Environmental parameter setting method and apparatus for plant cultivation box

Abstract

The present application relates to the technical field of plant cultivation, and provides an environmental parameter setting method and apparatus for a plant cultivation box. The method comprises: on the basis of an environmental monitoring device, acquiring current monitoring values of environmental parameters in a plant cultivation box; inputting the current monitoring values of the environmental parameters into a plant growth model to obtain current growth state data of plants in the plant cultivation box; inputting the current growth state data into a trained strategy network to obtain current optimal setting values of the environmental parameters; and performing setting on the basis of the current optimal setting values of the environmental parameters. According to the present application, automatic adjustment of environmental parameter settings of a plant cultivation box is realized, without requiring manual intervention; the requirements for management experience of a planter are reduced while plant growth conditions in the plant cultivation box are met, so that the planting process is more simplified and convenient.

Description

Environmental parameter setting method and device for plant cultivation box

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to Chinese patent application No. 202410322320.2, filed on March 20, 2024, entitled “Method and Device for Setting Environmental Parameters of a Plant Cultivation Box,” which is incorporated herein by reference in its entirety. Technical Field

[0003] The present application relates to the field of plant cultivation technology, and in particular to a method and device for setting environmental parameters of a plant cultivation box. Background Art

[0004] In traditional plant cultivation boxes, growers usually manually set environmental parameters such as temperature, humidity and light to provide conditions suitable for plant growth.

[0005] However, the requirements for these environmental parameters can vary significantly between different plants or at different growth and development stages of the same plant. For growers who lack relevant knowledge, correctly setting these environmental parameters can be a challenging task.

[0006] Therefore, it is necessary to provide a method for setting environmental parameters of a plant cultivation box, which can meet the growth conditions of plants in the plant cultivation box while reducing the requirements for the management experience of the grower.

[0007] Summary of the Invention

[0008] The present application provides a method and device for setting the environmental parameters of a plant cultivation box, which is used to solve the defects of the existing technology that the environmental parameters are manually set by the grower and the grower needs to have relevant experience. The method automatically adjusts the environmental parameter settings of the plant cultivation box without manual intervention, while meeting the growth conditions of the plants in the plant cultivation box and reducing the requirements for the grower's management experience, making the planting process simpler and more convenient.

[0009] The present application provides a method for setting environmental parameters of a plant cultivation box, comprising:

[0010] Based on the environmental monitoring equipment, obtain the current monitoring values ​​of the environmental parameters in the plant cultivation box;

[0011] Inputting the current monitored values ​​of the environmental parameters into a plant growth model to obtain current growth status data of the plants in the plant cultivation box;

[0012] Inputting the current growth state data into the trained strategy network to obtain the current optimal setting value of the environmental parameter;

[0013] The environmental parameters are set according to the currently optimal setting values.

[0014] In some embodiments, before inputting the current growth status data into the trained strategy network to obtain the current optimal setting value of the environmental parameter, the method further includes:

[0015] Determining a control target for the plant cultivation box;

[0016] Modeling the environmental parameter decision process in the plant cultivation box as a Markov decision process;

[0017] Based on a reinforcement learning algorithm, the Markov decision process is solved to obtain the trained policy network that meets the control objective.

[0018] In some embodiments, modeling the environmental parameter decision process in the plant cultivation box as a Markov decision process includes:

[0019] Using the intermediate quantity output by the plant growth model as the state space in the Markov decision process;

[0020] Using the setting values ​​of the environmental parameters in the plant cultivation box as the action space in the Markov decision process;

[0021] determining a state transition function in the Markov decision process according to the plant growth model;

[0022] A reward function in the Markov decision process is determined based on a control target of the plant cultivation box.

[0023] In some embodiments, solving the Markov decision process based on a reinforcement learning algorithm to obtain the trained policy network that meets the control objective includes:

[0024] Building a reinforcement learning training environment based on the current and predicted growth status data output by the plant growth model;

[0025] In the reinforcement learning training environment, the Markov decision process is solved based on the reinforcement learning algorithm to obtain the trained policy network that meets the control objective.

[0026] In some embodiments, the method further comprises:

[0027] Acquiring growth status monitoring data of the plants in the plant cultivation box based on the phenotypic monitoring equipment;

[0028] The parameters of the plant growth model are calibrated according to the growth status monitoring data.

