KR20240043986A - A method for providing an appropriate cultivation guid in a living horticultural environment - Google Patents

Abstract

The IoT-based plant cultivation system includes a water tank for storing the nutrient solution supplied to the plants, an air pump for sucking air from outside the plant cultivation device and injecting it into the nutrient solution, and a sensor unit for detecting information related to the cultivation environment. , a camera for acquiring image data by photographing the plant, an actuator for controlling variables related to the cultivation environment, a memory in which the operating conditions and operation contents of the actuator are recorded, and based on the operating conditions and the operation contents At least one processor controls the operation of the plant cultivation device and includes a wired or wireless communication interface, and the sensor unit includes a temperature sensor, a humidity sensor, an optical sensor, and an EC sensor, and information related to the cultivation environment. It includes temperature information, humidity information, illuminance information, and EC information of the nutrient solution inside the plant cultivation device, and to achieve the purpose of exploring cultivation conditions to improve the growth rate of the plant, a plurality of Internet of Things devices are remotely used. Alteration technology is applied to set different control points according to the experimental design method for the plant cultivation environment, and using IoT technology, plant growth rate, above-ground and root zone environmental information, and light information are collected from multiple plant cultivation devices. You can collect and tune it with better cultivation guide information through reinforcement learning.

Description

A method for providing an appropriate cultivation guide in a living gardening environment {A METHOD FOR PROVIDING AN APPROPRIATE CULTIVATION GUID IN A LIVING HORTICULTURAL ENVIRONMENT}

The present invention relates to a method for providing users with environmental information necessary for plant growth in a living gardening environment.

Recently, the number of single-person households is gradually increasing. As the number of people who have an emotional attachment to the plants they grow at home increases, the number of people looking for 'companion plants' is increasing. However, it is not easy for ordinary people, other than professional farmers, to cultivate specific plant species without proper knowledge of the characteristics of the plant species. In addition, it is not easy for ordinary people who carry out daily life and livelihood activities to manage the environment for growing plants and diagnose growth conditions.

In residential horticulture, the goal is to provide an appropriate cultivation guide, not an optimal cultivation guide, considering aesthetic purposes rather than a commercial agricultural industry where production and profit are prioritized. A cultivation guide refers to providing a summary of environmental information necessary for plant growth. For example, it may mean providing target control points for the root zone, above-ground zone, and light environment for each plant species.

In line with the diversified needs of modern people, the number of domestic horticultural crops is much more diverse than that of crops widely grown commercially, and it is necessary to provide appropriate information tailored to the needs of non-experts.

It can be used as a cultivation guide by providing information on the appropriate above-ground air temperature, humidity, illuminance, and root zone electrical conductivity for each crop. For example, the Rural Development Administration's smart farm optimal environment information service website provides the optimal environment for three types of commercial crops (ripe tomatoes, strawberries, and paprika).

There are no clear standards for the optimal cultivation guide for plants in a hydroponic cultivation environment, and is set based on the user's experience or customary standards. Since plant cultivation in a plant cultivator located in a home indoor environment, which is a living gardening environment, has different humidity and light levels from commercial facility gardening environments, it is necessary to develop a cultivation guide optimized for the indoor environment.

In order to discover and establish appropriate cultivation methods for various species of plants and various environments, deriving optimal cultivation methods through experiments on individual plant species of each species is not efficient in manpower and resources, so acquisition of data using IoT technology , an automatic tuning method for cultivation methods needs to be developed.

A method can be provided to maintain the optimal growth of plants and to easily manage the growth status of plants by ordinary individuals, not commercial agricultural industry operators, using smartphones, computers, etc.

The technical challenge that this embodiment aims to achieve is not limited to the technical challenges described above, and other technical challenges can be inferred from the following embodiments.

