lower limit of carbon dioxide concentration (the lower limit of the numerical range for carbon dioxide concentration within the facility) 35. Upper limit of carbon dioxide concentration (lower limit of the numerical range for carbon dioxide concentration within the facility) 200 Carbon Dioxide Application System 500 Carbon Dioxide Concentration Control System Facility F INF-1 First Factor Information INF-2 Second Factor Information P plant body PI Prediction Indicators

Claims

1. A carbon dioxide concentration control device that controls the carbon dioxide concentration in the cultivation environment in which plants are grown, An information acquisition unit that acquires humidity change information showing the measurement results of humidity changes in the cultivation environment, and carbon dioxide concentration information showing the measurement results of carbon dioxide concentration in the cultivation environment. A carbon dioxide concentration control device comprising: a concentration control unit that controls a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired by the information acquisition unit, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired by the information acquisition unit.

2. The prediction model is constructed using machine learning with training data that uses the humidity indicated by the humidity change information at a predetermined time, the first time derivative of the humidity change, and the second time derivative of the humidity change as example data, and a reference index indicating the state of water stress in the plant at the predetermined time as the ground truth data. The carbon dioxide concentration control device according to claim 1, wherein the output result is a predictive index indicating the predicted state of water stress on the plant at the time when the humidity change of the cultivation environment is measured.

3. The aforementioned reference index is an index based on at least the first time derivative of the arbitrary diameter reference change, obtained by subtracting the diameter of the arbitrary part of the plant at the start of the first period from the diameter of the arbitrary part of the plant at the predetermined time point, and then dividing this value by the diameter of the arbitrary part of the plant at the start of the first period to obtain the arbitrary diameter reference change. The carbon dioxide concentration control device according to claim 2, wherein the predictive index is an index based at least on the first time derivative of the arbitrary diameter change, obtained by subtracting the diameter of an arbitrary part of the plant at the start of a second period (a period later than the first period) from the diameter of an arbitrary part of the plant at the time when the humidity change of the cultivation environment is measured, and dividing this value by the diameter of the arbitrary part of the plant at the start of the second period to obtain the arbitrary diameter change.

4. The carbon dioxide concentration control device according to claim 2 or 3, wherein the prediction model is a decision tree.

5. The concentration control unit, The first factor information indicating the predictive factors of the prediction indicator, recorded at the root node of the decision tree, The second factor information, which is different from the first factor information, is recorded in the intermediate node of the decision tree, The prediction index recorded in the leaf node of the decision tree, The predetermined carbon dioxide concentration setting value for the cultivation environment, Based on this, the concentration control unit generates an algorithm to determine whether or not to allow the carbon dioxide application device to apply carbon dioxide, The carbon dioxide application device is controlled by the algorithm described above. The algorithm includes the following conditions for determining whether or not the concentration control unit will allow the carbon dioxide application device to apply carbon dioxide: The numerical range of humidity in the aforementioned cultivation environment, The numerical range of the first derivative value of the time change in humidity of the aforementioned cultivation environment, The numerical range of the second derivative value of the time change in the aforementioned cultivation environment, The numerical range of carbon dioxide concentration in the aforementioned cultivation environment, A carbon dioxide concentration control device according to claim 4, which includes the following:

6. The carbon dioxide concentration control device according to any one of claims 1 to 3, wherein the humidity change is at least one of vapor pressure deficit, relative humidity, and absolute humidity.

7. The information acquisition unit further acquires solar radiation information showing the measurement results of the solar radiation in the cultivation environment. The carbon dioxide concentration control device according to any one of claims 1 to 3, wherein the concentration control unit controls the carbon dioxide application device based on the solar radiation information acquired by the information acquisition unit.

8. A program for causing a computer to function as a carbon dioxide concentration control device according to claim 1, wherein the program causes the computer to function as the information acquisition unit and the concentration control unit.

9. A carbon dioxide concentration control system for controlling the carbon dioxide concentration in the cultivation environment in which plants are grown, An information acquisition unit that acquires humidity change information showing the measurement results of humidity changes in the cultivation environment, and carbon dioxide concentration information showing the measurement results of carbon dioxide concentration in the cultivation environment. A carbon dioxide concentration control system comprising: a concentration control unit that controls a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the output result obtained by inputting the humidity indicated by the humidity change information acquired by the information acquisition unit, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and the carbon dioxide concentration information acquired by the information acquisition unit.

10. A carbon dioxide concentration control method, which is performed by one or more computers, for controlling the carbon dioxide concentration in a cultivation environment in which plants are being grown, An information acquisition step involves acquiring humidity change information showing the measurement results of humidity changes in the cultivation environment, and carbon dioxide concentration information showing the measurement results of carbon dioxide concentration in the cultivation environment. A carbon dioxide concentration control method comprising: an output result obtained by inputting the humidity indicated by the humidity change information acquired in the information acquisition step, the first time derivative of the humidity change, and the second time derivative of the humidity change into a prediction model, and a concentration control step that controls a carbon dioxide application device for applying carbon dioxide to the cultivation environment based on the carbon dioxide concentration information acquired in the information acquisition step.

Images