[0029] The present application also provides a device for setting environmental parameters of a plant cultivation box, comprising:

[0030] A first acquisition module is used to obtain current monitoring values ​​of environmental parameters in the plant cultivation box based on the environmental monitoring equipment;

[0031] a second acquisition module, configured to input the current monitored value of the environmental parameter into a plant growth model to acquire current growth status data of the plants in the plant cultivation box;

[0032] An input module, configured to input the current growth state data into the trained strategy network to obtain the current optimal setting value of the environmental parameter;

[0033] The setting module is used to set the environmental parameters according to the current optimal setting values.

[0034] The present application also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for setting environmental parameters of a plant cultivation box as described above is implemented.

[0035] The present application also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for setting environmental parameters of a plant cultivation box as described in any one of the above is implemented.

[0036] The present application also provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-described methods for setting environmental parameters of a plant cultivation box.

[0037] The environmental parameter setting method and device for a plant cultivation box provided in the present application obtain the current monitoring values ​​of the environmental parameters in the plant cultivation box based on environmental monitoring equipment, input the current monitoring values ​​of the environmental parameters into a plant growth model, obtain the current growth status data of the plants in the plant cultivation box, input the current growth status data into a trained strategy network, obtain the current optimal setting values ​​of the environmental parameters, and set them according to the current optimal setting values ​​of the environmental parameters to achieve automatic adjustment of the environmental parameter settings of the plant cultivation box without manual intervention. While meeting the growth conditions of the plants in the plant cultivation box, the requirements for the management experience of the grower are reduced, making the planting process simpler and more convenient. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the present application or the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0039] FIG1 is a flow chart of a method for setting environmental parameters of a plant cultivation box provided in the present application;

[0040] FIG2 is a schematic diagram of a process for training a strategy network based on a reinforcement learning algorithm provided by the present application;

[0041] FIG3 is a schematic diagram showing the principle of the method for setting environmental parameters of a plant cultivation box provided in the present application;

[0042] FIG4 is a schematic structural diagram of an environmental parameter setting device for a plant cultivation box provided in the present application;

[0043] FIG5 is a schematic structural diagram of the electronic device provided in this application. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in this application will be clearly and completely described below in conjunction with the drawings in this application. Obviously, the described embodiments are part of the embodiments of this application, rather than all the embodiments. It should be noted that, in the absence of conflict, the embodiments of this application and the features in the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0045] FIG1 is a flow chart of a method for setting environmental parameters of a plant cultivation box provided by the present application. As shown in FIG1 , the present application provides a method for setting environmental parameters of a plant cultivation box, comprising the following steps:

[0046] Step 110: Obtain current monitoring values ​​of environmental parameters in the plant cultivation box based on the environmental monitoring equipment.

[0047] Step 120: Input the current monitored values ​​of the environmental parameters into the plant growth model to obtain the current growth status data of the plants in the plant cultivation box.

[0048] Step 130: Input the current growth state data into the trained strategy network to obtain the current optimal setting values ​​of the environmental parameters.

[0049] Step 140: Setting according to the current optimal setting value of the environmental parameter.

[0050] Specifically, an environmental monitoring device (such as a sensor with an environmental monitoring function) is provided in the plant cultivation box, so that the environment of the plant cultivation box can be monitored in real time based on the environmental monitoring device.

[0051] Use environmental monitoring equipment to monitor the plant cultivation box and obtain the current monitoring values ​​of environmental parameters in the plant cultivation box. Environmental parameters include but are not limited to: light intensity, temperature, humidity, and carbon dioxide concentration.

[0052] Subsequently, the current monitored values ​​of the environmental parameters are input into the plant growth model, which outputs data on the current growth status of the plants in the cultivation box. This growth status data includes, but is not limited to, the number of leaflets, number of internodes, number of fruits, leaf weight, petiole weight, internode weight, fruit weight, leaf area, biomass, and demand.

[0053] Then, the current growth state data is input into the trained strategy network, and the trained strategy network outputs the current optimal setting value of the environmental parameter based on the input current growth state data.

[0054] Finally, the current optimal setting value of the environmental parameter is obtained, and the control device is set according to the current optimal setting value of the environmental parameter output by the strategy network.