The IoT-based plant cultivation system includes a water tank for storing the nutrient solution supplied to the plants, an air pump for sucking air from outside the plant cultivation device and injecting it into the nutrient solution, and a sensor unit for detecting information related to the cultivation environment. , a camera for acquiring image data by photographing the plant, an actuator for controlling variables related to the cultivation environment, a memory in which the operating conditions and operation contents of the actuator are recorded, and based on the operating conditions and the operation contents At least one processor controls the operation of the plant cultivation device and includes a wired or wireless communication interface, and the sensor unit includes a temperature sensor, a humidity sensor, an optical sensor, and an EC sensor, and information related to the cultivation environment. It includes temperature information, humidity information, illuminance information, and EC information of the nutrient solution inside the plant cultivation device, and to achieve the purpose of exploring cultivation conditions to improve the growth rate of the plant, a plurality of Internet of Things devices are remotely used. Alteration technology is applied to set the plant cultivation environment differently according to the experimental design method, and using IoT technology, plant growth rate, above-ground and root zone environmental information, and light information are collected from multiple plant cultivation devices. You can collect and tune it with better cultivation guide information through reinforcement learning.

Individuals who want to grow plants at home or in the office for purposes such as hobbies, ornamentation, or air purification have the advantage of being able to grow plants easily and with fun. Appropriate cultivation guides can be provided to users based on plant species, growth stage, and growth height in a living gardening environment that considers aesthetic purposes rather than a commercial agricultural industry where production and profit are prioritized.

1 illustrates a personal plant cultivation system, according to one embodiment.
Figure 2 shows a screen on which cultivation environment information detected by a sensor unit is displayed in a predetermined application, according to an embodiment.
Figure 3 shows a screen displayed on a predetermined application to control an actuator, according to one embodiment.
Figure 4 shows a plant cultivation device linked to an external sensor unit, according to an embodiment.
Figure 5 shows block coding for inputting settings for controlling a plant cultivation device on a personal terminal, according to an embodiment.
Figure 6 shows a configuration for a plant cultivation device to perform an air purifying function, according to an embodiment.
Figure 7 shows a plant cultivation device, according to one embodiment.
Figure 8 shows a prediction model used to predict the growth state of a plant and data input to the prediction model, according to an embodiment.
Figure 9 shows the general growth stages of plants.
Figure 10 shows a plant image with the background removed, according to one embodiment.
Figures 11 and 12 show a fuzzy membership function for specifying the current growth stage of the plant according to the number of days of cultivation and growth rate, according to an embodiment.
Figure 13 shows a fuzzy control system used in an IoT-based plant cultivator, according to an embodiment.

Below, several embodiments will be described clearly and in detail with reference to the accompanying drawings so that those skilled in the art (hereinafter referred to as skilled in the art) can easily practice the present invention. will be.

1 shows a plant cultivation system, according to one embodiment.

Referring to Figure 1, a personal plant cultivation system 10000 for an individual to easily cultivate plants may include a user's personal terminal 100, a server 200, and a plant cultivation device 300.

The personal plant cultivation system 10000 may include a personal terminal 100 of a user who cultivates plants using the plant cultivation device 300. The personal terminal 100 may include various electronic devices such as smartphones, PCs, tablet PCs, and wearable devices.

Server 200 may be a central server, multimedia computer, laptop computer, desktop computer, grid computing resource, virtualized computer resource, cloud computing resource, peer-to-peer distributed computing device, or a combination thereof. A computing device may include at least one processor and at least one memory.

The database (DB) may be located inside or outside the server 200 and collects various information about the cultivation environment, image data of the plant, operation of the actuator, growth status of the plant, etc. from a plurality of plant cultivation devices. It can be received and saved. The server 200 can operate as a cloud server in conjunction with a database (DB). The server 200 may obtain useful information about plant cultivation using a database (DB) and feed the obtained information back to the plant cultivation device 300 or the personal terminal 100.

The plant cultivation device 300 is a device for inserting and cultivating plants and may include components for creating an appropriate environment for cultivating plants and checking growth conditions. The plant cultivation device 300 can cultivate plants using a hydroponic cultivation method such as Ebb & Flow, a spray aquarium, or a liquid aquaculture system. According to one embodiment, the plant cultivation device 300 may be mainly located in indoor spaces such as homes, offices, and buildings, and may be manufactured in a sufficiently small size so that it can be located in indoor spaces. For example, the width, length, and height of the plant cultivation device 300 may be limited to less than 1 meter.