[0055] The environmental parameter setting method for a plant cultivation box provided in the present application obtains the current monitoring values ​​of the environmental parameters in the plant cultivation box based on environmental monitoring equipment, inputs the current monitoring values ​​of the environmental parameters into a plant growth model, obtains the current growth status data of the plants in the plant cultivation box, inputs the current growth status data into a trained strategy network, obtains the current optimal setting values ​​of the environmental parameters, and sets them according to the current optimal setting values ​​of the environmental parameters, thereby realizing automatic adjustment of the environmental parameter settings of the plant cultivation box without manual intervention, while meeting the growth conditions of the plants in the plant cultivation box and reducing the requirements for the management experience of the grower, making the planting process simpler and more convenient.

[0056] In some embodiments, before inputting the current growth state data into the trained strategy network to obtain the current optimal setting value of the environmental parameter, the method further includes:

[0057] Determine the control objectives of the plant cultivation box;

[0058] The decision-making process of environmental parameters in the plant cultivation box is modeled as a Markov decision process;

[0059] Based on the reinforcement learning algorithm, the Markov decision process is solved to obtain a trained policy network that meets the control objectives.

[0060] Specifically, a control target of the plant cultivation box is determined, for example, the control target is to maximize the yield while minimizing the energy consumption caused by the control.

[0061] The decision-making process for the environmental parameters within the plant cultivation chamber is modeled as a Markov decision process, which frames the problem of setting the environmental parameters within the plant cultivation chamber as a reinforcement learning problem. The setting of these environmental parameters is related to the update frequency of the running policy network. This can be done by refreshing the settings after the policy network runs periodically, such as once a day or every two hours. Alternatively, the settings can be set irregularly, such as retraining the algorithm trainer after accumulating a certain amount of data and updating the policy network irregularly, then running the policy network again to refresh the environmental parameters.

[0062] Using reinforcement learning algorithms (such as Soft Actor Critic (SAC)), the Markov decision process (or reinforcement learning problem) is solved to obtain a trained policy network. This trained policy network must meet the control objectives of the plant cultivation box. The trained policy network can determine the optimal settings for environmental parameters under different growth conditions.

[0063] The environmental parameter setting method of the plant cultivation box provided in the present application determines the control target of the plant cultivation box, abstracts the environmental parameter decision-making process in the plant cultivation box into a Markov decision process, and solves the Markov decision process based on the reinforcement learning algorithm to obtain a trained policy network that meets the control target. The obtained trained policy network is conducive to the subsequent use of the trained policy network to adjust the environmental parameter settings in real time.

[0064] In some embodiments, the environmental parameter decision process in the plant cultivation box is modeled as a Markov decision process, including:

[0065] The intermediate quantity output by the plant growth model is used as the state space in the Markov decision process;

[0066] The setting values ​​of the environmental parameters in the plant cultivation box are used as the action space in the Markov decision process;

[0067] According to the plant growth model, determine the state transition function in the Markov decision process;

[0068] Based on the control objectives of the plant cultivation box, the reward function in the Markov decision process is determined.

[0069] Specifically, the Markov decision process consists of a state space S, an action space A, a state transition function P, a reward function R, and a discount factor γ. These elements together form a mathematical model of the Markov decision process, which is used to describe the decision-making process under uncertainty.

[0070] (1) State space S

[0071] The intermediate variables in the plant growth process can reflect the growth status of the plant, and the state space is used to characterize the growth and development of the plant in a complete growth cycle.

[0072] For leafy crops, the state quantities such as plant growth days, leaf element number, leaf weight, leaf area, produced biomass and required biomass output from the plant growth model are selected as the state space of plant growth.

[0073] For fruit-bearing crops, the state quantities such as plant growth days, number of leaf elements, number of internodes, number of fruits, leaf weight, petiole weight, internode weight, fruit weight, root weight, leaf area, produced biomass and required biomass output from the plant growth model are selected as the state space of plant growth.

[0074] (2) Action space A

[0075] The setting values ​​of the environmental parameters to be optimized are selected as the action space, which is continuous.