The plant cultivation device 300, the server 200, and the personal terminal 100 may be connected to each other based on a communication interface. Communication interfaces include a wired local area network (LAN), a wireless local area network (WLAN) such as Wi-fi (Wireless Fidelity), a wireless personal area network (Wireless Personal Area Network) such as Bluetooth; WPAN), Wireless Universal Serial Bus (Wireless USB), Zigbee, Near Field Communication (NFC), Radio-frequency identification (RFID), Power Line communication (PLC), or 3rd Generation (3G), 4th Generation (4G), LTE It may include a modem communication interface that can connect to a mobile cellular network, such as Long Term Evolution. The Bluetooth interface may support Bluetooth Low Energy (BLE).

The personal plant cultivation system 10000 may be an IoT network system including a plurality of IoT (Internet of Things) devices. The plant cultivation device 300 and the personal terminal 100 may be IoT devices capable of transmitting or receiving data by communicating with at least one other device through a wired/wireless communication interface, and the personal plant cultivation system 10000 May include access points and gateways. The IoT network system uses transmission protocols such as UDP (User Datagram Protocol) and TCP (Transmission Control Protocol) and 6LoWPAN (IPv6 Low-power Wireless Personal Area Networks) for information exchange (communication) between two or more components within the IoT network system. protocols, IPv6 Internet routing protocol, and application protocols such as constrained application protocol (CoAP), hypertext transfer protocol (HTTP), message queue telemetry transport (MQTT), and MQTT for sensors networks (MQTT-S).

In this embodiment, the plant cultivation device 300 and the personal terminal 100 may be connected to each other through an access point, and the access point may be included in the plant cultivation device 300 or the personal terminal 100, or may be included in another IoT device. there is. Hereinafter, the personal terminal 100 and the plant cultivation device 300 may be connected through an application running on the personal terminal 100, and connected means that the personal terminal 100 is connected to the plant cultivation device through the server 200 ( It may include both an embodiment in which the personal terminal 100 is connected to the plant cultivation device 300 and an embodiment in which the personal terminal 100 is connected to the plant cultivation device 300 through an access point. The meaning that the personal terminal 100 controls the plant cultivation device 300 refers to an embodiment in which the personal terminal 100 connects to the server 200 and controls the plant cultivation device 300 through the server 200 and the personal terminal (100) may include all embodiments in which the plant cultivation device 300 is controlled through an access point.

The plant cultivation device 300 includes a water tank 310 for storing the nutrient solution for hydroponic cultivation supplied to the plants, and an air pump 320 for sucking air from outside the plant cultivation device 300 and injecting it into the nutrient solution within the water tank 310. , a sensor unit 330 for detecting information related to the cultivation environment of the plant cultivation device 300, a camera 340 for photographing plants being cultivated and acquiring image data, and a camera 340 for controlling various variables related to the cultivation environment. It may include an actuator 350 and at least one processor 360 that controls the overall operation of the plant cultivation device 300 and includes a wireless communication interface for communication with the outside. The plant cultivation device 300 may further include a memory (not shown), and at least one processor 360 may generate a control signal of the actuator 350 based on logic recorded in the memory (not shown). . Hereinafter, nutrient solution may mean not only water in which nutrients necessary for plant growth are dissolved, but also pure water that does not contain nutrients.

Referring to Figure 6, the plant cultivation device 300 can suck in external air through the air pump 320 to clean the air, and the sucked external air can be input into the nutrient solution in the water tank 310 through the air line. You can. Fine dust contained in the sucked outside air may settle inside the water tank 310 or be absorbed into the roots of plants. Fine dust in the nutrient solution that is not absorbed into the roots of the plant can be removed when changing the water in the water tank 310. In this embodiment, the air purification function can be implemented by natural biochemical interactions between plants and water without the need to add a separate filter to clean or purify the air, thereby reducing maintenance costs by eliminating the need to replace the filter. It is environmentally friendly as there is no need to dispose of used filters.

According to one embodiment, the air pump 320 may operate in conjunction with an external fine dust sensor. For example, the air pump 320 may operate when the fine dust concentration received from the fine dust sensor is greater than or equal to a reference value. Alternatively, the air pump 320 may be operated by an operation signal received from an external device including a fine dust sensor.

According to one embodiment, the bed on the water tank (the part where the plants are placed) may include a woven filter capable of adsorbing fine dust. To prevent changes in the pH of the nutrient solution due to fine dust, the nutrient solution may contain a buffer solution to maintain PH. The plant grown in the plant cultivation device 300 may be one of, but is not limited to, Bengal rubber tree, skin dapsus, yellow palm tree, spathiphyllum, ivy, and table palm tree.