[0076] (3) State transition function P

[0077] A plant growth model is selected to simulate crop growth, and its state transition function is determined based on the plant growth model. For example, the plant growth model receives daily temperature, humidity, CO2 concentration, and light intensity information and returns the plant's daily number of leaflets, number of internodes, number of fruits, leaf weight, petiole weight, internode weight, fruit weight, root weight, leaf area, biomass produced, and biomass required.

[0078] (4) Reward function R

[0079] The reward function is designed around the control objectives of the plant cultivation box. For example, if the control objectives are to maximize yield while minimizing energy consumption, the reward function is designed around these two variables. Based on the plant growth model, for leafy crops, the daily increase in leaf weight is generated; for fruiting crops, the daily increase in fruit weight is generated. The increase in leaf weight or fruit weight is multiplied by the unit sales price, and the energy consumption is multiplied by the unit energy price. The reward function is set with net profit as the target.

[0080] Taking fruit-bearing plants as an example, the difference between the daily income from the fruit weight increase and the cost from energy consumption is used as the corresponding reward. k :

[0081] Where, represents the fruit weight increment corresponding to the kth day, represents the fruit weight corresponding to the kth day, represents the fruit weight corresponding to the k-1th day, and All given by the GreenLab model; C k represents the energy consumption corresponding to the kth day; p fruit Indicates the unit selling price of the fruit, p energy Represents the unit energy price.

[0082] (5) Discount factor γ

[0083] The discount factor is used to discount the future state action value to the current moment for calculation. It can be set in advance, for example, it can be set to 0.99, assuming that the future state action value is also important.

[0084] The interaction between the agent and the environment in reinforcement learning can be represented by a Markov decision process. After the agent learns the state of the environment, it takes an action and returns this action to the environment. After the environment learns the agent's action, it enters the next state and passes the next state to the agent. This is how the agent and environment interact in reinforcement learning, making the Markov decision process the fundamental framework of reinforcement learning. The goal of reinforcement learning is to find the optimal policy, a mapping from state to action, that maximizes the cumulative reward. The algorithm trainer models the entire interaction between the agent and the environment in reinforcement learning. It uses sampled data from the Markov decision process to train the reinforcement learning model and regularly updates the model parameters of the policy network.

[0085] The environmental parameter setting method for a plant cultivation box provided in this application models the environmental parameter decision-making process in the plant cultivation box as a Markov decision process by constructing a state space, an action space, a state transfer function and a reward function.

[0086] In some embodiments, the plant growth model is a GreenLab model, an ALMIS model, an L-Peach model, or a LIGNUM model.

[0087] Specifically, the plant growth model is capable of simulating the dynamic growth and development process of different varieties of plants, and outputs different plant growth state data as variables in the state space based on the environmental monitoring data in the plant growth chamber obtained in real time.

[0088] Plant growth models include, but are not limited to, GreenLab model, ALMIS model, L-Peach model, LIGNUM model, etc., or process-based plant growth models.

[0089] In some embodiments, a Markov decision process is solved based on a reinforcement learning algorithm to obtain a trained policy network that meets the control objective, including:

[0090] Build a reinforcement learning training environment based on the growth status data output by the plant growth model;

[0091] In the reinforcement learning training environment, the Markov decision process is solved based on the reinforcement learning algorithm to obtain a trained policy network that meets the control objectives.

[0092] Specifically, the plant growth change process can be learned from the plant growth model. Therefore, a reinforcement learning training environment is built based on the growth status data (including current growth status data and future growth status data) output by the plant growth model.

[0093] After building a reinforcement learning training environment for the plant growth process, we explore the optimal environment parameter settings in the reinforcement learning training environment. After a large amount of training, we obtain a trained policy network. Taking the SAC algorithm as an example, the specific implementation steps of reinforcement learning are as follows:

[0094] Step 1: Initialize plant status.

[0095] Step 2: When the number of iteration steps has not reached the preset starting number, the environmental parameter setting values ​​are randomly sampled from the action space of the constructed Markov decision process. If the number of iteration steps reaches the starting number, the current plant growth state is used as the input of the policy network to obtain the environmental parameter setting values ​​output by the policy network under the current plant growth state.

[0096] Step 3: If the number of samples in the experience replay pool is greater than the batch size required for the update, update the value network, policy network, and the temperature parameter α.