Referring again to FIG. 1, the sensor unit 330 may include an optical sensor for measuring the illuminance inside the plant cultivation device 300, a temperature sensor for measuring the temperature, and a humidity sensor for measuring humidity. According to one embodiment, the sensor unit 330 may further include an EC (Electrical Conductivity) sensor for measuring changes in electrical conductivity of the nutrient solution due to photosynthesis of plants. Information related to the cultivation environment acquired by the sensor unit 330 may include at least one of illuminance information, temperature information, humidity information, and EC information of the nutrient solution inside the plant cultivation device 300. Figure 2 shows a screen on which cultivation environment information detected by a sensor unit is displayed in a predetermined application, according to an embodiment.

Referring again to FIG. 1, the camera 340 is used to obtain image data by photographing plants being cultivated, and may include, for example, an RGB camera. By applying an image processing algorithm to the image data acquired by the camera 340, various growth data such as plant height and leaf area index can be obtained.

The actuator 350 may include a light control unit for controlling illuminance or light, a nutrient solution control unit for controlling the nutrient solution, and a temperature control unit for controlling the temperature. The actuator 350 may operate based on a control signal received from the processor 360. For example, the light control unit may include an LED light source such as an RBW LED array and may be turned on/off or have its intensity adjusted under a control signal. The nutrient solution control unit may include an irrigation pump, and operation may be determined under a control signal or the amount of supplied nutrient solution may be adjusted. The temperature control unit may include a ventilation fan (or Peltier) and may be configured to raise or lower the temperature of the plant cultivation device 300.

Information related to the cultivation environment acquired by the sensor unit 330 (e.g., various sensor values), information about the operation of the actuator 350 (e.g., whether the irrigation pump is running and operation time), and a camera ( Each of the image data acquired by 340) may be acquired at a reference time (for example, 10 minutes), and the acquired information and data may be transmitted to the server 200 and/or the personal terminal 100. .

The user checks the cultivation environment of the plant cultivation device 300, operation information of the actuator 350, images of plants in the plant cultivation device 300, etc. through the personal terminal 100 and determines the optimal cultivation environment for plant growth. The plant cultivation device 300 can be controlled to achieve this. For example, a predetermined application for controlling and monitoring the plant cultivation device 300 may be installed on the user's smartphone. By monitoring the plant cultivation device 300 and controlling the actuator 350 through the application, temperature, Variables related to the cultivation environment, such as nutrient solution and illumination level, can be adjusted. Each reference value of variables related to the cultivation environment is different for each plant, and values pre-stored in the server 200 may be used.

The plant cultivation device 300 or the actuator 350 may operate under a control signal received from the personal terminal 100, the server 200, or an operation unit (not shown) of the plant cultivation device 300. The control signal includes information related to the cultivation environment (e.g., various sensor values acquired by the sensor unit 330), operation information of the actuator 350 (e.g., operation status and operation time of various actuators 350) ), and image data acquired by the camera 340 (for example, a two-dimensional image of a plant). For example, the user displays the plant cultivation device 300 (see FIG. 7) based on at least one of information related to the cultivation environment accumulated during a predetermined time period in the past, accumulated operation information of the actuator, and accumulated image data. You can directly control the actuator 350 through or control the actuator 350 through an application. Alternatively, the server 200 generates a control signal for the actuator 350 based on at least one of information related to the cultivation environment accumulated during a predetermined time period in the past, accumulated operation information of the actuator, and accumulated image data, and sends it to the plant. It can be transmitted to the cultivation device 300.

According to one embodiment, control of the plant cultivation device 300 may include manual control, semi-automatic control, and fully automatic control. Manual control is a form in which the user directly controls the actuator 350 of the plant cultivation device 300 based on image data obtained from the camera 340 and values related to the cultivation environment obtained from the sensor unit 330. The server 200 can guide the operation of the actuator 350 to the user based on artificial intelligence technology, and the user can control the actuator 350 according to the guide. Figure 3 shows a screen displayed on a predetermined application to control an actuator, according to one embodiment.

Semi-automatic control is a form in which the user pre-sets the actuator 350 of the plant cultivation device 300 to perform a certain operation under certain conditions. For example, the user can preset the LED light source to turn on at a specific time. Users can preset to increase the amount of water supplied to plants by 30% when the accumulated temperature (sum of daily average temperatures after planting) reaches 100℃. Users can preset to perform irrigation for 1 minute and reset the cumulative illuminance value when the daily cumulative illuminance value reaches 10000 lm·h.

According to one embodiment, a user may set the operating conditions and operation contents of the actuator 350 of the plant cultivation device 300 using block coding. Block coding is a tool for designing various programs by combining or arranging blocks with defined logic, and may include Entry, Scratch, etc. The user has the advantage of being able to easily design the control logic of the plant cultivation device 300 through block coding even if the user does not know high-level languages such as C language and JAVA. Figure 5 shows block coding for inputting settings for controlling a plant cultivation device on a personal terminal, according to an embodiment. Referring to Figure 5, the user can design the side fan to turn on after 8 a.m. and the LED light source to turn on at an intensity of level 10.

According to one embodiment, while the user's personal terminal 100 and the plant cultivation device 300 are connected, the operating conditions and operation contents of the actuator 350 designed by the user through block coding through the personal terminal 100 are By being transmitted to the plant cultivation device 300 through the server 200, the actuator 350 of the plant cultivation device 300 can be controlled. In this embodiment, the plant cultivation device 300 can be controlled in real time by a user using block coding-based commands transmitted from the personal terminal 100, which can be a form of manual control.

According to another embodiment, the operating conditions and operation contents of the actuator 350 designed by the user through block coding may be transmitted to the plant cultivation device 300 and recorded in the memory (not shown) of the plant cultivation device 300. . In this embodiment, the plant cultivation device 300 operates as an embedded device in which block coding-based operating conditions and operation contents are recorded in the device itself, so the plant cultivation device 300 is connected to the personal terminal 100 and the server 200. Even when not connected, it can be controlled according to operating conditions and operation contents previously stored in memory (not shown), and this can be a form of semi-automatic control.

Automatic control is a form in which variables related to the cultivation environment of the plant cultivation device 300 are automatically adjusted when the user inputs only the type of plant and the date of planting into the application or the plant cultivation device 300. In this embodiment, the actuator 350 of the plant cultivation device 300 may operate without user intervention. For example, the server 200 may determine the operation of the actuator 350 to create an optimized cultivation environment based on artificial intelligence technology, and the actuator 350 of the plant cultivation device 300 may be operated by the server 200. It can be controlled based on commands received from.

The server 200 can predict the growth state of plants based on artificial intelligence technology and determine the operation of the actuator 350 for an optimal cultivation environment. Referring to Figure 8, the server 200 can predict the growth state of a plant based on the prediction model 1000. When the accumulated data of the days since planting of the plant, temperature, humidity, EC (electric conductivity), irrigation pump operation, and illuminance are input to the prediction model (1000), the prediction model (1000) outputs the predicted growth status. can do.

The prediction model 1000 can be generated based on machine learning techniques by applying the above-described six types of input data and the actually obtained growth state of the plant as input and output, respectively. Machine learning techniques include Support Vector Machine (SVM), Random forest, Support Vector Machine (SVM), Naive Bayes, Adaptive Boosting (AdaBoost), and Gradient Boosting ( Gradient Boosting, K-means clustering, Artificial Neural Network, etc. For example, the prediction model 1000 is a Long Short-Term (LSTM), a type of Recurrent Neural Network (RNN) that includes an input layer, a hidden layer, and an output layer. Output results can be generated using an artificial neural network in the form of a model.

According to one embodiment, six input data (days since planting, temperature, humidity, EC, irrigation pump operation, illuminance) input to the prediction model are obtained every predetermined time interval (e.g., 10 minutes). You can. If 6 pieces of input data are acquired every 10 minutes, the number of pieces of input data acquired per day may be 6

As input data input to the prediction model 1000, input data acquired over a plurality of days can be accumulated and used. When using input data acquired during n_day, the number of input data may be 6 n_day is a variable representing the unit period for predicting growth status using the prediction model (1000) or accumulating input data for learning the prediction model (1000), and is a real value. For example, if you want to predict growth status or train the prediction model (1000) using the prediction model (1000) using input data acquired for 3 days, the input data includes 6 can do.