[0097] Step 4: Based on the currently output environmental parameter settings, execute the training environment in the plant growth reinforcement learning environment to obtain the plant's growth status, corresponding reward, and a Boolean value indicating whether the plant is harvested at the next time step. The number of iterations is incremented by 1. The time step can be days or hours, etc.

[0098] Step 5: Store the plant state, environmental parameter setting value, reward, plant growth state at the next time step, and Boolean value of whether the crop is harvested into the experience replay pool, and update the current plant state to the plant growth state at the next time step in step 4.

[0099] Step 6: Determine whether the current crop has been harvested. If not, loop through steps 2 to 5.

[0100] Step 7: Determine whether the set number of iteration steps has been reached. If not, loop through steps 1 to 7.

[0101] Step 8: SAC algorithm training is completed, and the trained policy network is obtained.

[0102] The environmental parameter setting method for the plant cultivation box provided in this application builds a reinforcement learning training environment based on the plant growth model, limits the solution of the Markov decision process to the reinforcement learning training environment, and improves the solution efficiency.

[0103] In some embodiments, the method for setting environmental parameters of a plant cultivation box provided in this application further includes:

[0104] Based on the phenotypic monitoring equipment, the growth status monitoring data of the plants in the plant cultivation box is obtained;

[0105] The parameters of the plant growth model are calibrated based on the growth status monitoring data.

[0106] Specifically, the phenotypic detection device can be a camera (such as an infrared camera) that measures morphological phenotypes (such as the number of leaves and leaf area, etc.), or a device that measures physiological phenotypes (such as chlorophyll content and photosynthetic rate, etc.).

[0107] Based on the phenotypic monitoring equipment, the growth status monitoring data of the plants in the plant cultivation box is obtained. Due to the monitoring capabilities of the phenotypic monitoring equipment, the growth status monitoring data obtained by the phenotypic monitoring equipment is only partial growth status data of the plants in the plant cultivation box.

[0108] The acquired growth status monitoring data includes current and historical growth status monitoring data. The acquired growth status monitoring data is parsed and stored by the phenotypic data parsing module. At certain time steps, the parameters of the plant growth model are calibrated through data fitting and parameter inversion methods to further improve the accuracy of the plant growth model.

[0109] The following is a specific example of lettuce planting to illustrate the method for setting environmental parameters of the plant cultivation box provided by the present application.

[0110] The GreenLab model was used to define lettuce plants. The area occupied by a single plant was set to 0.08 square meters, and the entire growth cycle was set to 60 days, that is, the harvest was 60 days after sowing.

[0111] The daily temperature is set to T set , which remains unchanged throughout the day. Limit the lighting to be turned on from 6:00 to 18:00 every day, set to I set , closed from 18:00 to 6:00 the next day (to provide crops with a certain dark environment to meet their respiratory process), the optimization target is the daily Tset and I set The optimization step is set to 1 day, assuming that the plant's environmental requirements do not vary much during the day. It is assumed that the plant is in a dark environment without the influence of outdoor light, and all light energy comes from the fill light, and all light energy consumption is from I set Produced by I set The energy consumption within a day can be calculated according to the following formula:

[0112] Where, represents the light energy consumption on the kth day, Indicates the light intensity setting value for day k.

[0113] For the temperature loss, T in As a benchmark, it is believed that the ideal temperature is greater than T in Calculate from T in to T set Energy consumption due to additional heating:

[0114] Where, represents the temperature consumption on the kth day, HEC is the conversion coefficient between the temperature difference and the energy consumption caused by heating, represents the temperature setting value for the kth day, T in Indicates the reference temperature.

[0115] Use M leaf represents the leaf weight of lettuce at harvest time, K is the total number of days in the plant growth cycle, and the optimization problem can be expressed as:

[0116] Where C k represents the energy consumption on the kth day, represents the light energy consumption on the kth day, represents the temperature consumption on the kth day, M leaf Indicates the leaf weight of lettuce at harvest time. represents the temperature setting value for the kth day, represents the light intensity setting value on the kth day, HEC is the conversion coefficient between temperature difference and energy consumption caused by heating, H represents humidity, CO2 represents carbon dioxide, and cropModel represents the plant growth model.

[0117] The decision-making process of daily light intensity and temperature setting values ​​during the entire growth period of lettuce is modeled as a Markov decision process.<S,A,P,R,γ> . In the state space S, the number of growing days, leaflet number, leaf weight, leaf area, biomass produced and biomass required of the lettuce output by the GreenLab crop model are used as the state space. The set values ​​of daily temperature and light intensity that need to be optimized are used as the action space A of reinforcement learning, and the upper and lower bounds of the temperature and light intensity in the action space are set. The GreenLab model receives information on temperature, humidity, CO2 concentration and light intensity within a day, and outputs the state quantity of plant growth. Therefore, the state transfer function P is determined by the GreenLab model. The discount factor γ is used to discount the future state action value to the current moment for calculation. In the example, it is set to 0.99, and it is considered that the future state action value is also important.

[0118] The reward function R is set as follows:

[0119] Where, represents the weight increment of lettuce leaves corresponding to the kth day, represents the weight of lettuce leaves corresponding to the kth day, represents the weight of lettuce leaves corresponding to the k-1th day, and All given by the GreenLab model; C k represents the energy consumption corresponding to the kth day; p leaf represents the selling price of lettuce leaves per unit, p energy Represents the unit energy price.

[0120] FIG2 is a flow chart of a reinforcement learning algorithm-based policy network training process provided by the present application. The reinforcement learning SAC algorithm is used to train a policy network for setting the optimal temperature and light intensity throughout the lettuce growth cycle, including the following steps:

[0121] (i) Initialization of lettuce state;

[0122] (ii) When the number of iterations has not reached the starting number, randomly sample the temperature and light intensity setting values ​​from the action space of the constructed Markov decision process. If the number of iterations reaches the starting number, use the current state as the input of the policy network to obtain the temperature and light intensity setting values ​​output by the policy network under the current lettuce growth state;

[0123] (iii) If the number of samples in the experience replay pool is larger than the batch size required for the update, update the value network, policy network and the temperature parameter α;

[0124] (iv) Based on the current output temperature and light intensity settings, execute in the reinforcement learning training environment for lettuce growth, obtain the growth state of the lettuce at the next time step, the corresponding reward, and a Boolean value indicating whether the lettuce is harvested, and increase the number of iteration steps by 1;

[0125] (v) The growth state of the lettuce, the set values ​​of temperature and light intensity, the reward, the growth state of the lettuce in the next time step, and the Boolean value of whether the lettuce is harvested are stored in the experience replay pool, and the current growth state of the lettuce is updated to the growth state of the lettuce in the next time step in (iv);

[0126] (vi) determining whether the current lettuce has been harvested, and if not, looping through steps (ii) to (vi);

[0127] (vii) determining whether the set number of iteration steps has been reached, and if not, looping through steps (i) to (vii);

[0128] (viii) The SAC method is trained and a strategy network is obtained that can determine the optimal set values ​​of temperature and light intensity.

[0129] The trained policy network, which can determine the optimal set values ​​for temperature and light intensity during lettuce growth, was applied to the lettuce cultivation process in the smart plant cultivation box. The specific process is as follows:

[0130] The number of growing days for the lettuce in the smart plant cultivation box is calculated. Current monitoring values ​​of environmental parameters within the smart plant cultivation box (such as temperature, humidity, CO2 concentration, and light intensity) acquired by environmental monitoring equipment are input into the lettuce plant growth model to obtain the current growth status data of the lettuce, namely the number of leaf elements, leaf weight, leaf area, biomass produced, and biomass required. This current growth status data is then input into the trained policy network to determine the current optimal settings for temperature and light intensity.

[0131] The above operations are dynamically performed throughout the entire growth cycle of the lettuce, realizing intelligent control of the plant cultivation box with the goals of high yield and energy saving.

[0132] In addition, the growth status monitoring data of lettuce in the smart plant cultivation box obtained based on phenotypic monitoring is stored and can be further used to calibrate relevant parameters of the GreenLab model.

[0133] FIG3 is a schematic diagram of the principle of the method for setting environmental parameters of the plant cultivation box provided by the present application, which includes the following three aspects:

[0134] The first aspect: monitor the plant cultivation box through environmental monitoring equipment to obtain the monitoring values ​​of environmental parameters, input the monitoring values ​​of environmental parameters into the plant growth model, obtain the current growth status data of the plants in the plant cultivation box, input the current growth status data into the trained strategy network, obtain the optimal setting values ​​of the environmental parameters, and set them according to the optimal setting values ​​of the environmental parameters through the control equipment.