According to one embodiment, all of the six types of input data obtained after planting the plant can be used for prediction or learning, but the input data can be accumulated for each unit period. When the number of days elapsed after formulation is N, and the six types of data acquired after formulation are to be accumulated and used every predetermined unit period (n_day), the number of input data input to the prediction model (1000) is as follows [mathematics] It is the same as Equation 1].

[Equation 1]

6

(6

24

n_day)Cumulative number of input data after formalization = (N - n_day)

6

(6

24

n_day)

According to one embodiment, if input data obtained by cultivating a specific plant species in M plant cultivation devices is used to learn the prediction model 1000, the total number of input data is as follows [Equation 2] same. Assume that the values for the elapsed days after planting obtained from M plant cultivation devices are the same.

[Equation 2]

(N - n_day)

6

(6

24

n_day) Cumulative number of post-planting input data obtained from M plant cultivation periods = M

(N - n_day)

6

(6

24

n_day)

(M: Number of plant cultivation devices used for learning, N: Number of days since planting)

The output of the prediction model 1000 may include the growth state of the plant. The growth state may include data such as plant height or leaf area index, and this can be obtained by analyzing image data acquired from the camera 340 using an image processing algorithm.

When the prediction model 1000 is generated based on an artificial neural network, the loss representing the performance of the artificial neural network may be determined to be smaller as the output growth state is the same as the actually obtained growth state. According to one embodiment, training for an artificial neural network may be performed using a back propagation algorithm that updates synaptic weights so that loss is reduced.

According to one embodiment, the plant cultivation device 300 may be controlled so that plants can grow well based on the generated prediction model 1000. According to one embodiment, the server 200 may determine input data of the prediction model 1000 to ensure the best growth condition, which is the output data of the prediction model 1000. For example, the server 200 may repeatedly input various input data into the prediction model 1000 using Bayesian optimization to determine input data that can provide the best growth condition. In this embodiment, the server 200 performs plant cultivation such as temperature, illuminance, humidity, EC, whether the irrigation pump is in operation, etc. so that the plants of the plant cultivation device 300 can grow best based on the prediction model 1000. Values related to the cultivation environment of the device 300 or the operation of the actuator 350 can be determined, and based on the determined value or the operation of the actuator 350, the user can manually control the plant cultivation device 300 or use the plant cultivation device 300 ) can be automatically controlled.

Figure 4 shows a plant cultivation device linked to an external sensor unit, according to an embodiment.

Referring to Figure 4, the personal plant cultivation system 10000 may further include an external sensor unit 400 located outside the plant cultivation device 300. The plant cultivation device 300 and the external sensor unit 400 may be located in the same indoor space, and the external sensor unit 400 may be located around the plant cultivation device 300. The plant cultivation device 300 may operate in conjunction with the external sensor unit 400. For example, if the plant cultivation device 300 is located in a home, the external sensor unit 400 may also be located in the home. However, the external sensor unit 400 does not necessarily have to be located in the same indoor space as the plant cultivation device 300, and the plant cultivation device 300 and the external sensor unit 400 may be located indoors and outdoors, respectively.

According to one embodiment, the external sensor unit 400 may include a temperature sensor, a humidity sensor, a fine dust sensor, etc. Based on the sensing value received from the external sensor unit 400, the operation of the plant cultivation device 300 may be determined or a control signal may be received from the external sensor unit 400. For example, if the room temperature value received from the temperature sensor installed in the room is higher than the reference value, the fan among the actuators 350 of the plant cultivation device 300 may be driven. If the fine dust concentration in the room received from the fine dust sensor installed in the room is higher than the reference value, the air pump 320 of the plant cultivation device 300 is driven so that the air in the room can be sucked into the water tank of the plant cultivation device 300.

Figure 7 shows a plant cultivation device, according to one embodiment.

The plant cultivation device 7000 of FIG. 7 represents an embodiment in which the plant cultivation device 300 described above with reference to FIG. 1 or FIG. 4 is actually implemented, but the plant cultivation device 300 is not limited thereto.