[0135] Second aspect: The plant growth model generates a large amount of current and future growth status data, which is input into the algorithm trainer for algorithm training. The algorithm trainer then updates the strategy network from time to time.

[0136] The third aspect: monitor the plant cultivation box through the phenotypic monitoring equipment to obtain the growth status monitoring data of the plants in the plant cultivation box, and calibrate the plant growth model after parsing the growth status monitoring data through the phenotypic data interpretation module.

[0137] The environmental parameter setting device for the plant cultivation box provided in the present application is described below. The environmental parameter setting device for the plant cultivation box described below and the environmental parameter setting method for the plant cultivation box described above can be referenced to each other.

[0138] FIG4 is a schematic diagram of the structure of the environmental parameter setting device of the plant cultivation box provided by the present application. As shown in FIG4 , the present application provides an environmental parameter setting device of the plant cultivation box, comprising:

[0139] A first acquisition module 410 is configured to acquire current monitoring values ​​of environmental parameters in the plant cultivation box based on environmental monitoring equipment;

[0140] A second acquisition module 420 is configured to input the current monitored value of the environmental parameter into a plant growth model to obtain current growth status data of the plants in the plant cultivation box;

[0141] An input module 430 is used to input the current growth state data into the trained strategy network to obtain the current optimal setting value of the environmental parameter;

[0142] The setting module 440 is configured to set the environmental parameters according to the current optimal setting values.

[0143] In some embodiments, the apparatus further comprises:

[0144] a determination module, configured to determine a control target of the plant cultivation box;

[0145] A construction module is used to model the environmental parameter decision process in the plant cultivation box as a Markov decision process;

[0146] A solution module is used to solve the Markov decision process based on a reinforcement learning algorithm to obtain the trained policy network that meets the control objective.

[0147] In some embodiments, the building blocks are specifically configured to:

[0148] Using the intermediate quantity output by the plant growth model as the state space in the Markov decision process;

[0149] Using the setting values ​​of the environmental parameters in the plant cultivation box as the action space in the Markov decision process;

[0150] determining a state transition function in the Markov decision process according to the plant growth model;

[0151] A reward function in the Markov decision process is determined based on a control target of the plant cultivation box.

[0152] In some embodiments, the solution module is specifically configured to:

[0153] Building a reinforcement learning training environment based on the growth status data output by the plant growth model;

[0154] In the reinforcement learning training environment, the Markov decision process is solved based on the reinforcement learning algorithm to obtain the trained policy network that meets the control objective.

[0155] In some embodiments, the apparatus further comprises:

[0156] a third acquisition module, configured to acquire growth status monitoring data of the plants in the plant cultivation box based on the phenotypic monitoring device;

[0157] A calibration module is used to calibrate the parameters of the plant growth model according to the growth status monitoring data.

[0158] It should be noted here that the environmental parameter setting device for the plant cultivation box provided in this application can implement all the method steps implemented in the above-mentioned method embodiment and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.

[0159] FIG5 is a schematic diagram of the structure of an electronic device provided by the present application. As shown in FIG5 , the electronic device may include: a processor 510, a communication interface 520, a memory 530, and a communication bus 540. The processor 510, the communication interface 520, and the memory 530 communicate with each other via the communication bus 540. The processor 510 may call the logic instructions in the memory 530 to execute the method for setting the environmental parameters of the plant cultivation box, which includes:

[0160] Based on the environmental monitoring equipment, the current monitoring values ​​of the environmental parameters in the plant cultivation box are obtained; the current monitoring values ​​of the environmental parameters are input into the plant growth model to obtain the current growth status data of the plants in the plant cultivation box; the current growth status data is input into the trained strategy network to obtain the current optimal setting values ​​of the environmental parameters; and the settings are performed according to the current optimal setting values ​​of the environmental parameters.

[0161] In addition, the logic instructions in the above-mentioned memory 530 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0162] On the other hand, the present application also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can perform the environmental parameter setting method of the plant cultivation box provided by the above methods, which includes:

[0163] Based on the environmental monitoring equipment, the current monitoring values ​​of the environmental parameters in the plant cultivation box are obtained; the current monitoring values ​​of the environmental parameters are input into the plant growth model to obtain the current growth status data of the plants in the plant cultivation box; the current growth status data is input into the trained strategy network to obtain the current optimal setting values ​​of the environmental parameters; and the settings are performed according to the current optimal setting values ​​of the environmental parameters.

[0164] In another aspect, the present application further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the environmental parameter setting method for a plant cultivation box provided by each of the above methods, the method comprising:

[0165] Based on the environmental monitoring equipment, the current monitoring values ​​of the environmental parameters in the plant cultivation box are obtained; the current monitoring values ​​of the environmental parameters are input into the plant growth model to obtain the current growth status data of the plants in the plant cultivation box; the current growth status data is input into the trained strategy network to obtain the current optimal setting values ​​of the environmental parameters; and the settings are performed according to the current optimal setting values ​​of the environmental parameters.

[0166] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0167] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0168] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application pertains. The terms used herein in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0169] It should be further noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0170] In this application, "at least one" means one or more, and "a plurality" means two or more than two. The terms "first," "second," "third," "fourth," etc. (if any) in this application are used to distinguish similar objects, rather than to describe a particular order or sequence.

[0171] In the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0172] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for setting environmental parameters of a plant cultivation box, comprising: Based on the environmental monitoring equipment, obtain the current monitoring values ​​of the environmental parameters in the plant cultivation box; Inputting the current monitored values ​​of the environmental parameters into a plant growth model to obtain current growth status data of the plants in the plant cultivation box; Inputting the current growth state data into the trained strategy network to obtain the current optimal setting value of the environmental parameter; The environmental parameters are set according to the currently optimal setting values.

2. The method for setting environmental parameters of a plant cultivation box according to claim 1, wherein: Before inputting the current growth state data into the trained strategy network to obtain the current optimal setting value of the environmental parameter, the method further includes: Determining a control target for the plant cultivation box; Modeling the environmental parameter decision process in the plant cultivation box as a Markov decision process; Based on a reinforcement learning algorithm, the Markov decision process is solved to obtain the trained policy network that meets the control objective.

3. The method for setting environmental parameters of a plant cultivation box according to claim 2, wherein: The process of modeling the environmental parameter decision process in the plant cultivation box as a Markov decision process includes: Using the intermediate quantity output by the plant growth model as the state space in the Markov decision process; Using the setting values ​​of the environmental parameters in the plant cultivation box as the action space in the Markov decision process; determining a state transition function in the Markov decision process according to the plant growth model; A reward function in the Markov decision process is determined based on a control target of the plant cultivation box.

4. The method for setting environmental parameters of a plant cultivation box according to claim 2, wherein: The step of solving the Markov decision process based on a reinforcement learning algorithm to obtain the trained policy network that satisfies the control objective includes: Based on the growth status data output by the plant growth model, a reinforcement learning training loop is built. territory; In the reinforcement learning training environment, the Markov decision process is solved based on the reinforcement learning algorithm to obtain the trained policy network that meets the control objective.

5. The method for setting environmental parameters of a plant cultivation box according to claim 1, wherein: The method further comprises: Acquiring growth status monitoring data of the plants in the plant cultivation box based on the phenotypic monitoring equipment; The parameters of the plant growth model are calibrated according to the growth status monitoring data.

6. A device for setting environmental parameters of a plant cultivation box, comprising: A first acquisition module is used to obtain current monitoring values ​​of environmental parameters in the plant cultivation box based on the environmental monitoring equipment; a second acquisition module, configured to input the current monitored value of the environmental parameter into a plant growth model to acquire current growth status data of the plants in the plant cultivation box; An input module, configured to input the current growth state data into the trained strategy network to obtain the current optimal setting value of the environmental parameter; The setting module is used to set the environmental parameters according to the current optimal setting values.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for setting environmental parameters of the plant cultivation box according to any one of claims 1 to 5 is implemented.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method for setting environmental parameters of a plant cultivation box according to any one of claims 1 to 5 is implemented.

9. A computer program product comprising a computer program, wherein When the computer program is executed by a processor, the method for setting environmental parameters of a plant cultivation box according to any one of claims 1 to 5 is implemented.

Images