Referring to FIG. 7, the plant cultivation device 7000 includes an actuator such as an air inlet or outlet (AIO), a Peltier (PT), a watering inlet (WI), a watering pump (not shown), or an LED array (LDR). It can be included. The air inlet or outlet (AIO) may operate for ventilation of the plant cultivation device 7000, and the Peltier (PT) may operate to regulate the temperature within the plant cultivation device 7000. The irrigation inlet (WI) is for injecting nutrient solution, and the nutrient solution can be supplied through an irrigation pump. The LED array (LDR) is used to control illuminance (or light) and may include white LED, red LED, or blue LED.

The plant cultivation device 7000 includes a first sensor unit (S1) including a temperature sensor, a humidity sensor, and a light sensor at the top where the plants are located, and a second sensor unit (S2) including an EC sensor at the bottom where the water tank is located. may include.

The plant cultivation device 7000 may include a camera C for photographing the growth state of the plant. The camera (C) may be located at the top where the plants are located.

The plant cultivation device 7000 can control the nutrient solution supplied to the plants by controlling the operation of the irrigation pump. The plant cultivation device 7000 includes a first bottom part (B1) containing a water tank, a third bottom part (B3) including a sponge medium on which plants are planted and a bed that is a frame for placing plants, and a first bottom part (B1) ) and a second bottom portion (B2) located between the third bottom portion (B3). The nutrient solution in the water tank located in the first bottom (B1) is drawn up to the second bottom (B2), which is the middle layer between the sponge medium and the water tank, through an irrigation pump, and the nutrient solution fills the second bottom (B2) with the roots of the plants. can be absorbed. When the operation of the irrigation pump is stopped, the nutrient solution in the second bottom part (B2) may be configured to go back down to the water tank in the first bottom part (B1).

The second bottom portion (B2) may include a hole for recovering the nutrient solution from the bottom of the second bottom portion (B2) to the water tank so that the nutrient solution does not rise above the standard height.

The plant cultivation device 7000 displays sensor data acquired from the first sensor unit (S1) and the second sensor unit (S2) to the user or uses an air inlet or outlet (AIO), Peltier (PT), and irrigation pump (WI). It may further include a display (DP) capable of controlling actuators such as. The display (DP) may include a Liquid Crystal Display (LCD), Light Emitting Diode (LED), etc. and may be touch-sensitive.

In order to discover and establish appropriate cultivation methods for various species of plants and various environments, deriving optimal cultivation methods through experiments on individual plant species of each species is not efficient in manpower and resources, so acquisition of data using IoT technology , an automatic tuning method for cultivation methods needs to be developed. Hereinafter, a method for providing a plant cultivation guide in a living gardening environment will be described with reference to FIGS. 9 to 14.

Sensors built into IoT-based plant cultivators can measure above-ground temperature, above-ground humidity, illuminance, and rhizosphere EC. The above-ground part is the part that is above ground excluding the plant's roots and its surrounding environment, and the rhizosphere part is the part where the plant's roots are located and its surrounding environment. EC (Electrical Conductivity) is pure water, which does not conduct electricity, so EC = 0. In water mixed with fertilizer, the electrolyte allows electricity to flow, so the higher the fertilizer concentration, the higher the EC value.

The above items can be measured at regular intervals and the data can be transmitted to the cloud server. An IoT-based plant cultivator may include a camera for photographing plants.

The IoT-based plant cultivator can be equipped with actuators including an LED module controlled by the dimming function (brightness control), a peristaltic pump, nutrient solution (water + fertilizer) supply control, a fan for ventilation with the outside, and a heating module for heating. You can.

According to one embodiment, the initial value of the growth guide may be set to the initial value of the growth guide according to a conventional cultivation method. It can be set by inferring the appropriate above-ground/rhizosphere/light environment through existing plant databases at universities, academic institutions, public institutions, and open databases.

Referring to Figure 9, the general growth stages of plants are divided into Lag phase, Log phase (or Exponential phase), and Steady State (or Stationary phase), and form a sigmoid function.

IoT-based plant cultivators can measure plant growth over time. To measure the degree of plant growth, the degree of plant growth can be measured by acquiring image information from a camera mounted on a plant cultivator and using the acquired image information. By separating the plant and background areas, the plant can be detected, the area measured, and the growth rate determined. Figure 10 shows a plant image with the background removed, according to one embodiment. At this time, image processing technologies can be used and OpenCV's background subtraction function, Measure Rosette Area tool, etc. can be used. Cultivation patterns that change over time can be recorded and used to determine growth stages.

An IoT-based plant cultivator can derive appropriate cultivation guides for each growth stage. The concentration of nutrients in the rhizosphere, temperature and humidity in the above-ground part, and light environment required for each growth stage are different. (For example, log phase requires high fertilizer concentration, especially high nitrogen concentration, for sufficient growth)

Through an IoT-based plant cultivator, data can be collected to derive appropriate cultivation guides for each growth stage. In the initial values of the cultivation guide, the changes (alterations) of each factor (i.e. temperature, humidity, EC, light) are set at multiple levels and applied to the IoT-based plant cultivator. When setting each control factor at multiple levels, it can be configured using the Box-Benhken Model, etc.

Ultimately, the cultivation guide can be changed to conditions that allow for faster growth. Changes to these cultivation guides must be made through observation of at least one growing season, so they are updated at regular intervals.

The cultivation guide can be updated using the ‘reinforcement learning’ method, which rewards cultivation environments that produce high growth rates. Cultivation guides (temperature, humidity, EC, and light conditions that are the target points of control) can be written in three growth stages (Lag, Log, Stationary Phase) depending on the cultivation time and growth of the plant.

In other words, by observing environmental information obtained from a large number of plant cultivators distributed to consumers and the resulting plant growth, optimal control points are derived in the Lag phase, Log Phase, and Stationary Phase growth stages, and depending on the number of cultivation days and growth. Three growth stages can be set.

According to one embodiment, an IoT-based plant cultivator can determine the growth stage by applying a fuzzy inference system. IoT-based plant cultivators can use a ‘fuzzy inference system’ that determines the current growth stage by inputting the cultivated fruit and growth rate and outputs an optimal cultivation guide (control point).

This is because ambiguity arises at the boundaries of each growth stage, and accordingly, it is also ambiguous as to which growth stage the corresponding cultivation guide should be applied to. For example, when growth occurs at the maximum rate on the 60th day of cultivation, On the 30th day, it may be difficult to determine whether it is a lag phase or a log phase.

IoT-based plant cultivators can resolve ambiguities at the boundary points of each growth stage by applying the Fuzzy Inference System (FIS). By specifying the input fuzzy membership function as shown in Figures 11 and 12, the current growth stage of the plant can be specified according to the number of days of cultivation and growth rate.

Figure 13 shows a fuzzy control system used in an IoT-based plant cultivator, according to an embodiment.

The output function consists of three membership functions, and the fuzzy control system generates a 1 by 4 matrix of optimal cultivation guides according to each growth stage (lag, log, steady) derived in the previous process [temperature, humidity, EC, Light quantity] can be output.

The descriptions are intended to provide example configurations and operations for implementing the invention. The technical idea of the present invention will include not only the embodiments described above, but also implementations that can be obtained by simply changing or modifying the above embodiments. In addition, the technical idea of the present invention will also include implementations that can be easily achieved by changing or modifying the embodiments described above.

Claims (1)

A plant cultivation device for hydroponically cultivating plants;
a server connected to the plant cultivation device based on a wired or communication interface; and
A user's personal terminal is connected to the plant cultivation device through a predetermined application and controls the plant cultivation device,
The plant cultivation device,
A water tank for storing the nutrient solution supplied to the plants;
an air pump for sucking air from outside the plant cultivation device and injecting it into the nutrient solution;
A sensor unit for detecting information related to the cultivation environment;
a camera for acquiring image data by photographing the plant;
an actuator for controlling variables related to the cultivation environment; and
Comprising at least one processor that controls the operation of the plant cultivation device and includes a wired or wireless communication interface,
Information related to the cultivation environment, operation information of the actuator, and the image data are transmitted to the server,
In order to achieve the purpose of exploring cultivation conditions to improve the growth of the plants, an alternation technology is applied to set different control points in the plant cultivation environment of multiple IoT devices remotely according to the experimental design method,
An IoT-based plant cultivation system that uses IoT technology to collect plant growth, above-ground and root zone environmental information, and light information from multiple plant cultivation devices, and tunes them into better cultivation guide information through reinforcement learning.

Images